Assessing Intergovernmental Fiscal Relations Financing Higher Education Controlling Health Care Expenditures Managing Fiscal Risks in Public-Private Partnerships

VOLUME I: MAIN REPORT VOLUME II: COUNTRY CASE STUDIES (CD)

Contents FOREWORD ...... 7 Daniela Gressani CONTRIBUTORS ...... 9 CONFERENCE AGENDA ...... 10 OVERVIEW ...... 11 Thomas Laursen I. FISCAL CHALLENGES FOR THE EU8 COUNTRIES ...... 27 Paulina Bucon and Emilia Skrok 1. FISCAL DEVELOPMENTS 1995-2004 ...... 27 2. MEDIUM-LONG TERM FISCAL PROSPECTS ...... 43 3. QUALITY OF FISCAL POLICY ...... 46 II. INTERGOVERNMENTAL FISCAL RELATIONS ...... 58 William Dillinger 1. INTRODUCTION ...... 58 2. EVALUATION CRITERIA ...... 58 3. THE SYSTEM EX ANTE ...... 59 4. POLITICAL AND ORGANIZATIONAL REFORMS ...... 60 5. ASSIGNMENT OF FUNCTIONS ...... 62 6. REVENUES ...... 66 7. DEBT CONTROLS ...... 73 8. DIRECTIONS FOR FURTHER REFORM ...... 75 III. FINANCING HIGHER EDUCATION ...... 80 Mary Canning, Martin Godfrey, and Dorota Holzer-Zelazewska 1. INTRODUCTION ...... 80 2. HIGHER EDUCATION SYSTEMS IN A COMPARATIVE PERSPECTIVE ...... 81 3. DIRECTIONS FOR FURTHER REFORM ...... 93 4. CONCLUSIONS ...... 96 IV. CONTROLLING HEALTH EXPENDITURES ...... 99 Mukesh Chawla 1. INTRODUCTION ...... 99 2. HEALTH EXPENDITURES ...... 103 3. HEALTH SECTOR DEFICITS AND DEBTS ...... 107 4. KEY EXPENDITURE AREAS IN THE HEALTH SECTOR ...... 109 5. SPENDING ON AGEING POPULATIONS AND MEDICAL TECHNOLOGY ...... 114 6. DIRECTIONS FOR FURTHER REFORM ...... 117 7. CONCLUSIONS ...... 129 V. MANAGING FISCAL RISKS IN PUBLIC–PRIVATE PARTNERSHIPS ...... 135 Hana Polackova Brixi, Nina Budina, and Timothy Irwin 1. INTRODUCTION ...... 135 2. THE FISCAL EFFECTS OF PPPS ...... 137 3. HOW FISCAL INSTITUTIONS AFFECT THE FISCAL COST OF PPPS ...... 143 4. DIRECTIONS FOR FURTHER REFORM ...... 147 5. CONCLUSIONS ...... 154 SUMMARY OF CONFERENCE DISCUSSIONS ...... 157 Thomas Laursen Boxes Box 01. Objectives of Intergovernmental Finance Systems ...... 17 Box I.1. EU8 Tax Reforms: ...... 33 Figure A. EU8 Effective Corporate Income Tax Rates (macro backward-looking approach)...... 33 Table A. Top statutory tax rate on personal income ...... 34 Table B. Top statutory tax rate on corporate income ...... 34 Table C. VAT rates ...... 34 Box I.2. Public Spending Measurement Issues ...... 36 Box I.3. Factors Affecting Expenditure Developments ...... 36 Figure A. Public Expenditure and Output Gap (%) ...... 37 Figure B.1. Political Economy at Work in the Baltic Countries ...... 37 Figure B.2. Political Economy at Work in the Visegrad Countries ...... 37 Box I.4. Case Study: Ireland ...... 47 Box I.5. Case Study: Slovakia ...... 48 Figure A. GG expenditure and GDP growth (% of GDP and %, respectively) ...... 49 Figure B. GG Revenue, Expenditure and Balance (% of GDP) ...... 49 Figure C. Composition of GG Revenues (% of GDP) ...... 49 Figure D. Composition of GG Expenditures (% of GDP) ...... 49 Box I.6. EBA Methodology ...... 53 Box II.1. A Snapshot of the Economics of Expenditure Assignment ...... 59 Box II.2. Hungary’s Municipal Bankruptcy Procedure ...... 75 Box II.3. Calculating the Formula Grant in England and Wales ...... 76 Box IV.1. Expenditure on Drugs (OECD countries; 1998-2003) ...... 110 Box IV.2. Health Reforms in Slovakia ...... 118 Box IV.3. Managing Expenditure on Drugs – the Experience of Slovakia ...... 121 Box V.1. How Does the Use of PPPs Differ by Sector? ...... 137 Box V.2. PPPs in the EU8 ...... 139 Box V.3. Building Risk Management Capacity in Turkey ...... 146 Tables Table I.1. Effect of Pension Reforms in the EU8 (% of GDP) ...... 29 Table I.2. EU8 Tax Burden 1995-2004 (% of GDP) ...... 32 Table I.3. EU8 Economic Classification of Public Expenditure 2003 (% of GDP) ...... 39 Table I.4. EU8 Functional Classification of Public Expenditure 2003 (% of GDP) ...... 39 Table I.5. Snow Ball Effect on Debt (% GDP) ...... 41 Table I.6. EU8 Expenditure and Revenue at Central and Local Government Level 2003 ...... 42 Table I.7. EU8 Debt Stabilizing Primary Fiscal Balance 2004 (% of GDP) ...... 44 Table I.8. EU8 Medium-Long Term Spending Projections (% of GDP) ...... 44 Table I.9. EU8 Sustainability Gap Indicators (% of GDP) ...... 46 Table I.10. EU8 Patterns of Large Fiscal Contractions 1998-2004 ...... 47 Table I.11. Composition of Public Spending in 2002 – Ranking ...... 51 Table I.12. EU8 Panel Growth Regressions 1995-2004 (dependent variable: growth rate of per capita GDP) ...... 54 Table I.13. Robustness of Base Variables ...... 54 Table I.14. Tax Structure and Output Growth (shares of total taxes) ...... 55 Table I.15. Expenditure Structure and Output Growth ...... 56 Table II.1. Timetable of Post-Soviet Reforms ...... 60 Table II.2. Local Share of Total Public Education Spending ...... 63 Table II.3. Subnational Spending on Health and Social Assistance (% of total public sector spending on these functions) ...... 64 Table II.4. Personal Income Tax Shared with Subnational Governments (2002) ...... 67 Table II.5. Property Taxes (% of GDP) ...... 69 Table II.6. Summary of Municipal Debt Regulations ...... 74 Table III.1. Higher Education Indicators (EU15 and EU8 countries) ...... 83 Table III.2. Tertiary Graduates in Science and Technology Employed as Professionals or Technicians (2003) ...... 84 Table III.3. Higher Education and the Labor Market (EU8 and EU15 countries) ...... 84 Table III.4. Public Financing of Higher Education (EU8 and EU15 countries, 2001) ...... 85 Table III.5. Financial Aid to Students (EU8 and EU15 countries, 2001) ...... 86 Table III.6. Higher Education Expenses Borne by Parents and Students (first degree, public institutions: Poland, Hungary and Latvia; US$ PPP) ...... 88 Table III.7. Poland: Enrollment Rate by Age Group and Household Consumption Quintile (2003) .... 88 Table III.8. Extent of Autonomy Enjoyed by Universities (selected EU15 and EU8 countries) ...... 91 Table III.9. Appointment of Leaders of Universities (selected EU15 and EU8 countries) ...... 92 Table IV.1. EU8 – Standardized Death Rates (per 100,000); different causes (2003) ...... 101 Table IV.2. EU8 – Scope of Services Covered by Social Health Insurance ...... 104 Table IV.3. Health Sector Costs in Slovakia (1998-2003; percent of total) ...... 105 Table IV.4. Lithuania – Debt of Health Insurance Fund to Providers (billion Litas) ...... 109 Table IV.5. Beds per 100,000 Inhabitants (EU8; 1993-2002) ...... 112 Table IV.6. Physician Salaries (selected EU8 countries; average monthly in local currency) ...... 113 Table V.1. Selected Liabilities from PPPs and Other Privately Financed Infrastructure Projects ...... 140 Table V.2. Possible Real Fiscal Effects of PPPs on Net Worth ...... 141 Table V.3. Improving Fiscal Institutions for PPPs at a Glance ...... 154 Figures Figure O1. Employment and Relative Poverty Risk in the EU (2001 or latest data) ...... 11 Figure O2. Employment and Debt in the EU (2004) ...... 12 Figure O3. Size of Public Sector and Per Capita Income in the EU (2004) ...... 12 Figure O4. Tax wedge and Employment Rate in the EU ( 2004) ...... 13 Figure O5. Fiscal Decentralization and Real GDP Growth in the EU8 1995-2003 ...... 14 Figure I.1. EU8 Fiscal Developments 1998-2004 (% of GDP) ...... 28 Figure I.2. EU8 General Government Revenue and Expenditure (% of GDP) ...... 30 Figure I.3. Visegrad Countries - Divergence from EU Program Fiscal Targets (% of GDP) ...... 31 Figure I.4. EU8 Change in Tax Burden 1995-2004 Relative to Initial Level ...... 31 Figure I.5. EU8 Composition of Tax Burden 1995-2004 (% of GDP) ...... 33 Figure I.6. EU8 Tax Wedge on Labor Costs - Low Wage Earners ...... 35 Figure I.7. General government, total revenue and grants (% GDP) ...... 35 Figure I.8. EU8 Public Expenditure Composition (% of GDP) ...... 38 Figure I.9. General government, total expenditure (% GDP) ...... 40 Figure I.10. EU8 Hidden Fiscal Deficit 2000-03 (% of GDP) ...... 41 Figure I.11. EU8 Balances at Various Levels of Government 2001-2003 ...... 42 Figure I.12. EU8 Fiscal Plans 2005-07 (% or % of GDP) ...... 43 Figure I.13. EU8 Long-Term Public Debt Projections (% of GDP) ...... 45 Figure I.14. Tax Ratio and Output Growth (for all EU8, average 1995-2003) ...... 49 Figure I.15. Theoretical Efficiency of Various Categories of Government Expenditure ...... 50 Figure II.1. Structure of Local Government Expenditures as Shares of GDP ...... 62 Figure II.2. Structure of Local Government Revenues as % GDP ...... 66 Figure III.1. Gross Enrollment Rates in Tertiary Education (EU8 & selected EU15 countries, .1990/91 – 2002/2003) ...... 82 Figure III.2. Share of University Students in Public and Private Institutions (EU8 countries, 2002) .... 86 Figure III.3. Chances of Obtaining Higher Education by Family Background (indicated by father’s occupation: the Czech Republic, Poland, Hungary, Finland, and the USA, 1998; ratios for individuals aged 20-35) ...... 89 Figure IV.1. GDP and Health Expenditures, Selected EU, OECD and Other Countries (2002 or latest available year) ...... 99 Figure IV.2. Expected Years Spent in Poor Health (Selected EU, OECD and other Countries) ...... 100 Figure IV.3. Health Expenditures in EU8 Countries, Percent of GDP, 2003 ...... 103 Figure IV.4. Public and Private Expenditures on Health in Latvia ...... 103 Figure IV.5. Structure of Health Expenditures in Poland, 1999-2003 ...... 105 Figure IV.6. Structure of Public Expenditures on Health in Hungary ...... 106 Figure IV.7. Structure of Public Expenditures on Health in Latvia ...... 106 Figure IV.8. Structure of Health Expenditures, OECD average (2001) ...... 107 Figure IV.9. Fiscal Position of the Health System, Hungary ...... 107 Figure IV.10. Fiscal Position of the Slovak Health System (percent of GDP) ...... 108 Figure IV.11. Structure of Hospital Debts in Poland ...... 108 Figure IV.13. Total Pharmaceutical Expenditure in Latvia (2003 prices) ...... 110 Figure IV.14. EU8 Actual and Projected Total Fertility Rate (TFR) ...... 114 Figure IV.15. EU8 Actual and Projected Life Expectancy at Birth ...... 114 Figure IV.16. EU8 Countries: Population over 60, 1995-2025 ...... 115 Figure IV.17. Diffusion of MRI Units and CT Scanners, Selected Countries ...... 116 Figure IV.18. Price and Volume of Pharmaceuticals, Selected Countries, 2002 ...... 119 Figure IV.19. Expenditures on Drugs in Slovakia ...... 121 Figure V.1. Comparing Public Finance to a PPP with a Long-Term Purchase Contract ...... 142

Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 7 FOREWORD

The new Central European and Baltic members of the European Union – the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, and Slovenia, henceforth the “EU8 countries” – all face the challenge of creating greater fiscal space for needed development spending, not least on infrastructure, to accelerate the process of real income convergence with the incumbent members of the EU.1 At the same time, they are under increasing pressure to improve the quality and efficiency of public service delivery, both to strengthen the formation of human capital needed to support high and sustainable growth and to ensure broad and equitable opportunities for all citizens.

Fiscal space is constrained first of all by the need to maintain sustainable fiscal positions in the context of the Economic and Monetary Union, but also by limited absorptive capacity for additional, productive public spending. The Stability and Growth Pact requires EU members to maintain at least balanced budgets over the economic cycle, and to contain fiscal deficits below 3% of GDP and debt levels below 60% of GDP, albeit with some flexibility. These thresholds are also key conditions for the adoption of the Euro by the new member states. Finally, the Visegrad countries (the Czech Republic, Hungary, Poland, and Slovakia) need to reduce excessive fiscal deficits to meet the requirements for joining the monetary union.

At the same time, fiscal space is constrained by existing pressures on public expenditures and fiscal revenues. Social spending is already compressed in some EU8 countries, and most of the EU countries face significant spending pressures over the medium-long term related to population aging (pensions, health, etc.). Public sectors also remain relatively large given income levels in some of these countries, and tax burdens may need to be reduced further to enhance efficiency, competitiveness, and employment in the EU.

Under these conditions, fiscal space can generally only be created through rationalizing existing spending programs. While some progress has been made in most EU8 countries, there is strong evidence of remaining inefficient or even wasteful expenditures, and – although politically difficult – curtailing such spending would clearly be justified on economic grounds. Reviewing persisting subsidies to certain sectors or enterprises is the most obvious place to start, as it would result in increased fiscal space as well as reduced distortions. But the biggest potential for savings in most countries lies in the social spending area where large inefficiencies are apparent. In most countries, the scope for creating fiscal space on the revenue side is limited but could be achieved through broadening tax bases and sources.

The aim of this study is to examine more specific ways of creating fiscal space in the EU8 countries. It considers ways to enhance the efficiency of key spending programs and bring in additional private financing, while supporting the broader growth process and social equity and governance concerns. The study is not meant to be a comprehensive review of fiscal policy in the EU8 countries, but rather to address a limited number of selected, priority issues. This study is also not meant to include in-depth analysis of country specific systems and specific country-oriented policy recommendations, but rather – given the commonality of issues across the EU8 countries as well as other countries in the region – to provide a cross-country comparative perspective, convey existing knowledge and current thinking on reform issues, and distill lessons from reforms both within and outside of the region as well as best practices where such can be identified. Finally, this study is part of a larger, multi-year effort that will continue to focus on priority issues for fiscal reform in the EU8 countries.

The study was prepared in close partnership with governments and non-government experts in the EU8 countries. The priority issues were selected in consultation with Finance Ministers from the region, and government officials from the various countries – as well as local experts and consultants – were closely involved in the work through the provision of information, background analysis, and discussion of work in progress. International experts were also involved either as direct contributors or peer reviewers. Early drafts of the study were discussed at the technical level in a series of regional thematic workshops, and the final overall study was discussed at a conference in Prague hosted jointly by the Ministry of Finance of the Czech Republic and the World Bank on November 15-16, 2005. We hope that government officials from the EU8 countries can benefit from the study through the associated dissemination of knowledge and analytical

1 “In its broadest sense, fiscal space can be defined as the availability of budgetary room that allows a government to provide resources for a desired purpose without any prejudice to the sustainability of a government’s financial position” (Heller, 2005: Understanding Fiscal Space, IMF Policy Discussion Paper, March). 8 capacity building, and also that policy-makers will find the study useful as they consider various reform alternatives. We also hope that other countries in the region as well as international partner organizations such as the European Commission will find the study relevant and useful.

This study includes a review of the current fiscal performance and prospects in the EU8 countries; intergovernmental fiscal relations; higher education financing; health care financing; and fiscal risks related to public-private partnerships. The main message of the study is that much progress has been made in placing fiscal policy on a more efficient and sustainable path in the process leading to EU accession, but that there are opportunities for a stronger orientation of fiscal policy toward growth with equity. In the social spending areas, the study argues for enhancing competition within adequate regulatory and quality control frameworks; increasing the role of private financing; improving incentives on both the supply and demand sides of the market; and providing more targeted support for low-income groups. Such reforms would not only enhance efficiency and fiscal sustainability, but also provide for more equitable access to key social services and strengthen governance in major institutions. In the institutional areas, the study argues for institution building to play a major role in managing fiscal risks from PPPs and in making intergovernmental fiscal relations more stable and predictable. For PPPs, this includes measures to enhance risk analysis, disclosure, accounting and management (including medium-long term fiscal projections). For intergovernmental fiscal relations, it includes measures to ensure adequate funding of central government mandated sector specific spending as well as greater security in the funding of discretionary expenditures.

The World Bank stands ready to further support the EU8 countries in these endeavors.

Daniela Gressani Country Director, Central Europe and the Baltic Countries Europe and Central Asia Department World Bank

Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 9 CONTRIBUTORS Authors: Hana Polackova Brixi, Senior Economist, World Bank (Beijing) Nina Budina, Senior Economist, World Bank (Washington, DC) Paulina Bucon, Research Assistant, World Bank (Warsaw) Mary Canning, Lead Education Specialist, World Bank (Warsaw) Mukesh Chawla, Lead Economist, World Bank (Washington, DC) William Dillinger, Lead Public Sector Management Specialist, World Bank (Washington, DC) Martin Godfrey, Consultant, World Bank (Washington, DC) Dorota Holzer-Zelazewska, Research Analyst, World Bank (Warsaw) Timothy Irwin, Senior Economist, World Bank, (Washington, DC) Thomas Laursen, Lead Economist, World Bank (Warsaw) Emilia Skrok, Economist, World Bank (Warsaw) Peer reviewers: Nicholas Barr, Professor of Public Economics, London School of Economics, UK Paolo Belli, Senior Economist (Health), World Bank (Washington, DC) Andrew Burns, Senior Economist, World Bank (Washington, DC) Kenneth Davey, International Development Department, University of Birmingham, Emeritus Professor of Development Administration Imre Hollo, Economist, European Investment Bank Andrzej Rys, Director, JAGIELLONIAN UNIVERSITY, Centre for Innovation, Technology Transfer, and University Development, Poland Jamil Salmi, Sector Manager, World Bank (Washington, DC) Allen Schick, Professor, Maryland School of Public Affairs, USA Sergei Shatalov, Lead Country Economist, World Bank (Washington, DC) Drahomira Vaskova, Deputy Director, Ministry of Finance, Czech Republic, Max Watson, Wolfson College, Oxford, Deborah Wetzel, Sector Manager, World Bank (Washington, DC) Conference discussants and session chairmen: Jaak Aaviksoo, Rector of the University of Tartu, Professor of Optics and Spectroscopy, Estonia Lucie Antosova, Institute of Economic Studies, Charles University, Prague Vladimir Bezdek, Advisor to the Bank Board, Czech National Bank Cheryl Gray, Sector Director, Poverty Reduction and Economic Management, Europe and Central Asia region, World Bank Roger Grawe, Senior Advisor, Roma Education Fund; former Country Director for Central Europe and the Baltic States, Europe and Central Asia region, World Bank Daniela Gressani, Country Director for Central Europe and the Baltic States, Europe and Central Asia region, World Bank Stepan Jurajda, Deputy Director for Research, CERGE-EI, Prague Filip Keereman, Head of Unit, Member States V: Czech Republic, Lithuania, Poland, Slovenia, Slovakia (ECFIN. B.5), European Commission, DG Economic and Financial Affairs Thomas Laursen, Lead Economist for Central Europe and the Baltics, Europe and Central Asia region, World Bank Danny Leipziger, Vice President and Head of Poverty Reduction and Economic Management Network, World Bank Tom McCarthy, Chief Executive, Irish Management Institute, Ireland Witold Orlowski, Professor, Warsaw School of Economics; former Advisor to the President of Poland Gabor Peteri, Director of Development, LGID, Hungary Tomas Prouza, Deputy Minister of Finance, Czech Republic Andrzej Rys (see above) Ondrej Schneider, Deputy Head of the Institute for Economic Studies, Head of the Department of European Economic Integration and Economic Policy, Faculty of Social Science, Charles University, Czech Republic Paul Bernd Spahn, Macro Fiscal Advisor, ITA Regional Center, Bosnia and Herzegovina Jan Svejnar, Professor of Economics, University of Michigan, USA Sweder van Wijnbergen, Professor of Economics, Universiteit van Amsterdam, the Netherlands Drahomira Vaskova, Deputy Director, Ministry of Finance, Czech Republic Andreas Woergoetter, Head of Country Studies V Division, Economics Department, OECD Miroslav Zamecnik, Consultant, Czech Republic 10 CONFERENCE AGENDA Prague, November 15-16, 2005

November 15:

20.00-22.00 Welcome Banquet Hrzansky Palace (hosted by Finance Minister Sobotka).

November 16:

8.30 Registration

9.00 Welcome Remarks (Daniela Gressani and Tomas Prouza)

9.15 Introductory remarks (Tomas Prouza and Danny Leipziger)

9.45 Fiscal Background (Emilia Skrok and Paulina Bucon) Discussant: Jan Svejnar Discussant: Filip Keereman Chair: Thomas Laursen 10.30 Coffee

11.00 Assessing Intergovernmental Fiscal Relations (William Dillinger). Discussant: Gabor Peteri Discussant: Paul Bernd Spahn Chair: Roger Grawe

12.00 Managing Fiscal Risks of PPPs (Nina Budina) Discussant: Sweder van Wijnbergen Discussant: Vladimir Bezdek Chair: Drahomira Vaskova 13.00 Lunch

14.00 Financing Higher Education (Mary Canning). Discussant: Jaak Aaviksoo Discussant: Tom McCarthy Chair: Stepan Jurajda 15.00 Coffee

15.30 Controlling Health Expenditures (Mukesh Chawla) Discussant: Andrzej Rys Discussant: Miroslav Zemecnik Chair: Lucie Antosova

16.30 Concluding panel discussion Andreas Woergoetter Ondrej Schneider Witold Orlowski Chair: Cheryl Gray

17.30 Closing remarks (Drahomira Vaskova and Thomas Laursen). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 11

OVERVIEW2 i. EU8 countries continue – to varying extents – to be affected by their socialist legacies in terms of their economic and social systems. Generally, the social models of these countries are characterized by relatively low employment and efficiency, with employment rates ranging from a low of 52 percent in Poland to a high of 64 percent in the Czech Republic, still below the EU15 average.3 Equity, on the other hand, varies significantly among the EU8 countries, with relative “poverty rates” below the EU15 average in the Czech Republic, Hungary, and Slovenia, but significantly higher in Slovakia, Poland, and the Baltic countries.4 Following Sapir (2005), this places the Czech Republic, Hungary, and Slovenia among the European countries with low efficiency and high equity, similar to most continental economies (including France and Belgium), while the other EU8 countries are characterized by low efficiency and low equity, similar to the Mediterranean economies (including Spain, Italy, and Greece) (Figure O1). This is in sharp contrast to the high efficiency-high equity models of the Nordic countries.5

Figure O1. Employment and Relative Poverty Risk in the EU (2001 or latest data). 94

92 CZ SE 90 HU SI FI LU Nord ) 88 AT NL %

( Cont

e BE DE t

a FR r 86 y t r EU8 e

v EU15 o P

- 84 PL LV 1 LT 82 EE IT Med ES UK Anglo 80 GR PT SK IE 78 50 55 60 65 70 75 80 Employment rate (%) Notes: Poverty rate is the “risk” of relative poverty (measured as the share of the population with income, after social transfers, below 60% of median). Data for EU15, EU8, Anglo, Cont, Nord and Med are calculated as simple averages. LV (2002); SK (2003). Source: Eurostat. ii. While, this report relies mostly on the EU countries as main comparators, this may not always be the best benchmarks. The structural public finance and labor market problems in Europe are well known, and EU8 countries may also want to look beyond this region for guidance on key economic policy issues and best practices, including to other OECD countries and fast-growing emerging markets. Nevertheless, with recent EU accession, participation in the union’s broader economic and structural policy guidelines, and EU-wide data comparability, Europe is the natural first place for comparison. iii. The ultimate objective of economic reforms in general and social and labor market reforms in particular in the EU8 countries should be to move towards the NE corner of the efficiency-equity diagram. The experience and models of European countries suggests that there is no inherent trade-off between these two objectives, at least in the longer run (Schuknecht and Tanzi, 2000).6 Further, the EU8 countries are not likely to be close to the frontier where such a trade-off may exist. The aim of the World Bank’s programmatic fiscal reform analysis in the region is to discuss how public finance and social reforms can help make a push in the right direction while creating fiscal space for additional productive, growth enhancing public spending (and potentially tax

2 Prepared by Thomas Laursen. 3 Efficiency is characterized here is providing adequate incentives to work. Employment is measured as the share of employed 15-64 year olds to total population in that age group. 4 Poverty rates measured as the share of population with incomes below 60% of the median (as a proxy for the risk of falling into poverty). 5 Boeri (2003) refers to the “fatal attraction” of Western Europe in shaping social policies in Central and Eastern Europe. 6 This contrasts with earlier work that argued for a clear – and possibly increasing – trade-off between equity and efficiency (Okun, 1975; and de Mooj and Tang 2004). 12 reforms) within tight fiscal constraints and the need to ensure fiscal sustainability. As pointed out by Sapir (op cit.), efficiency is also a necessary condition for long-term sustainability. This claim is tentatively supported by the negative relationship between employment and debt in the EU, although the correlation is not strong, there is significant variation, and all the usual caveats to such bivariate correlation analysis apply (Figure O2).

Figure O2. Employment and Debt in the EU (2004).

115 IT GR 105

BE 95

85

DP 75 o G FI

f FR 65 DE s%

ba HU PT EU15 d 55 SE t NL

GG e ES 45 PL AT SK UK 35 CZ EU8 SI IE 25 LT 15 50 55 60 65 70 75 Employment rate (%) Source: Eurostat. iv. Fiscal space in the EU8 countries is needed to increase investment in physical – notably infrastructure – and human capital to raise the countries’ long-run growth potential and thus ensure relatively rapid convergence of real incomes and living standards with those prevailing in the rest of the EU. While education levels are generally high in the EU8, the structure of employment and production suggests that strengthening the quality of human capital remains an important challenge. Similarly, while much progress has been made, infrastructure indicators continue to compare unfavorably with those of Western Europe. Fiscal space is also required to cofinance large and increasing inflows of EU regional aid funds, the absorption of which also will be contingent on stronger technical and administrative capacity of all levels of government. Further, some EU8 countries (notably Hungary and Slovenia) retain relatively large public sectors given their income levels, and in most EU8 countries the tax burden on labor remains very high, discouraging employment (Figures O3 and O4). Thus, a further lowering of the tax burden – not least on labor – is also desirable to stimulate investment and employment growth.

Figure O3. Size of Public Sector and Per Capita Income in the EU (2004). 60 SE

55 AT

BE 50 FI FR

DP SI IT EU15 45 HU PT NL s GR DE LU a of G PL CZ 40 UK

e EU8 v EE ES eue% n LV SK

GG r 35 IE

LT 30

25 0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 GDP per cap ita PPS Source: Eurostat. v. EU8 countries face tight fiscal constraints over the medium term and difficult challenges in ensuring fiscal sustainability over the longer term. As members of the EU, countries are subject to the Stability and Growth Pact (SGP), and the commitment to prepare for Euro adoption entails the need to prepare convergence programs consistent with the Maastricht criteria and – where relevant – submission to the Excessive Deficit Procedure. These Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 13

Figure O4. Tax wedge and Employment Rate in the EU ( 2004). 50

BE SE 45 DE IT PL HU LV CZ EU8 40 LT SI FI SK AT NL () EE o b% 35 GR EU15

ES n FR o ar

g 30 PT dl ee

a LU UK

Txw 25

20

IE 15 50 55 60 65 70 75 Employment rate (%)

Notes: LT and SI (2003); EE (2002). Source: Eurostat. institutional constraints boil down to commitments to gradually reduce fiscal deficits to below 3 percent of GDP – and in principle to continue fiscal consolidation and achieve broadly balances fiscal positions – and maintain public debt below 60 percent of GDP. In the revised SGP, new member states that have introduced private pension pillars are given some respite in that the fiscal cost of such reforms can be phased in gradually over the period 2005-10. While all EU8 countries are below the debt threshold (although Hungary is teetering on the edge of it), only the Baltic countries and Slovenia meet the deficit target. The Visegrad countries (the Czech Republic, Hungary, Poland, and Slovakia) plan to achieve this only during 2006-08, and uncertainty remains about the broad political commitment and ability to adhere to these timetables in some of these countries. Over the longer term, all countries face significant fiscal pressures from population aging that in most cases could render public finances unsustainable in the absence of further reform of social insurance systems.7 vi. Most countries in the region have significant potential for creating the required fiscal space. First, several countries – notably the ones that retain large public sectors – have room for reducing wasteful spending on direct or indirect subsidies, rationalizing social transfer programs, enhancing the efficiency of health care and education systems, and streamlining public administration. Second, most countries can further strengthen tax administration, broaden tax bases and sources, and reduce tax expenditures. Third, some countries can advance further on the privatization agenda, raising resources not only for financing pension and other reforms that reduce implicit government liabilities, but also reducing the need for ongoing fiscal support in various forms and paving the way for improved efficiency and profitability of the given enterprises and thus enhancing the future revenue potential. There is also scope for increased use of public- private partnerships where there are good reasons to believe that the private sector is more efficient than the public sector in creating and operating infrastructure and public services. On the other hand, shifting spending off the budget will not by itself create real fiscal space. Fourth, in several countries there is scope for strengthening public finance management (procurement, debt management, etc.). vii. The Lisbon strategy calls for the emphasis of public finances to be broadened from its focus on stability to include the contribution to sustainable growth, full employment, social cohesion, and competitiveness. The quality of fiscal policy remains an important concern in most EU8 countries, as indeed in many incumbent EU member countries. While the impact of fiscal policy on short-term growth has been extensively studied and discussed, surprisingly little remains understood about the role of public finances in promoting longer term growth.8 Modern endogenous growth theory clearly identifies a potentially important role through potential effects on factor accumulation (both human and physical capital), the provision of public goods, addressing externalities, etc. (see e.g. Barro, 1990). Most models distinguish between productive and non-productive

7 Countries with strong fiscal positions and low public debt levels as well as those that have implemented comprehensive pension reforms (including establishing a close link between contributions and benefits in their PAYG systems) are generally in a stronger position to deal with adverse demographics. 8 Recent literature has focused on conditions for expansionary fiscal contractions (see e.g. Giavazzi and Pagano, 1990; and Alesina and Perotti, 1995). 14 spending and distortionary and non-distortionary taxes, and several of these postulate a non-linear, hump- shaped relationship between aggregate spending or taxation and economic growth. Nevertheless, such classifications are ambiguous, and the empirical evidence is similarly non-conclusive (Volume II Annex 2). This being said, it is hardly controversial that public investment in key infrastructure and education – at least up to a certain point – are conducive or even critical to sustained growth, while high direct taxes that affect investment and employment have a negative impact. Flat tax systems have become popular in transition economies, but the empirical evidence on their role in promoting growth remains mixed and tentative, revenue effects are uncertain, and they may reduce the role of public finances in income redistribution. viii. The role of fiscal decentralization in promoting efficiency and growth is, on the other hand, controversial. While subnational fiscal autonomy may promote the effectiveness and efficiency of government interventions through better knowledge of local market conditions, there is also a risk of regulatory capture by local interests and expenditure indiscipline. Institutions undoubtedly matter a great deal – if these are not strong, privatization of key municipal enterprises and service providers may often be a better alternative. Some studies do find a positive association between decentralization and local government revenue autonomy on the one hand, and output growth on the other, but again the empirical evidence is far from conclusive.9 This is hardly surprising as for example the decentralization of education in Hungary has had questionable results in terms of education outcomes. However, it is interesting to note that fiscal decentralization in the EU8 appears to have been weakly associated with faster output growth, although there is obviously significant variation and growth clearly depends on many other factors (Figure O5).

Figure O5. Fiscal Decentralization and Real GDP Growth in the EU8 1995-2003.

3 10 0 2) 50

eal LV

9 8 o r

1- EE 9 ( ate f

tr 6 n% LT CU i 4 ggrowh ainL SK SI HU PL CZ

vera e 2

er capit EU15 nala DP p An u 0 G 15 2 0 2 5 3 0 3 5 Expenditure of local government level as a s hare of general government expenditure in 2003 Note: The share of local government in total spending was broadly unchanged during 1995-2003. Source: Eurostat; and staff calculations. ix. The World Bank’s analytical work on fiscal reform issues in the EU8 is of a multi-year, programmatic nature, with a focus on the following five main areas: (i) education; (ii) health; (iii) social insurance; (iv) public administration reform; and (v) fiscal risks. The work adopts a regional, cross-country approach reflecting the fact that many of the issues and reform considerations are similar across most countries in the region and that directions for reform may have broad applicability. At the same time, the progress in different areas of reform has been uneven across the region, and there may be valuable lessons to be learned from the reforms. The studies in this volume focus on two priority areas for public finance reform: health care financing and higher education financing. In both areas, there has been some, albeit uneven, progress on reform in recent years, but much remains to be done in most countries to promote service quality, efficiency, equitable access, financial sustainability, and governance. EU membership and enhanced mobility of people across borders in the union is already affecting both sectors, including through upward pressure on salaries, and more is likely to come. There is little doubt that enhanced competition, increased reliance on user payments, and more targeted subsidies is in the cards. In both health care and education, the system of intergovernmental fiscal relations plays an important role, and the volume sets the stage for the subsequent discussion by reviewing the current situation and key issues in this area. Along a somewhat different dimension, the study also includes a discussion of the fiscal risks of public-private partnerships (PPPs) and the conditions under

9 See e.g. Davoodi and Zou (1998). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 15 which such arrangements may help to create real fiscal space.10 The regional focus of and priorities for the Bank’s analytical support have been widely endorsed by Finance Ministers in the region. Fiscal challenges for the EU8 countries (Chapter I) x. Fiscal trends in recent years have varied considerably among the EU8 countries. Two groups of countries emerge within the EU8 on the fiscal scene: those with relatively strong fiscal positions, modest debt, and small governments (the Baltic countries, and to some extent Slovakia); and those with relatively weak fiscal positions, sizeable debt, and large governments (Hungary, Poland, and the Czech Republic). Slovenia is a special case, with strong public finances but a large government. The composition of revenues and expenditures, meanwhile, remains an important issue in all EU8 countries, although the particular problems differ from country to country. Chapter I by Emilia Skrok and Paulina Bucon examines recent fiscal developments, assesses medium- long term fiscal prospects, and discusses the quality of fiscal policy in the EU8. xi. Following the temporary widening of fiscal deficits in the wake of the Asian and Russian crises in most countries, some countries pursued ambitious and sustained fiscal consolidation strategies while others allowed deficits to remain high or even widen further. In particular, the Baltic countries, Slovenia, and Slovakia all undertook a sustained adjustment effort, with fiscal deficits now around or below three percent of GDP (and even surplus in Estonia). Debt levels are low in the Baltic countries, moderate in Slovenia, and reaching comfortable levels in Slovakia.11 Fiscal policy has been more erratic in the other Visegrad countries.12 The Czech Republic managed to reverse a sharp widening of the fiscal deficit in 2001-02, but new pressures are building fast. Poland pursued an expansionary fiscal policy in the period 2001-04, especially in the last couple of years where output growth recovered strongly, but took steps to regain control in 2005. Hungary has to a large extent lost control over its fiscal policy, and debt levels are hovering around the critical 60 percent of GDP limit. xii. Fiscal consolidation efforts have been supported by strong output growth and in some countries expenditure reform and/or discipline while tax reforms have tended to lower revenues. Several countries have been pursuing tax reforms aimed at lowering the overall tax burden, and in most EU8 countries general government revenues as a share of GDP are now significantly lower than in the mid-1990s and than in the EU15. Slovakia has been a frontrunner in these efforts (starting from a relatively high level), reducing the tax burden by more than 10 percentage points of GDP. Cuts in corporate and personal income taxes have tended to lower the share of direct to total taxes, while the reliance on social security contributions and indirect taxes has increased.13 Overall, labor taxes remain relatively high in most EU8 countries. xiii. Expenditure developments have generally mirrored revenue developments, albeit with cyclical conditions and political cycles affecting short-term movements in some countries. Slovakia has been the only EU8 country to undertake a comprehensive restructuring of its social spending programs, with more piecemeal reforms in other countries that have tended to rely more on various administrative measures (notably Hungary). EU8 countries have generally also benefited from a decline in interest rates. On the whole, the structure of spending has not changed much over the past decade or so, although some countries have managed to reduce subsidies and public consumption to the benefit of higher investment. Increasing subsidies in the Czech Republic (and recently also Estonia and Lithuania), higher social spending and public consumption from already high levels in Hungary, and cut backs in social spending from low levels in Latvia and Lithuania are a source of some concern. Public investment is particularly low in Latvia. xiv. Intergovernmental fiscal relations have been influenced by an ongoing process of decentralization, with a tendency toward increasing budgetary autonomy of lower levels of government. The size of local governments relative to total general government in the EU8 is around 20-25 percent, broadly similar to EU15. Local governments in the EU8 have generally not been a source of fiscal deficits in recent years reflecting tight borrowing constraints in most countries, although they may have been behind some of the “hidden” fiscal deficits (which have amounted to an average of more than one percent of GDP annually in recent years).

10 While PPPs can play a role in the provision of healthcare and education services, this is not the focus in this report. 11 Debt developments have generally been influenced favorably by strong output growth, lower interest rates, and currency appreciation. 12 Fiscal outcomes in the Visegrad countries have generally fallen well short of targets agreed with the EU in the context of pre- accession economic programs and in some countries post-accession convergence programs. 13 More recently, changes in indirect taxes have been influenced by EU accession. 16 xv. Medium-term fiscal plans are laid out in countries’ Euro adoption convergence programs with the EU. The Baltic countries and Slovenia aim to maintain fiscal deficits below three percent of GDP and adopt the Euro from 2007-08.14 Slovakia is targeting to achieve the fiscal deficit threshold in 2006, and plans to adopt the Euro from 2009. Poland and Hungary were planning to achieve the fiscal target in 2007, but a new government taking office in Poland and recurring fiscal slippages in Hungary raise uncertainty about these commitments and Euro adoption may well be pushed back to beyond 2010 in both countries. Despite major outstanding reforms, the Czech Republic looks to be on target to reduce its fiscal deficit to below the critical level in 2008, with Euro adoption feasible in 2010-11. Existing fiscal plans would be consistent with maintaining public debt ratios below the required 60 percent of GDP. xvi. Looking further ahead, public finances – notably pension and health spending – in all EU8 countries will come under increasing pressure from population aging. While some countries (notably Poland and Estonia) have positioned themselves well through comprehensive pension reforms, others face rapidly rising and unsustainable debt levels in the absence of further reforms – not least in case of failure to comply with medium-term fiscal programs. xvii. The quality of fiscal policy remains an important concern in most EU8 countries. The Lisbon strategy calls for the emphasis of public finances to be broadened from its focus on stability to include the contribution to sustainable growth, full employment, social cohesion, and competitiveness. In this context, the quality of fiscal policy refers to the structure and efficiency of taxes and spending. The quality of fiscal adjustments also has important bearings on the success and sustainability of such adjustments. High quality fiscal adjustments – including expansionary episodes of fiscal adjustment – have been found to rely on expenditure cuts rather than tax increases, and on tackling those expenditures that are politically sensitive such as transfers, subsidies, and wages. The EU8 countries have undertaken several episodes of fiscal adjustment since the late 1990s, with about one-half of these episodes dominated by expenditure contraction.15 Fiscal contractions have relied on cuts in investment (Czech Republic, 1999 and 2004; Estonia, 2003; and Latvia, 2003), social spending (Lithuania, 2002; Latvia, 2001 and 2002; and Slovakia, 2003), and public consumption. xviii. The extent to which fiscal adjustments and public finances in the EU8 have been and are conducive to growth is difficult to assess. The literature remains somewhat inconclusive as far as the impact of size of government and structure of public finances on growth is concerned, although some consensus seems to be emerging that there may be a hump-shape on the former relationship – too little and too much government is bad, that certain spending areas tend to be more productive than others (notably investment in critical infrastructure and human capital is better than subsidies and excessive government employment), and that some taxes are more distortionary than others (indirect taxes less distortionary than direct taxes). Much has to do with the efficiency with which government collect and spend resources, both in a technical and economic sense. The tentative findings in Chapter I suggest that in the EU8, higher tax burdens do tend to be associated with weaker growth performance, and that the higher the share of direct taxes and social security contributions in total taxes, the more harmful is the tax system. Findings are less conclusive regarding expenditure composition, although investment and social benefits tended to be positively and negatively, respectively, related to growth. On the whole, EU8 public finance reforms in recent years appear to have been in the right direction, although much remains to be done. Intergovernmental fiscal relations (Chapter II) xix. Overall, the EU8 countries have made remarkable progress in reforming their systems of intergovernmental fiscal relations, and the eight new EU countries now have much more in common with the longer-standing members of the Union than they do with the Soviet-era systems they inherited. Subnational governments are an important part of the public sector in the EU8 countries, accounting for about one-quarter of total government spending. They provide basic public services in both the social sectors (education, health, and social assistance) and in infrastructure (water supply, sewerage and transport). How well they perform these functions has an important influence on the quality of both human and physical capital in these countries, and therefore on the prospects for economic growth. It also has equity implications, as the social services performed by municipal governments have important distributional effects. Further, subnational government operations may affect macroeconomic stability

14 Estonia, Lithuania, and Slovenia joined the ERM2 in July 2004, with Latvia following suit in April 2005. The first three countries target euro adoption for January 1, 2007, and Latvia one year later. Recently, these plans in Latvia and Estonia have been put into question due to increases in inflation. 15 However, since 2001, only Slovakia has implemented expenditure dominated fiscal adjustments. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 17

Box 01. Objectives of Intergovernmental Finance Systems What are the objectives of a system of intergovernmental finance? According to conventional economic theory, the primary aim should be to promote efficiency in the allocation of resources while the central government would be responsible for stabilization and income distribution. Theory argues that if the benefits of particular services are largely confined to local jurisdictions, welfare gains can be achieved by permitting the level and mix of such services to vary according to local preferences. Local consumers, confronted with the costs of alternative levels of service can express their preferences by voting for rival political candidates (i.e., voting with their hands) or moving to other jurisdictions (voting with their feet). In this respect, local politics can approximate the efficiencies of a market in the allocation of local public services by “pricing” municipal services and relying on the local political process and household mobility to clear the market. On this basis, a successful system of intergovernmental relations would be one in which each level of government (1) performs only functions whose benefits are confined to its boundaries and (2) derives its revenues from sources (such as local taxes and fees) that confront its residents with the costs of these functions. In practice, there are a number of limitations to this model. First, the benefits of public services can be difficult to define in geographical terms. While most of the benefits of spending on public health, for example, may be confined to a local jurisdiction, failure to control infectious diseases can have national or international implications. Second, spending on public services – particularly education and health – can be important instruments of distributional policy. Where these functions are assigned to local governments, their distributional implications cannot be ignored. Third, high levels of local taxation or widespread deficit spending could weaken the central government’s control over fiscal policy. In practice, the intergovernmental fiscal systems of the EU8 countries must be held to somewhat different standard: (i) Transparency and predictability: The principal challenge confronting the EU8 countries was to devise a system for financing local government that would be perceived as fair and stable. The emergence of multi-party democracy in the EU8 required transparency in order to allay charges of partisan favoritism and reduce the transaction costs that would result from continued reliance on bargaining. Predictability was required to provide local governments with a stable basis for budget planning; (ii) Equity: Because local governments assumed responsibility for social services, the new system had to ensure that such services would be available to people living in poor jurisdictions; (iii) Macroeconomic control: The system had to ensure that local government fiscal behavior would not threaten macroeconomic stability. Central governments needed sufficient control over aggregate local taxing to conduct fiscal policy and sufficient control over local borrowing to ensure that aggregate deficit targets were achievable; and (iv) Effectiveness: Finally, the systems needed to encourage effectiveness in the production of services under local control. The EU8 countries inherited over-scaled schools and hospitals and inefficiently run public utilities. Incentives for rationing services therefore had to be incorporated in the new system of intergovernmental finance. as independent fiscal actions by local governments could undermine central control over fiscal policy. Chapter II by William Dillinger takes stock of developments in intergovernmental fiscal relations and discusses current challenges in this area in the EU8. xx. Organizational reforms generally took the form of scaling back the responsibilities of deconcentrated units of central administration while creating autonomous units of self-government and, in the larger EU8 countries, new, consolidated regional governments. Privatization had two important implications for local governments. First, it eliminated important providers of public services. Second, it drastically altered the revenue base. Profits (where such were generated) from local-government-owned enterprises were no longer a source of revenue, and as central governments revised the tax structure to capture revenues from the growing private sector, the fundamental basis for tax sharing had to be altered. Functional decentralization required an increase in local government revenues. Privatization changed the nature of the tax base, and changed political conditions increased pressure for local fiscal autonomy while raising the risk that local governments would use that autonomy to thwart central government fiscal and equity objectives. xxi. The new local government legislations assigned an extensive list of functions to the newly independent municipalities. Education is the largest single item of local government expenditure, with local governments (including intermediate levels of elected government) accounting for about two-thirds of total public sector spending on education in most EU8 countries. The extent of municipal management control over education, however, is limited, owing to legal constraints which set out the framework for teachers’ salaries, teaching loads, and rules for recruitment and promotion. Local governments play a major role in managing healthcare in the EU8, but only a minor role in its financing. All these countries now finance the majority of healthcare costs through compulsory insurance schemes. In most of the countries, health insurance funds (HIFs) provide 18 funding directly to healthcare providers (primary physicians and secondary and tertiary healthcare facilities), and subnational governments thus account for only a small share of total public sector health spending in most countries. In all EU8 countries, central governments provide the vast majority of social protection, including old age and disability pensions, sickness benefits, and short term unemployment benefits.16 Local governments, nevertheless, are responsible for a variety of assistance to other vulnerable segments of the population, including the long-term unemployed and their dependents. They have varying degrees of discretion in determining who is eligible for such benefits and what form they should take. xxii. With the privatization of the housing stock, municipal responsibilities for housing management and maintenance have declined sharply in the EU8 countries. Municipalities have, on the other hand, assumed responsibility for water supply, sewage, district heating, solid waste management, and third tier roads, although these do not necessarily account for a large proportion of budgetary expenditure.17 Water, sewage, district heating and solid waste management are generally organized as separate, tariff-supported, enterprises. Poor standards of original design and construction, high energy requirements, changing patterns of demand, and long periods of deferred maintenance have resulted in deteriorating and obsolete infrastructure networks. xxiii. The revenues of local governments consist of discretionary funds, largely revenue sharing and funding for specific functions, principally education and health.18 Revenue sharing generally relates to the personal income tax (PIT) – in all EU8 countries, the shared taxes are administered by the central government at nationally uniform rates, and fixed shares of the revenues are then assigned to local governments. The distribution mechanisms for the PIT include a high degree of equalization. None of the EU8 countries assign a major unrestricted – in the sense that municipalities have the authority to set the rate and retain the revenue – tax base to local governments. The most widespread form of local taxation is the property tax, but local discretion over tax rates varies among countries and even in those that have moved to marked-based valuation, the property tax generates very little revenue. xxiv. All the EU8 countries have supplementary arrangements for financing education and health. The post-Soviet-era systems of education financing reflect an ongoing tension between ensuring a basic level of education financing in all jurisdictions, regardless of the strength of local tax bases, and encouraging efficiency in the use of education funding. All the EU8 countries finance the largest component of education – teachers’ salaries – through some form of intergovernmental transfer. A few countries rely solely on central government administrative procedures to ration the level of local spending on education, while most distribute funding for education on a per pupil (capitation) basis and allow local governments considerable discretion over how these funds are used. The latter approach ensures a minimum level of education financing in all jurisdictions while allowing the central government to ration the level of such spending through its control over the per-capita amount. In particular, it forces rationalization in jurisdictions with under-enrolled schools. However, the success of capitation based financing is challenged by the inability of local governments to dismiss staff as well as the unwillingness or inability of central governments to increase the level of capitation transfers to reflect centrally-mandated increases in costs, notably wages. xxv. The EU8 countries have also attempted to use the health financing system as a tool to achieve efficiency. All have created HIFs to finance the majority of primary, secondary, and tertiary healthcare costs. These funds are financed from payroll taxes, supplemented by central government payments on behalf of certain categories of non-contributors (pensioners, workers on maternity or sick leave, and unemployed). Health funds were generally created to protect healthcare spending from the vicissitudes of the annual government budget process (although funding remains subject to fluctuations in the payroll tax base and the rate of the payroll tax). In principle, they also function as a ceiling on healthcare spending: with few exceptions, spending on primary, secondary, and tertiary healthcare is not allowed to exceed the revenues of the HIF. However, in practice hard budget constraints are often breached through arrears and extraordinary budget support. Attempts to use fiscal instruments to ration the supply of health services on a more micro level have had mixed results. One

16 Pension systems have been partially privatized in several countries in the region. 17 They do, nevertheless, constitute an important source of contingent liabilities (as these enterprises often have high levels of arrears) and a claim on scarce investment resources. 18 In most of the EU8 countries, the specific purpose of this funding is to cover the full costs of the function to which it is assigned. Poland and Hungary are exceptions, with the level of funding for education determined independently of the costs of this function in the expectation that any shortfall will be funded from discretionary revenues. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 19 of the major sources of inefficiency in the Soviet model of healthcare was its excessive use of secondary and tertiary facilities. To varying degrees the EU8 countries have addressed this issue by introducing “family medici ne” models of primary care, but performance incentives have proven incompatible with cost containment due to the tendency of providers to respond to fee-for-service forms of compensation by increasing the volume of services they provide. Rationing secondary and tertiary care has been even more difficult. xxvi. EU8 countries have also strengthened controls on local government borrowing. Following liberal access of local governments to capital markets in the early 1990s and several instances of default, the majority of EU8 countries have introduced quantitative ceilings on municipal borrowing and debt service. Several of the EU8 countries also impose case by case administrative controls on local borrowing and restrict certain forms of borrowing. Further, several countries impose penalties on municipalities that fail to meet their debt obligations, and a few have introduced formal municipal bankruptcy laws (although the use of these measures has been limited so far). Central government bailouts remain a possibility in most countries. xxvii. Chapter II argues that the EU8 countries have made substantial progress in the reform of intergovernmental relations as measured against the criteria outlined above. Most countries have come a long way on transparency and predictability, although Hungary remains a source of concern. Countries in the region also manage to achieve a high degree of redistribution in the financing of social services, perhaps even too much: while there is a strong case for equalizing spending on social assistance, the case for equalizing spending on other functions financed from discretionary revenues – road maintenance, community amenities, or general administration – is less obvious. Local government finances have not been a source of fiscal and macroeconomic instability, reflecting limited fiscal autonomy and strict borrowing constraints in most countries. On the other hand, attempts to use the system of intergovernmental relations to encourage greater efficiency in the production of public services have been less successful. While the majority of the EU8 countries now finance education on a capitation basis, this has not been sufficient to prompt the closure of under-enrolled classrooms or schools. In the health sector, efforts to encourage primary providers to increase the volume of services they provide have been thwarted by over-billing, and efforts to ration secondary and tertiary care using variants of the German points system have run aground for similar reasons. xxviii. Over the longer term, Dillinger points out that sustainability of EU8 intergovernmental fiscal arrangements will depend on the ability of central governments to keep their agreements with local governments. In monopolizing the major sources of tax revenues, central governments have arrogated to themselves the power to decide how large a share of total public sector revenues local governments should receive, and how they should be spent. In the process, they have made themselves vulnerable to the claim that the level of funding which they provide to local governments is insufficient. In the case of sector specific financing, the implication is that central government mandated increases in costs of service provision must be accompanied by increases in the level of financing. In the case of funding discretionary expenditures, the main issue is risk related to the annual negotiations between the central and local governments. Experience from Western Europe suggests that countries might deal with these issues through either increasingly sophisticated estimates of the costs of providing discretionary – in effect treating these as if they were mandatory (England is a good example of this approach), or allow the local governments to decide for themselves on the appropriate level of spending on the basis of their constituents’ willingness to pay. The latter approach would require the assignment of a broad based tax such as the PIT to the local level. Some Scandinavian countries have gone this way (and is also cited in the European Charter of Local Self-Government and actively promoted by the Council of Europe), but it appears to have few supporters among the EU8. At present, central governments prefer to retain near-total control over tax policy, and local governments prefer to avoid the onus of taxation. Financing higher education (Chapter III) xxix. The EU8 countries have enthusiastically embraced mass higher education, and are rapidly closing the gap between themselves and the EU15 in enrollment rates and the incidence of higher education in the population of working age. This expansion is helping these countries to insert themselves into the global knowledge economy, as a high proportion of their graduates get jobs in the knowledge-intensive services sector. At the same time, resources per student are declining, raising concerns about the quality of higher education. Also, the emergence of private higher education institutions and introduction of fees for student who do not gain regular admission, have resulted in a dual-track system that is affecting the equity of access to higher education. Chapter III by Mary Canning, Martin Godfrey, and Dorota Holzer-Zelazewska reviews the experience with a variety of financing mechanisms in the EU8 countries and seeks to develop some 20 useful policy options to “level the playing field” for countries contemplating further reforms, which would include the introduction of variable fees, needs-based grants and loans to increase private financial flows into higher education while facilitating equal access across the board. xxx. Average enrollment in higher education in the EU8 countries is now equivalent to well over one-half of the relevant age group, and the annual number of tertiary graduates (per 1000 people) now exceeds that of the EU15 (although, on average, the proportion of the 25-64 population with higher education remains larger in the latter group of countries). With the coming of a market economy and substantially greater opportunities for university graduates, both in local private sectors and now potentially elsewhere in the EU, the private benefits from university education have increased considerably both in terms of higher income and lower rates of unemployment. The earnings premium for higher education in the labor market is particularly high in countries where higher education still has some scarcity value. Unemployment rates (similar on average in both the EU8 and EU15) vary between countries but are invariably lower for higher education graduates than for those with lower levels of education. Overall, the private rate of return to higher education is still high in the EU8 and, given the pattern of subsidization, likely to be higher than the social rate of return. xxxi. Most EU8 countries have managed to protect the percentage of GDP that is devoted to public expenditure on education, and the spending on higher education is comparable to the EU15 (around one percent of GDP). However, EU8 countries spend a much higher proportion of GDP per head per student in public tertiary institutions than EU15 countries, and a much higher multiple of expenditure per primary school student. The number of Euros spent per tertiary student, adjusted for differences in purchasing power, is less than one-half of the level in the EU15. Expenditure per head seems to vary inversely with enrollment rates, which suggests at least the possibility of a trade-off between the accessibility and quality of higher education from the student’s point of view. While such cross-country comparisons between countries with different salary levels, even when adjusted for purchasing-power differences, are fraught with problems, there is little doubt that an increase in unit expenditure will be needed.19 xxxii. The small role of fees in most EU8 countries does not mean that parents and students avoid paying for higher education. Many countries have instituted a dual system, maintaining fee-free higher education for regularly admitted state-supported students, while adding a special fee-paying track for those who fail to gain such admission. Institutions have also imposed and increased user charges for formerly heavily- subsidized board and residence facilities, which are borne by students or their parents, as are the costs of living for those who do not make use of such facilities. Also, although a significant proportion of educational budgets remains devoted to stipends, maintenance grants or other types of financial aid to students, they have been eroded by inflation to virtually nominal levels in many EU8 countries.The dual-track system means that there are two classes of students – those whose education expenses are nominal, and those who pay fees equivalent to several thousands of dollars. Those who obtain fee-free, state-subsidized places are disproportionately from privileged families (which have contributed to their academic success); poorer students, who are less successful in entrance examinations and cannot afford the alternative fee-paying track, are generally excluded from higher education. In order to facilitate entry to higher education for students who need financial support to pay fees and/or maintenance costs, several EU8 countries have instituted loan systems. xxxiii. The trend towards increased participation in higher education can probably not be combined with the improvement in quality and relevance that is needed in order to be competitive within the EU and to increase participation in the global knowledge economy, unless the amount of money available to institutions increases. At the same time, government budgets are severely constrained, and it is difficult to justify an increase in public allocations to higher education which yields high private rates of return to people who are disproportionately from privileged backgrounds. Also, the increasing internationalization of higher education and cross-border labor mobility from the new toward the old EU member countries reduces assurances that university graduates will eventually pay for their education through income taxes. Thus, it is almost inevitable that average fees in tertiary institutions would have to be increased to a level which covers a significant proportion of total costs. The level at which tuition fees are set is clearly not only

19 Tertiary institutions in the EU8 spend a lower percentage of their budgets on research and development than do their counterparts in the EU15, but even if this is taken into account, the contrast in unit educational spending remains. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 21 an economic but also a political issue. There is a strong case, on the grounds of revenue maximization, efficiency, autonomy, and equity, for making such fees variable, rather than fixed and uniform. xxxiv. The extension of fees to all students and an increase in their average level would imply a need for wide-ranging student loan systems. The authors argue in favor of income-contingent loans carrying “unsubsidized” interest rates.20 Repayments would be income contingent, calculated as a given percentage of the borrower’s earnings until the loan has been repaid. Those who take out loans effectively get their higher education free at the point of use, repaying only later if and when they can afford to do so. Thus, even the most risk-averse students (often assumed to come disproportionately from low- income backgrounds), may be willing to borrow in order to finance their studies. Several countries have introduced student loan schemes, but with varying success. The authors note that there are several pre- conditions for effective student loan schemes, including ways to keep track of people’s movements and systems of withholding at the point of wage and salary payment. Further, to the extent private banks are involved, there would need to be a system of clear and credible government guarantees and proper fiscal accounting for these. Student loans could be supplemented with scholarships and grants for those who truly need it.21 However, income contingent loans (and other subsidies) raise difficult issues of both feasibility and equity. Means-testing is complicated by the fact affluence may depend not only on current income, and further, there is no assurance that wealthy parents would pay for their children’s education. Means-tested loans and grants to higher education students are, at best, a blunt instrument and need to be backed up by other measures to promote equity in access. xxxv. Such reforms would allow the inequitable dual-track system to give way to a single-track system, under which all students pay tuition fees, and to a distribution of state subsidies based primarily on need, rather than on academic ability. Thus, a system of deferred tuition fees (i.e. combined with a well-designed system of student loans), supported by a system of scholarships or grants targeted to those who need them, is an essential part of any strategy for expanding quantity, improving quality and achieving equitable access. Students should be eligible for government subsidies whether they are attending a public or a private institution. At the same time, it is the role of the government to set and monitor quality standards for all tertiary institutions, regardless of ownership. The combination of variable fees, needs-based grants, and loans would also help to increase the current and future labor market relevance of the specializations chosen by students. xxxvi. The chapter also stresses that further reforms in the criteria for determining the amount of public money to be allocated to each tertiary institution would help increase transparency of funding and efficiency. Having started with fairly simple “money-follows-the-student” formulae, including competitive and experimental funding mechanisms to encourage innovation and research, would be useful in meeting the required policy goals. Further governance reforms, particularly in institutional management and procedures for appointing leaders of autonomous tertiary institutions, are likely to be needed to ensure that lump sums allocated through such mechanisms are used in a way that is sensitive to the public interest. While deregulation and increased autonomy of universities have made higher education more sensitive to student choice and labor market developments, closer links between the university and the private sector are desirable, also to attract private funds through endowments and other mechanisms. xxxvii. Decreased reliance on government funding and increased reliance on price incentives would not mean the elimination of a role for governments in relation to higher education. They would still be an important source of funds, organize and oversee student loan schemes, and be responsible for the promotion of equitable access. They would also have to ensure that quality assurance systems are in place, and would be able to design and use formula-funding schemes to achieve national objectives that go beyond those of the immediate market and to modify excessive competition between institutions. What is needed, rather than detailed interference in academic processes, is a combination of standard-setting and financing systems designed to ensure high-quality outcomes which would be communicated to stakeholders through accurate, impartial and easily available data. The rest could be left to higher education institutions, autonomous but accountable in their governance arrangements.

20 Interest rate subsidies add a lot to the costs of a loan scheme, divert funding from quality improvement to student support, and are deeply regressive (mainly benefiting successful professionals in mid-career whose loan repayments are switched off early because of the subsidy). 21 Institutions and governments could offer a few competitive scholarships to the strongest candidates to provide an incentive for competition in the qualifying process. 22 Controlling health care expenditures (Chapter IV) xxxviii. Following the transition from central planning toward market-based economies, the EU8 countries introduced a number of reforms in the finance, management and organization of the health sector. Reforms removed the state monopoly on healthcare by privatization and decentralization, and started the process of establishing an economic relationship between the healthcare provider and the patient through changes in healthcare financing. Almost all countries adopted the Bismarckian social health insurance model, and in all cases one or more Social Health Insurance Agencies started operating independently of the state budget. In most countries, subsequent reforms also addressed the pharmaceutical sector, and brought about changes in the health financing structure in an effort to address the manifold problems brought about by the insurance system. However, reforms have generally been less successful in securing sustainability of health care financing, improving efficiency, enhancing equity in healthcare financing and delivery, and managing the quality of health services. Chapter IV by Mukesh Chawla and Marzena Kulis takes stock of recent trends in health expenditure and discusses directions for reform consistent with the objectives of stabilizing the fiscal situation in these countries without adversely affecting the production, delivery and utilization of health services. xxxix. EU8 countries have generally been successful in safeguarding resources for the health sector and a reasonable health status of their people. Total spending on health in the EU8 countries is at levels roughly commensurate with per capita GDP in these countries, and is not out of line with other European countries and high-performing middle income countries. In some EU8 countries like Estonia and Latvia, however, health spending is on the low side relative to per capita GDP. The health status of the people in the EU8 countries is at levels commensurate with their levels of health spending and income, but is generally poor compared to the EU15. The difference between the EU15 and EU8 average life expectancy has decreased in recent years, but most EU8 countries have some five years shorter life expectancy than the EU15 average. xl. Health expenditures in the EU8 countries are mainly financed from public sources. On average, public financing accounts for about three-fourths of total health spending in EU8 countries, which is comparable to the EU15. Out-of-pocket payments in the health sector account for the bulk of the residual financing and have been increasing in recent years. Social health insurance is the dominant form of public financing of the health system in EU8 countries, although many also transfer funds from the state budget to the social insurance system. Most countries have a single agent who purchases health services on behalf of the insured. The public financing system in all EU8 countries provides generous coverage of health services, in most cases with little or no financial participation from the patient at the point of service. xli. Health expenditures in almost all EU8 countries are dominated by inpatient care and pharmaceuticals, which account for roughly 40 percent and 30 percent of total health expenditures, respectively. Primary and specialist outpatient care, long-term care, and administrative expenses comprise the balance. The general trend over time has been that of rising expenditures on pharmaceuticals, falling expenditures on inpatient care and increasing expenditures on outpatient care. This structure of spending is not very different from the OECD countries. The relatively fast growth in pharmaceutical costs in both groups of countries reflects higher volumes and the entry of new, more expensive drugs, rather than general price increases. xlii. In addition, excessive hospital infrastructure is a major drain on scarce health sector resources, and salaries and the cost of new technology are also rising fast. The EU8 countries inherited a disproportionately excessive number of hospitals and hospital beds, and while almost all countries have taken many bold steps to reduce the number of hospitals and acute-care hospital beds, but the number of beds per 100,000 inhabitants is still 30 percent higher than the average for EU15 countries. Salaries in the health sector – which account for more than 60% of health expenditures in these countries – are rising faster than average salaries. This reflects greater availability of outside opportunities following EU accession, and also the push to bring the ratio of salaries in the health sector to average salaries in the economy in line with EU15 proportions.22 Further, widespread and extensive use of new and expensive medical technology has made it a major cost- driver in the health systems in many EU15 countries, and this trend is likely to continue.23

22 The ratio of physician salaries to the national average is significantly lower than two (the average EU15 level) in many EU8 countries, but are likely to rise toward the EU15 average as medical professionals from the new member states are in great demand there. 23 The tax-funded health systems of Spain and the UK have the lowest density of MRI units and CT Scanners, and also report the lowest hospital capacities and inpatient expenditure shares, as well as the lowest total health expenditures. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 23 xliii. A defining feature of the health sector in almost all the EU8 countries is the widespread and growing indebtedness.24 Buoyed by generous benefit packages (offering ineffective and non-essential services), extensive infrastructures, and pressured in recent years by increasing pharmaceutical costs and higher salaries, expenditures on health in country after country have consistently surpassed the available resources, leaving huge unpaid bills for services already delivered. This situation is particularly severe in the Visegrad countries, although some of the Baltic countries are also beginning to feel the pressures. Regions with an excessive concentration of public hospitals and hospital beds expectedly also generate the largest debts in the health sector. Hospital contracts with health insurance companies do not cover the full costs of maintaining the large hospital infrastructure and associated fixed costs, and the resources allocated by the central and local governments (who own most public hospitals in most EU8 countries) are not sufficient to pay for all the expenses. In addition, public hospitals tend to provide services irrespective of the value of contracts concluded with the sickness funds, and bear costs that are refunded neither from health insurance nor from the state budget. xliv. Financing and delivering healthcare for the growing population of the elderly is likely going to be the largest future cost-driver in the health systems of the EU8. All countries in the region face the consequences of population ageing caused by reduced fertility and mortality rates on the one hand, and increasing life expectancies on the other. The share of people aged 60 years and older in the total population in EU8 countries, which was 16 percent in 1995, is projected to increase to 27 percent in 2025. Correspondingly, the old age dependency ratio is expected to rise sharply. Older persons are more likely to have greater health needs and are more likely to consume more expensive healthcare services, particularly during the last years of life. xlv. Thus, despite a decade of reforms, the underlying supply-side incentives in the health system continue to be weak and generally ineffective, almost matching the near-absence of demand-side incentives which promote cost-consciousness. While most countries have introduced risk-sharing at the primary care level by paying the providers on the basis of capitation, similar incentives are by and large absent at the secondary and tertiary levels. Most hospitals in the EU8 countries are owned by the local or central governments and most hospital managers have no claim to any residual balances that may result from good management of the facilities. Prices for health services are typically set at the level of the central government, usually at levels very close to the costs of production and delivery of services, and with few rewards and penalties for staying within/breaching the budget, health providers pay scant attention to the cost side of the equation. xlvi. The health insurance system in many countries continues to function like centralized collection and reimbursement agencies, and many of the tasks and responsibilities usually identified with an insurance system are lacking. On the delivery side, the shift from the more resource intensive inpatient care to less expensive outpatient care has not been accomplished, and the culture of over-hospitalization and seeking specialized care persists. Recent gains from the reduction in length of hospital stays per episode, and increases in bed occupancy rates in individual facilities have been negated by slow progress in addressing the over-supply of hospital infrastructure. On the financing side, not all healthcare systems have been able to find appropriate mechanisms of reimbursement so as to motivate providers to deliver better quality of services, and to produce them in an efficient and cost-effective manner. On the management side, even though most state-owned hospitals enjoy a fair degree of de jure autonomy, few efforts have been made to improve administrative capacity and management within hospitals. xlvii. The scope of services covered by the public system have generally not been clearly defined and articulated, and formal copayments remain limited in most countries. Further, the pervasiveness of informal payments in health in many EU8 countries has become a serious impediment to healthcare reform. Besides contributing to the general environment of corrupt practices and the growth of a parallel healthcare financing system, informal payments introduce perverse incentives in the health system, and compromise efforts to improve efficiency, accountability, and equity in the delivery of health services. The non-transparent and discretionary nature of informal payments adversely affect access to healthcare, particularly for the more vulnerable segments of the population who have to pay disproportionately large amounts for health services that are supposed to be available free of charge.

24 Debts refer to payments outstanding and past due, not liabilities which are a natural consequence of the financial management of health facilities. 24 xlviii. Much remains to be done to improve the quality of health care. Most EU8 countries do not have well-functioning quality controls that regularly incorporate evidence-based medicine in the production, delivery, and financing of healthcare services, and institutional mechanisms to review the quality of care at the health facility level are generally lacking. Clinical protocols for major diseases are outdated and non- compatible with evidence-based medicine and cost-effectiveness analysis. The existing quality improvement systems are fragmented and are not integrated across different elements of the healthcare system, including professional self-regulation, purchasing by health insurance funds, training of health professionals, and management of individual healthcare practices. xlix. The solutions to these three problems are obvious and include some combination of: (i) rationalization of the benefit package and introduction of supplementary health insurance; (ii) introduction of co-payments or other mechanisms to make people take greater responsibility of their health and healthcare-seeking behavior; (iii) creating competition among healthcare providers; (iv) consolidating and closing public hospitals and reducing hospital beds; (v) risk-pooling and rational purchasing of health services; and (vi) introducing provider payment mechanisms that promote incentives for cost management and quality enhancement. While rationalization of health spending may also pave the way for lowering health contributions and the tax burden on labor, the latter objective may also be achieved by moving toward greater reliance on broader based general taxes for financing health care. This is the direction in which several European countries have gone, and with good results – perhaps reflecting greater effectiveness of government interventions to control costs in the health sector in such systems. l. Creating a universally acceptable benefit package to be funded from public sources is an attractive, but difficult proposition. Health insurance funds (or governments) tend to exclude some treatments or entitlements, which do not solve the problem. For the fiscal sustainability of health systems, a flexible mechanism needs to be in place that allows for a benefit package that responds to population needs and can be serviced within the given budget constraints. This flexibility needs two types of decisions. First, setting priorities and matching them with the principle of covering catastrophic costs; and second, ensuring legislative support for the package, so that full political responsibility for such decisions is taken and adequate levels of resources are earmarked. li. To a great extent, demand for healthcare is controlled by the individual consumer, who has an incentive to over-consume if they are responsible for only a fraction of the costs. The key to containing excessive and unnecessary demand for health services therefore lies in making individuals and families sensitive to the costs of additional health spending, while still limiting each family’s maximum outlays to affordable levels. This was the logic behind the introduction of patient co-payments in the health sector in Slovenia and Slovakia, a move that in both countries resulted in significant savings. The adverse equity effects of co-payments can be neutralized to a large extent by exempting the poor and vulnerable from making official co-payments or through targeted grants, though it is not easy to set up a system by which the poor and the vulnerable are always appropriately identified. lii. Further reforms are also needed to rationalize hospital infrastructure and contain the rapid growth in pharmaceutical costs and new expensive technology. Estonia has been a frontrunner in terms of downsizing hospital infrastructure, and much can be learned from the reforms there. Lessons can also be learned from the ongoing restructuring of hospitals in Slovakia, focused on the largest cities where over-capacity is most severe and where access is less of a concern. Drug price regulation is important and necessary, but the experience of many countries shows that regulation of consumption of pharmaceuticals is also critical in order to contain expenditures. Most countries have adopted demand-side measures for controlling consumption, and cost- sharing has proven to be the most effective. Lowering drug costs requires a multi-faceted approach, necessarily involving the policy-makers working together with payers, doctors, pharmacists, and patients, and involving a package of measures aimed at pricing, prescription, and safety of drug use. Finally, EU8 countries may be well-advised to consider regulating the use of high-end technologies, similar to most EU15 countries.25 liii. Looking forward, the social health insurance system in the EU8 countries will come under even greater pressure unless urgent action is taken to address rising health expenditures. At the very least, the health reform package would have to consist of a combination of stricter supply-side measures, such as

25 Countries with tax-based systems such as the UK and Spain have national agencies for high technology and technology control measures in their national health policies. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 25 management of pharmaceutical expenditures and hospital restructuring, and demand-side measures, such as greater patient responsibility for their own health and greater patient contributions, including cost- sharing for pharmaceuticals. The objective of these measures in the short-run has to be the stabilization of the fiscal situation so that no new debts are created in the health system. Likewise, the emphasis in the medium term should be on improving efficiency and effectiveness, ensuring access to healthcare, and enhancing the quality of care. liv. Estonia, Slovenia and Latvia, all exposed to the same set of financial pressures as the other countries in the region, have effectively managed health care finances through good governance, strict adherence to the rules of financial discipline, and by simply not spending what they do not have. To be sure, all three countries also carried out a host of reforms to control pharmaceutical spending, reduce hospital infrastructure, improve hospital management and so on, but the underlying difference between these three countries and the others is that of sheer fiscal discipline. Also, important lessons may be learned from several other EU countries such as Austria that were in a similar state at the time of membership, but found ways of managing their budgets without bringing about huge destabilizing changes. Managing fiscal risks from PPPs (Chapter V) lv. Public–private partnerships (PPPs) have a long history around the world, and have become increasingly popular also in the EU8 countries. Poland, Hungary, and the Czech Republic, in particular, have all experimented with such arrangements, not least in the road sectors. Public-private partnerships may enhance efficiency in service delivery and help create fiscal space through real fiscal savings, but managing the associated fiscal risks and achieving these objectives requires a strong institutional framework. There is little doubt that in practice part of the attraction of PPPs is that they allow governments to secure much- needed investment in public services without immediately having to raise taxes or borrow – in the EU8 potentially getting around Maastricht criteria for debts and deficits. Chapter V by Nina Budina, Tim Irwin, and Hana Polackova-Brixi reviews the experience with the use of PPPs in the EU8, discusses the fiscal risks associated with such arrangements, and proposed institutional improvements to better manage such risks. lvi. PPPs involve fiscal risks. In some PPPs, governments commit to purchase the output of their private partners, whether or not they subsequently want the output. In these PPPs, governments take on debt-like obligations with a value roughly equal to the cost of the investment, even if they need not disclose any new liability. In other PPPs, final users pay, but governments guarantee certain outcomes. Governments might pay nothing now, but face the risk of paying later, perhaps when they can least afford it. The danger of PPPs is that governments’ desire to avoid reporting immediate liabilities may blind them to the future fiscal costs and risks. That not only increases fiscal vulnerability, but may also lead to poorly designed projects with a higher cost than necessary or the use of PPPs when either public finance or purely private finance would be better. lvii. The authors argue that the extent of fiscal risk depends intimately on the fiscal institutions that shape and constrain government decisions toward PPPs – that is, on factors such as fiscal targets (including the Maastricht criteria), budgeting procedures, accounting and auditing standards, and the assignment of responsibilities for fiscal decisions among different parts of government. Such fiscal institutions affect decision makers’ incentives. They may, for example, reward policy makers for minimizing short-term cash expenditures, or alternatively for reducing fiscal vulnerability. Institutions may also influence the information available to decision makers: they might generate information only on traditional liabilities created by publicly financed investment, for example, or also on liabilities incurred in PPPs. Finally, they affect decision makers’ capacity: for example, they may or may not allow governments to analyze the risks in PPP projects and decide which are best borne by the government. lviii. Better fiscal institutions can therefore increase the chance that PPPs will be well designed and appropriately used. First, governments can take steps to improve the awareness of risks among officials and politicians. Second, they can impose upon themselves and lower tiers of government stronger requirements to disclose information about PPP contracts and the fiscal obligations that they create. Third, governments can continue to improve their fiscal planning, budgeting, and accounting in ways that help them choose their expenditure and investment plans rationally. Fourth, they can improve their ability to manage risks by allocating responsibility for taking on risk, developing quantitative monitoring of exposure, and so forth. Under such circumstances, PPPs may be a valuable tool for promoting sound and fiscally responsible investment in infrastructure. 26 References Alesina, A., and R. Perotti (1995), “The Polital Economy of Budget Deficits,” IMF Staff Papers 42. Barro, R.J. (1990), “Government Spending in a Simple Model of Endogenous Growth,” Journal of Political Economy 98. Boeri, T. (2003), “Social Policy Models in Transition: Why Are They So Different from One Another?” In Holzman, R., M. Orenstein, and M. Rutkowski eds.: Pension Reform in Europe: Process and Progress. World Bank (Directions in Development). Davoodi, H., and H. Zou (1998), “Fiscal Decentralization and Economic Growth: A Cross-Country Study.” Journal of Urban Economics 43. De Mooij, R.A., and P.J.G. Tang (2004), “Reforming the Public Sector in Europe: Reconciling Equity and Efficiency.” CPB Netherlands Bureau of Economic Policy Analysis. European Commission (2004), “Public Finances in EMU 2004,” DG ECFIN. Giavazzi, F., and M. Pagano (1990), “Can Fiscal Contractions be Expansionary? Tales of Two Small European Countries.” In Blanchard, O., and S. Fischer (eds): NBER Macroeconomic Annual. MIT Press. Heller, P. (2005), “Understanding Fiscal Space,” IMF Policy Discussion Paper 05/4 (March). Okun, A.M. (1975), “Equality and Efficiency,” Brookings Institution, Washington. Sapir, A. (2005), “Globalization and the Reform of European Social Models,” Background document for the presentation at ECOFIN Informal Meeting in Manchester, September 9. Schuknecht, L., and V. Tanzi (2000), “Reforming Public Expenditure in Industrialized Countries: Are there Trade- offs?” Paper prepared for EC DG ECFIN workshop on Fiscasl Policy in the EMU – New Issues and Challenges (November) Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 27

I. FISCAL CHALLENGES FOR THE EU8 COUNTRIES26 1. During the past 10 years, there have been striking changes in fiscal policy in the EU8 countries. Most of the EU8 countries have engaged in fiscal consolidation aimed at reducing government deficits toward the Maastricht criterion for euro adoption of 3% of GDP. Fiscal consolidation has been complicated by the completion of the transition process which together with the implementation of the acquis communautaire has burdened the expenditure side of government budgets. At the same time, these have been pressured by growing tax competition. The structural nature of the fiscal imbalances in most of the countries certainly has not eased this task. The new member states have coped with these various pressures in strikingly different ways and with quite uneven results. Fiscal contractions based on expenditure restraint have been undertaken in the Baltic countries and Slovakia, while other EU8 governments have pursued erratic fiscal policies and postponed politically difficult reform measures. As a result, in 2004 budget deficits were comfortably below 3% of GDP only in the Baltic countries and Slovenia, while deficits in Poland and Hungary remained stubbornly high (close to 5% of GDP) despite an exceptionally supportive environment. The consolidation efforts have stabilized debt-to-GDP ratios and in some countries led to declining debt burdens, thus contributing to the sustainability of public finances.

2. However, looking further ahead, demographic and budgetary projections show that fiscal positions may easily turn unsustainable in the absence of further fiscal reforms. Although many EU8 countries implemented important reforms over the past years (pension systems reforms, fiscal decentralization, and education reforms, among others), many important challenges remain. Progress in health care system reforms have been slow in most countries, and in many social benefit systems remain insufficiently targeted. Strengthening both the sustainability and quality of public finances is of fundamental importance for delivering the Lisbon Strategy goals of building an inclusive labor market and stimulating the growth of employment, both essential for achieving greater social cohesion and combating poverty within the European Union.

3. It has become increasingly clear that, whatever steps are taken to improve fiscal sustainability and quality, it is equally important to ensure that domestic budgetary rules and institutions contribute towards sound public finances. In the EU8, fiscal discipline clearly seems to depend on the institutional setting: the experiences from recent years suggest that deficit bias, overoptimistic revenue projections, creative accounting and one-off measures may all be linked to underlying institutional weaknesses. 27

1. Fiscal Developments 1995-200428 General government balances

4. Fiscal developments in the EU8 have varied significantly in recent years. In all countries, the fiscal deficit widened in 1998-99 in the wake of Asian and Russian currency crises, but since then developments have differed substantially. The Baltic countries, Slovenia, the Czech Republic, and Slovakia have pursued ambitious fiscal consolidation, while Hungary and Poland have allowed deficits and public debt to widen to worrisome levels in recent years (Figure I.1).

26 Prepared by Paulina Bucon and Emilia Skrok. 27 The role of fiscal institutions in shaping fiscal policy outcomes in the new member states is discussed in Chapter IV of the EU 2005 Public Finance Report and in Ter-Minassian, T. (2005). The EC points out that a common source of fiscal slippage in the new member states was failure of spending ministers and local authorities to internalize the social costs of their demands (the so-called “common pool” problem) as well as fragmentation of the process of formation, monitoring and implementation of budgets. 28 This report is based on information available through end-August, 2005. 28

Figure I.1. EU8 Fiscal Developments 1998-2004 (% of GDP).

Czech Republic Estonia Latvia Lithuania GG debt (left scale) GG debt (left scale) GG debt (left scale) GG debt (left scale) GG balance (right scale) GG balance (right scale) GG balance (right scale) GG balance (right scale) 40 -2 8 3 20 0 25 -1 2 35 -3 -1 -2 1 30 -4 6 15 -2 20 0 -3 25 -5 -1 -3 -2 -4 20 -6 4 10 -4 15 -3 -5 15 -7 -5 -4 10 -8 2 -5 5 -6 10 -6 8 9 0 1 3 4 2 4 2 4 1 9 0 0 1 3 2 3 3 4 8 8 9 8 9 0 1 2 9 0 0 0 9 0 0 0 0 0 9 0 0 0 0 0 0 9 9 9 0 0 0 9 0 0 0 9 9 0 0 9 0 0 0 0 0 9 0 0 0 0 0 0 0 0 0 9 9 9 9 0 0 0 0 9 2 1 1 2 2 2 2 2 2 2 2 2 1 2 1 1 1 2 2 2 2 2 2 2 2 1 1 2 Hungary Poland Slovakia Slovenia GG debt (left scale) GG debt (left scale) GG debt (left scale) GG debt (left scale) GG balance (right scale) GG balance (right scale) GG balance (right scale) GG balance (right scale) 65 -2 50 -1 50 -2 30 -1 -3 -3 -2 -4 45 45 25 -4 -3 -5 -2 60 -6 -5 -4 -7 40 40 20 -6 -5 -8 55 -9 -3 -7 -6 35 35 -10 15 -11 -8 -7 -12 50 -9 30 -8 30 -13 10 -4 4 1 3 4 9 0 1 9 8 9 0 1 2 3 4 8 8 3 2 0 2 9 0 9 0 0 0 0 0 0 0 0 0 0 0 9 0 0 0 0 0 0 9 0 9 9 0 9 9 9 9 0 0 0 0 0 0 0 0 9 9 0 0 0 0 0 9 0 0 9 9 0 0 0 0 0 9 18 23 19 24 20 21 22 2 2 2 2 2 2 2 2 2 2 1 1 1 2 1 1 1 2 2 2 2

Notes: CZ - without one-off imputation of state guarantees in 2003; HU - excluding one-off expenditure in 2002; HU, PL - with the effect of pension reform. Source: Eurostat; staff calculations.

5. Among the Baltic countries, Estonia has been running a fiscal surplus in recent years, while both Latvia and Lithuania have reduced deficits to below three percent of GDP. In all three countries, public debt levels are low at less than 20 percent of GDP. Similarly, Slovenia has strengthened its fiscal credibility significantly, with deficits now contained to around two percent of GDP and the debt burden manageable. Slovakia has also taken great strides to bring its fiscal house in order, and is now close to complying with the Maastricht fiscal criteria. The same would appear to be true in the Czech Republic. In contrast, fiscal deficits in Hungary and Poland exceeded five percent of GDP in 2004, with debt levels – particularly in Hungary, close to the critical 60 percent of GDP limit. 29

29 Fiscal deficits include the effect of pension reform (see Table I.1). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 29

6. The fiscal consolidation that has taken place in most countries in the region in recent years reflects mainly strong output growth and administrative measures of various kinds, with real fiscal reforms confined to Slovakia and to a lesser extent the Czech Republic. Slovakia has implemented important fiscal reforms on both the revenue and expenditure side since the late 1990s, and especially in the last couple of years (Box I.5). There have also been important reforms in the Czech Republic, including in the area of social spending as well as public finance management. The positive development in 2004 was further affected by a change in budgetary rules that allowed unspent funds to be carried over to the following year. Meanwhile, efforts to control public finances in Hungary have generally been of an administrative nature, including in 2004 a new methodology for accounting VAT refunds. The fiscal deficit in all countries, except perhaps Estonia, thus remains of a structural rather than cyclical nature.

Table I.1. Effect of Pension Reforms in the EU8 (% of GDP).

Notes: *) CZ - without one-off imputation of state guarantees in 2003 (Eurostat figure: 11.7%); HU - excluding one-off expenditure in 2002 (Eurostat figure 8.5%). Source: Eurostat; staff calculations.

7. Revenue and spending patterns have also varied among in the EU8 countries in recent years. Several countries have been pursuing tax reforms aimed at lowering the overall tax burden, and general government revenues have been on a declining trend in the Baltic countries and Slovakia reaching a level of around 35 percent of GDP (Figure I.2). Expenditures in these countries have been scaled down correspondingly. Meanwhile, revenues have remained broadly stable at around 45 percent of GDP in the other countries in the region, with expenditures trending upward (except in Slovenia). 30

8. Thus, on the fiscal scene, two groups of countries emerge: those with moderate size of government and low deficits (the Baltic countries and Slovakia), and those with relatively large governments and significant fiscal problems (Hungary, Poland, and to some extent the Czech Republic). Slovenia falls in between, with large government but small deficit.

Figure I.2. EU8 General Government Revenue and Expenditure (% of GDP).

Czech Republic Estonia Latvia Lithuania

60 Total expenditure/ GDP 50 Total expenditure/ GDP 45 Total expenditure/ GDP 45 Total expenditure/ GDP Total revenue/ GDP Total revenue/ GDP Total revenue/ GDP Total revenue/ GDP 55 45 50 40 40

45 40

40 35 35 35 35

30 30 30 30 2 7 8 9 0 1 3 4 5 7 0 3 4 9 7 1 2 3 5 6 2 6 8 9 1 8 0 4 5 6 9 9 0 0 0 0 0 0 0 9 9 9 0 0 0 0 9 9 9 9 9 0 9 9 0 9 9 0 0 9 0 9 9 0 0 0 9 9 9 0 9 0 9 0 0 0 9 9 9 0 0 9 9 9 9 0 0 0 0 0 9 0 9 9 9 9 0 0 0 9 0 0 9 9 9 0 0 0 9 9 2 2 2 1 2 2 2 2 1 1 1 1 1 1 1 1 1 2 2 2 1 1 1 1 2 2 2 2 2 1 23 17 20 22 24 15 16 18 19 21 Hungary Poland Slovenia Slovakia

55 Total expenditure/ GDP 55 Total expenditure/ GDP 55 Total expenditure/ GDP 75 Total expenditure/ GDP Total revenue/ GDP Total revenue/ GDP Total revenue/ GDP 70 Total revenue/ GDP 65 50 50 50 60 55 50 45 45 45 45 40 35 40 40 40 30 4 6 7 0 1 2 3 5 5 6 7 8 8 9 0 1 1 2 2 5 4 6 7 9 0 3 3 4 4 8 9 0 0 0 9 0 0 9 9 9 9 9 9 9 0 0 9 9 9 0 0 0 0 9 9 9 9 9 9 0 0 0 0 0 0 0 9 0 9 9 0 0 0 0 0 0 9 9 9 9 9 0 9 9 9 9 9 9 9 9 0 0 0 0 0 9 9 9 0 0 0 9 9 9 9 0 0 0 0 0 0 22 2 21 23 15 16 17 18 19 20 2 2 2 1 1 1 2 1 1 2 1 1 1 1 1 1 1 2 2 2 2 2 2 2 1 1 1 2 2 2

Notes: CZ - without one-off imputation of state guarantees in 2003; HU - excluding one-off expenditure in 2002; HU, PL - with the effect of pension reform. Source: AMECO (ESA 95); staff calculations.

9. Fiscal outcomes in the Visegrad countries have generally fallen well short of targets agreed with the EU, although membership of the EU in 2004 forced additional discipline.30 A comparison of the budget outcomes achieved in the period 2000-2004 with the targets established in the Pre-Accession Economic Programs (2001-2003) reveals significant slippages in terms of both deficit and debt levels, but outcomes in 2004 were broadly in line with post-accession Convergence Programs (Spring 2004 and Autumn 2004) (Figure I.3). Deficit overshoots were caused mainly by higher than expected expenditure whereas revenues in some cases even exceeded the projected targets. On the expenditure side, discretionary spending stimulated by election cycles appears to have played a significant role in some countries.

30 The Baltic countries, notably Estonia and to some degree Latvia, were more successful in achieving program targets, despite lower than expected revenues. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 31

Figure I.3. Visegrad Countries - Divergence from EU Program Fiscal Targets (% of GDP).

General Government Balance

Czech Republic Hungary Poland Slovakia actuals ac t uals actuals ac t uals PEP 2001 4 PEP 2001 4 PEP 2001 4 PEP 2 0 0 1 PEP2002 PEP2002 4 PEP2002 3 PEP2002 PEP2003 3 PEP2003 PEP2003 2 PEP2003 2 2 CP 2004 May CP 2004 May CP 2004 May 1 2 CP 2004 May CP 2004 Nov CP 2004 Nov CP 2004 Nov 0 CP 2004 Nov 0 0 1 -1 -2 0 -2 -2 -4 -3 -1 -4 -4 -6 -2 -5 -6 -8 -6 -3 -8 -10 -7 -4 -8 -12 -5 -10 -9 -14 -10 -6 -12 2000 2001 2002 2003 2004 2005 2006 2007 2000 2001 2002 2003 2004 2005 2006 2007 2008 2000 2001 2002 2003 2004 2005 2006 2007 2000 2001 2002 2003 2004 2005 2006 2007

General Government Gross Debt

Czech Republic Hungary Poland Slovakia actuals actuals ac t uals ac t uals PEP 2001 65 PEP 2001 PEP 2001 60 PEP 2001 55 60 PEP2002 PEP2002 PEP2002 PEP2002 PEP2003 50 PEP2003 PEP2003 55 PEP2003 CP 2004 May CP 20 0 4 May CP 2 0 0 4 Ma y CP 2004 May 55 45 CP 2004 Nov 60 CP 2004 Nov * CP 2004 Nov 50 CP 2004 Nov

40 45 50 35 55 40 30 45 35 25 50 30 20 40 15 25 * including pension reform eff ect 10 45 35 20 2000 2001 2002 2003 2004 2005 2006 2007 2000 2001 2002 2003 2004 2005 2006 2007 2000 2001 2002 2003 2004 2005 2006 2007 2008 2000 2001 2002 2003 2004 2005 2006 2007

Source: PEPs; CPs; Eurostat. Revenue developments 10. EU8 countries have generally reduced the tax burden (i.e. the total amount of taxes and social security contributions) significantly over the last decade, with the largest reductions occurring in countries whose initial tax-to-GDP was the highest (Figure I.4).

Figure I.4. EU8 Change in Tax Burden 1995-2004 Relative to Initial Level.

2

EU15 CZ 0 25LT 30 35 40 45 -2

HU -4

LV EE PL -6

-8 Cumulative change 1995-2004

-10

SK -12 Total taxes, including social contributions, % GDP - base year 1995

Note: EU15 is simple average. Source: Eurostat; staff calculations. 32

Table I.2. EU8 Tax Burden 1995-2004 (% of GDP).

CZ EE LV LT HU 1995 2003 2004 1995 2003 2004 1995 2003 2004 1995 2003 2004 1995 2003 2004 TOTAL TAXES 36.236.236.137.933.432.833.729.128.628.628.527.743.639.239.2 of which Indirect taxes 12.311.411.913.913.112.713.711.511.412.311.911.217.815.816.1 of which VAT 6.3 6.5 n/a 9.8 8.9 n/a 9.3 7.3 n/a 7.7 6.8 n/a 7.7 9.1 n/a Excise duties 2.3 2.5 n/a 2.5 3.3 n/a 2.2 3.3 n/a 1.5 2.9 n/a 3.6 3.3 n/a Direct taxes 9.69.89.410.98.78.67.88.58.48.88.18.010.79.79.6 of which PIT 4.8 4.9 n/a 8.4 7.0 n/a 5.4 5.9 n/a 7.5 6.6 n/a 6.7 7.1 n/a CIT 4.6 4.6 n/a 2.4 1.7 n/a 1.8 1.5 n/a 1.3 1.4 n/a 3.8 2.2 n/a Social security contributions 14.415.114.713.111.511.212.19.08.87.68.68.515.013.513.4

PL SI SK EU8 EU15 1995 2003 2004 1995 2003 2004 1995 2003 2004 1995 2003 2004 1995 2003 2004 TOTAL TAXES 39.435.835.441.140.339.840.630.730.537.634.233.842.041.741.9 of which Indirect taxes 15.815.315.116.116.816.715.611.512.014.713.413.413.613.913.7 of which VAT 7.1 8.2 n/a 0.0 8.9 n/a 9.5 6.8 n/a 7.2 7.8 n/a 6.8 6.9 n/a Excise duties 3.5 3.9 n/a 0.2 3.2 n/a 2.5 2.1 n/a 2.3 3.1 n/a 2.8 2.7 n/a Direct taxes 12.87.27.07.18.48.411.67.25.89.98.58.212.512.912.9 of which PIT 8.1 4.1 n/a 6.1 6.0 n/a 3.6 3.3 n/a 6.3 5.6 n/a n/a n/a n/a CIT 3.3 2.2 n/a 0.5 1.9 n/a 6.1 2.8 n/a 3.0 2.3 n/a n/a n/a n/a Social security contributions 11.714.113.117.814.914.714.412.312.613.312.412.115.613.314.2

Source: Eurostat; AMECO (ESA 95); staff calculations.

For the group of countries as whole, the tax burden declined from 37.6% of GDP in 1995 to 33.8% GDP in 2004, driven by a drastic decline in Slovakia of more than 10 percentage points but also substantial reductions in Estonia, Latvia, Poland and Hungary (only the Czech Republic and Slovenia maintained broadly unchanged tax burdens during this period) (Table I.2) This compares to almost 42 percent of GDP in the EU15. The tax burden remains close to the EU15 average in Slovenia and Hungary, whereas the Baltic countries are substantially below.

11. The structure of taxation in the EU8 has also undergone significant changes. In particular, the share of direct taxes in total tax revenues has declined while the share of social security contributions and indirect taxes has increased. This reflects ongoing cuts in corporate and personal income tax rates, while high social security contribution rates have been largely maintained in the face of increasing pressures on social spending and some indirect tax rates have been increased (or bases broadened) in relation to EU accession. Nevertheless, as a share of GDP, all major tax categories have declined in the region as a whole and in most EU8 countries. The notable exceptions are social security contributions in the Czech Republic, Lithuania, and Poland, and both direct and indirect taxes in Slovenia, all of which are now higher than in the mid-1990s.

12. Compared to the EU15, EU8 countries in 2004 collected a relatively larger share of tax revenues from indirect taxes (and to a lesser extent social security contributions) and lower share from direct taxes. These differences are also reflected to a certain extent in the differences in the statutory tax rates, where the EU8 countries have higher-than-average consumption tax rates, but lower-than-average corporate and personal tax rates. Since direct taxes are more effective instrument for income redistribution, the differences between the EU8 and the EU15 suggest a stronger focus on redistribution in the old member countries. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 33

Figure I.5. EU8 Composition of Tax Burden 1995-2004 (% of GDP).

Social security contributions 50 Direct taxes Indirect taxes 40

30

20

10

0 4 5 0 4 5 0 4 5 4 0 5 0 4 5 0 4 5 0 4 5 0 4 5 0 4 5 0 0 9 0 0 0 9 0 0 9 0 9 0 0 0 9 0 0 0 9 0 9 0 0 9 0 0 9 0 9 0 0 0 9 0 0 9 0 0 0 9 0 9 0 0 9 0 0 9 0 0 9 0 0 9 2 1 2 2 2 1 2 2 1 2 2 1 2 2 2 1 2 1 2 2 1 2 2 1 2 2 1 CZ EE LV LT HU PL SI SK EU15

Source: Eurostat; AMECO (ESA 95); staff calculations.

13. Since the mid-1990s, a number of EU8 countries have implemented important reforms to their tax systems (Box I.1). The reforms differed in coverage and depth, but were generally aimed at reducing the tax burden on labor, particularly at the low-middle end of the pay scale, and at reducing corporate income tax rates while broadening bases. At the same time, however, in some countries personal and corporate income tax bases were eroded through the introduction of various tax relief schemes. Slovakia implemented a comprehensive tax reform that unified tax rates and eliminated most exemptions.

Box I.1. EU8 Tax Reforms EU8 countries have generally been reducing corporate income tax (CIT) rates while curtailing favorable special tax regimes (see tables below). Top CIT rates now vary between 28% in the Czech Republic and 15% in Latvia and Lithuania (Estonia has no tax on re-invested earnings). Some countries (including Slovakia) have also removed the double taxation of dividends. At the same time, some countries (including the Czech Republic) have introduced accelerated depreciation and increased R&D allowances.The tendency to reduce corporate tax rates and change favorable special tax regimes in the EU8 states has been reflected in effective tax rates. In the second half of the 1990s, effective corporate tax rates were growing in the EU-15, but falling in the EU-8 countries. Since then, both trends appear to have reversed and some convergence taking place.

Figure A. EU8 Effective Corporate Income Tax Rates (macro backward-looking approach). 25 CZ EE LV LT HU SK 20 EU- 8 EU- 15

15

10

5

0 1995 1996 1997 1998 1999 2000 2001 2002

Note: EU8 is simple average. Source: World Bank EU-8 Quarterly Economic Report, October 2004. 34

Recent changes in personal income taxes in the EU8 countries included mainly adjustments in tax allowances and exemptions aimed at social assistance (the Czech Republic, Estonia, Lithuania), but in some countries also lowering of tax rates. Most countries retain different personal income tax rates depending on income level and source, with only Slovakia moving to a single rate (but retaining a certain tax-free income threshold). Top marginal tax rates now vary between 50% in Slovenia and 19% in Slovakia.EU8 countries have continued their efforts towards harmonization of their indirect taxes (VAT and excises) with the EU15. Standard rates VAT now vary between 18% in the Baltic countries and 25% in Hungary. All countries have retained reduced rates for selected goods and services, while mostly rationalizing the items included in this group. The majority of the EU8 countries requested transitional periods for applying the standard rate on certain items, especially constructing services and heating (as well as the turnover threshold for SMEs), and for harmonizing certain excise duty rates (notably, excise duty rates on mineral oils, tobacco and alcohol are significantly lower in the EU8). These excises are being gradually raised. Table A. Top statutory tax rate on personal income.

1995 1998 2003 2004 CZ 40 40 32 32 EE 26 26 26 26 LV 25 25 25 25 LT 33 33 33 33 HU 44 44 40 38 PL 45 40 40 40 SI 50 50 50 50 SK 42 42 38 19

Table B. Top statutory tax rate on corporate income.

1995 1998 2003 2004 CZ 41 35 31 28 EE 26 26 26 26 LV 25 25 19 15 LT 29 29 15 15 HU 19.6 19.6 19.6 17.7 PL 40 36 27 19 SI 25 25 25 25 SK 40 40 25 19

Table C. VAT rates.

Standard rate 2004 Domestic zero Reduced 1998 2000 2003 rate1 rate(s) Standard rate Czech Republic 22 22 22 no 5.0 19.0 Estonia 18 18 18 yes 5.0 18.0 Latvia 18 18 18 yes 5.0 18.0 Lithuania 18 18 18 no 9.0 (5.0) 18.0 Hungary 25 25 25 yes 15.0 (9.0) 25.0 Poland 22 22 22 yes 7.0 (3.0) 22.0 Slovak Republic 23 23 20 yes 15.0 (5.0) 19.0 Slovenia - 19 20 no 8.5 20.0

"Domestic zero rate" means tax is applied at a rate of zero to certain domestic sales. It does not include zero rated exports.

14. With only marginal changes in the taxation of labor, tax wedges remain high in the EU8 countries, especially for low wage earners (Figure I.6).31 Hungary was most successful in recent years in reducing its tax wedge, albeit from a high level.

31 Tax wedge on labor costs is measured as the hypothetical income tax on gross wage earnings plus employee and employer social security contributions, expressed as a percentage of the total labor costs (defined as gross earnings plus employer social security contributions plus payroll taxes, where applicable). Low wage earners defined as single persons without children earning 67% of the average of a full-time production worker (APW). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 35

Figure I.6. EU8 Tax Wedge on Labor Costs - Low Wage Earners.3

SK

SL

PL

HU

LT

LV 2004 2002 EE 1998 CZ

EU15

0 1020304050 Note: Slovenia data are for 2003. Source: Eurostat.

15. The level of general government revenues and grants in the EU8 countries is relatively high compared to a broader group of countries but similar to the average in the Central Europe and Asia region (Figure I.7).

Figure I.7. General government, total revenue and grants (% GDP).

50

40

30

20

10

0 EU8 EU 15 Sub- North Africa Latin Saharan America Af ric a Pacific & North Carabean America & America South As ia Middle East Middle East Asia & Asia East Europe and Europe Central Asia Central

Note: 2000-2004 average Source: IMF World Economic Outlook Database; staff calculations.

Expenditure developments 16. Expenditure developments have largely mirrored revenue developments over the past decade, with even larger contractions in spending relative to revenue in countries that initially had large fiscal imbalances (notably the Czech and Slovak Republics).32 Thus, the EU8 average public expenditure-to-GDP ratio declined from 46.5% of GDP in 1995 to 42.7% of GDP in 2004. Incidentally, a similar pattern was followed in the EU15 countries, where, however, spending still amounted to nearly 48% of GDP in 2004.

32 There are several complications in making cross-country comparisons of public expenditure levels or composition (Box I.2). 36

Box I.2. Public Spending Measurement Issues Cross-country comparisons of levels or composition of public spending are complicated by a number of factors: • The extent to which countries rely on tax expenditures as opposed to direct expenditure. While there is no consistent information on the amount of tax expenditure across countries, largely reflecting the absence of an agreed operational framework for measuring tax expenditures, the impression is that these are substantial in some countries. For example, in Poland, personal income tax expenditure programs, introduced in 1992, were originally used to compensate lower-income taxpayers for government withdrawal of price subsidies. However, the number and cost of these programs rapidly increased in the following years. While the marginal rates for the three income tax brackets are 40%, 30%, and 19%, income tax exemptions and deductions allowed under tax expenditure programs lowered the effective rates applicable to these brackets to 28%, 18%, and 13%, respectively, in 2003. • The taxation of social benefits. In some OECD countries, transfers are subject to broadly the same tax treatment as wage income (e.g. in the Nordic countries), while in others they are predominantly untaxed (e.g. sickness benefits in Poland and Hungary). For a given amount of net social transfers paid by the public sector and received by households, the level of public expenditures will be higher in the first group of countries. • The reliance on (mandatory and/or voluntary) private insurance schemes for pensions, unemployment, and/or health care systems. • Financing projects through public-private partnerships (PPPs) as opposed to conventional public investment. Several countries are increasingly relying on PPPs to finance investment projects (e.g. Hungary, Poland, and Slovakia). While this would show up as lower public spending in the short term, fiscal costs may materialize later (see Chapter V for a further discussion of these issues). • The use of guarantees on loans as a substitute for direct capital transfers or subsidies. Guarantees issued to state-owned enterprises (e.g. Czech Republic and Poland) are generally not reflected in general government expenditure in the year they are extended although they may show up in the fiscal accounts later when the guarantees are called. Source: OECD ( 2004); Polackova et al. (2004).

17. In the Baltic countries, spending increased sharply towards the end of the 1990s in response to the Russia crisis, but subsequently reverted to its downward path. Spending levels of around 35% of GDP are now among the lowest in the EU. Developments in the Visegrad countries have varied significantly. The Czech Republic cut back spending dramatically in the mid-1990s, but since then it has remained broadly stable at around 45% of GDP. Following a surge in spending in the difficult years of 1996-97, Slovakia similarly undertook an ambitious rationalization of spending programs and downsizing of the public sector, and spending declined to around 40% of GDP by 2004. Poland made good progress in cutting back public spending in the second half of the 1990s, but this was largely reversed during the period 2001-04. Hungary also allowed a significant increase in spending during these years. Slovenia maintained a high and stable spending ratio.

18. While EU8 expenditure trends in a longer retrospective generally have been dominated by the need for downsizing bloated public sectors, shorter term developments in some countries have been influenced by external or domestic shocks as well as business and political cycles (Box I.3).

Box I.3. Factors Affecting Expenditure Developments The literature typically identifies three typical factors that may affect public expenditure developments: (i) over the medium-long term, increases in per-capita incomes raises demand for some public goods and services (e.g. healthcare and education), thus leading to higher government spending (Wagner’s law); (ii) failure to rein in spending after periods of higher spending related to shocks, business cycles, or political cycles (hysteresis effect); and (iii) cyclical developments related to the factors mentioned in (ii). It is clear from the discussion above, that there is little support for the “Wagner” effect in the EU8 countries. Expenditures in most countries have trended down as per capita incomes increased, mirroring tax reductions and lower revenues. This reflects the initially bloated public sectors in all EU8 countries. Similarly, there is no clear evidence of any hysteresis or ratchet like effects. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 37

Over the shorter term, there is some evidence of counter-cyclical expenditure developments in the Baltic countries, but not in the Visegrad countries or Slovenia (Figure A). In Slovakia, on the contrary, the widening of the output gap after 1998 was accompanied by a strong compression of public spending

Figure A. Public Expenditure and Output Gap (%).

Czech Republic Estonia Latvia Lithuania Total expenditure/GDP (left scale) Total expenditure/GDP (left scale) Total expenditure/GDP (left scale) Total expenditure/GDP (left scale) Output gap (right scale) Output gap (right scale) Output gap (right scale) Output gap (right scale) 60 1 50 1 45 2 45 2 0 1 55 0 1 -1 0 50 -1 45 -2 0 40 -3 -1 45 -2 -4 40 -1 -2 -5 -3 40 -3 40 -6 -2 35 -4 -7 35 -4 -3 -8 -5 30 -5 35 -9 35 -4 30 -6 5 7 9 6 8 0 3 4 2 1 5 7 8 8 6 6 7 0 5 5 6 9 1 0 7 4 9 1 2 3 8 9 1 4 3 2 4 0 2 3 9 9 9 9 0 9 0 0 0 0 9 9 9 9 9 9 9 9 9 9 9 0 9 9 0 0 9 0 9 0 0 0 0 0 0 0 0 0 0 0 9 9 9 9 9 0 0 0 0 0 9 9 9 9 9 9 9 9 9 9 9 0 9 9 0 0 9 0 9 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 2 2 2 2 2 1 1 1 1 1 1 1 1 2 1 1 1 2 2 2 1 1 1 2 1 2 2 2 2 2 2 2 2 2 2 Hungary Poland Slovakia Slovenia Total expenditure/GDP (left scale) Total expenditure/GDP (left scale) Total expenditure/GDP (left scale) Total expenditure/GDP (left scale) Output gap (right scale) Output gap (right scale) Output gap (right scale) Output gap (right scale) 55 2 55 1 70 1 55 2

0 65 1 0 1 60 50 -1 50 55 50 0 -2 -1 0 50 45 -3 45 45 -1 -2 -1 -4 40

45 -2 40 -5 35 -3 40 -2 5 5 5 7 7 6 6 6 8 9 7 9 8 0 8 1 9 1 0 2 0 1 2 4 4 2 4 3 3 3 9 9 9 9 9 0 0 0 0 0 9 9 9 9 9 9 9 9 9 9 9 9 0 9 0 9 0 9 0 0 0 0 0 0 0 0 0 0 0 0 9 9 9 9 0 0 9 0 0 0 9 9 9 9 9 9 9 9 9 9 9 9 0 9 9 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 15 16 18 17 20 21 19 22 24 23 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 2 1 2 2 2 2 2 2 2 2 2 2 2 2 2

Notes: output gap measures as the gap between actual and potential gross domestic product at 1995 market prices (% of potential GDP at market prices). CZ -without one-off imputation of state guarantees in 2003;HU - excluding one-off expenditure in 2002; HU, PL - with the effect of pension reform Source: AMECO (ESA 95); staff calculations.

There is also some evidence of political cycle dominated spending patterns in Estonia, Latvia, and Hungary, where spending has seemed to increase in periods leading up to Parliament elections (Figure B). On the other hand, no such patterns can be identified in the other EU8 countries.

Figure B.1. Political Economy at Work . Figure B.2. Political Economy at Work in the Visegrad in the Baltic Countries. Countries. Total expenditure/GDP 50 80 CZ HU PL SK Parliamentary elections 45 CZ HU PL SK

40 65

35

30 Total expenditure/GDP 50 EE LV LT Parliamentary elections 25 EE LV LT

20 35 1992 1994 1996 1998 2000 2002 1992 1994 1996 1998 2000 2002

Source: AMECO (ESA 95); staff calculations. Notes: CZ -without one-off imputation of state guarantees in 2003; HU - excluding one-off expenditure in 2002; HU, PL - with the effect of pension reform. Source: AMECO (ESA 95); staff calculations. 38

19. The composition of government spending has not changed much over the last decade, although a few EU8 countries have managed to improve the structure of spending somewhat (Figure I.8). Generally, lower interest rates have reduced the debt service burden, thus creating some fiscal space for productive spending. Slovakia has gradually reduced subsidies and collective consumption to the benefit of higher public investment and social spending. Nevertheless, subsidies remain relatively high and capital spending relatively low. In contrast, the Czech Republic has been increasing subsidies and only in recent years reversed a squeezing of capital expenditures. Hungary has made some progress since 2000 in reducing subsidies and raising public investment. However, sizeable interest savings have also been used to raise social spending and government consumption. In Poland, the structure of public finances has also improved somewhat with a rationalization of subsidies, social transfers, and collective consumption along with more resources for investment, although much remains to be done (and the large increase in “other” spending warrants scrutiny). In the Baltic countries, the recent tendency to increase subsidies (Estonia and Lithuania) and cut back social spending (Latvia and Lithuania) is a source of concern.

Figure I.8. EU8 Public Expenditure Composition (% of GDP).

60

50

40

30

20

10

0 0 5 0 4 5 4 0 5 4 5 0 0 4 4 5 0 4 5 0 5 5 0 4 5 0 4 4 9 0 0 9 0 9 0 0 9 0 0 0 9 0 0 0 9 0 9 0 0 0 0 9 0 9 0 0 0 9 9 0 9 0 0 9 0 0 0 9 0 0 0 9 0 0 0 0 0 0 9 0 9 9 1 2 2 1 2 1 2 2 2 2 1 2 1 2 2 1 2 2 1 2 2 2 1 2 2 1 2 CZ EE LV LT HU PL SI SK EU15 Subsidies Interes t Social transfers in kind Social benefits other than in kind Collective consumption Gross fixed capital formation Other

Notes: ESA 95 methodology. Social transfers in kind consist mainly of education and health services, although other services such as housing and cultural and recreational services are also frequently included. Social benefits other than in kind include: social security benefits in cash provided by social security schemes; private funded social benefits; unfunded employee social benefits (incl. (a) the continued payment of normal, or reduced, wages during periods of absence from work as a result of ill health, accident, maternity, etc.; (b) the payment of family, education or other allowances in respect of dependants; (c) the payment of retirement or survivors’ pensions to ex- employees or their survivors, and the payment of severance allowances to workers or their survivors in the event of redundancy, incapacity, accidental death, etc. (if linked to collective agreements); (d) general medical services not related to the employee’s work; (e) convalescent and retirement homes.); and (f) social assistance benefits in cash (ex. transfers made in response to natural disasters, recorded under other current transfers or under other capital transfers). Source: AMECO (ESA 95); staff calculations.

20. Social protection/benefits remain the largest category of public spending in all EU8 countries, although below the average EU15 level (Table I.3 and Table I.4). Social benefits and transfers in 2003 amounted to around 20% of GDP in the Baltic countries, Slovakia, and Slovenia, and 24-28% of GDP in the Czech Republic, Poland, and Hungary (ESA95). The second largest spending category is collective consumption in the range of 8-12% of GDP, in all cases but Lithuania higher than the EU15 average. It is interesting to note that the government wage bill appeared particularly large in Hungary and Slovenia, and to a lesser extent in Poland. Gross fixed capital formation varied between only 1.5% of GDP in Latvia to more than 4% of GDP in the Czech Republic, with most countries around 3% of GDP – somewhat higher than the EU15 average. Subsidies remained sizeable in the Czech Republic and to a lesser extent in Slovakia, Hungary, and Slovenia, while other countries in the region were below Western European levels. Finally, interest payments absorbed a significant share of resources in Hungary and Poland in line with their relatively high debt levels. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 39

Table I.3. EU8 Economic Classification of Public Expenditure 2003 (% of GDP).

EU15* CZ EE HU LV LT PL SK SI GFS 2001 Compensation of employees 11.8 4.3 n/a 12.7 8.9 8.9 9.7 8.6 12.8 Use of goods and services 6.6 4.8 n/a 6.7 6.0 5.7 8.2 6.4 10.3 Interest 3.0 0.8 n/a 4.2 0.8 1.3 2.8 2.5 1.6 Subsidies 1.6 9.7 n/a 2.9 0.8 0.3 0.7 1.8 1.4 Grants 0.7 0.1 n/a 0.0 0.0 0.0 0.0 0.1 0.0 Social benefits 20.2 12.5 n/a 16.3 10.0 12.8 18.3 17.4 18.6 Other expense 2.6 9.8 n/a 3.7 6.6 1.4 2.0 2.6 1.7 Expense 48.241.9n/a46.533.231.043.539.346.3

ESA 95 Subsidies 1.2 2.7 0.9 1.5 0.8 0.8 0.4 1.7 1.5 Interest 3.4 1.3 0.3 4.2 0.8 1.3 3.1 2.5 2.1 Collective consumption 8.2 12.3 8.8 10.8 11.0 7.6 9.1 11.0 8.4 expenditure Social benefits other than social 16.3 12.3 9.6 14.2 9.6 9.2 17.6 11.4 17.2 transfers in kind Social transfers in kind 12.5 11.7 9.9 13.4 10.8 10.8 8.5 8.9 11.9 Other current expenditure 1.9 1.2 0.9 1.7 0.8 0.2 1.5 2.2 1.3 Total current expenditure 43.5 41.6 30.3 45.7 33.7 29.9 40.2 37.8 42.4 Gross fixed capital formation 2.3 4.2 3.4 3.4 1.5 3.0 3.4 2.6 2.8 Total expenditure 47.753.235.850.735.734.249.039.248.2

Notes: * 2002. GFS 2001 on accrual basis except for LV and SI (on cash basis). GFS 2001 for EU15: simple average excluding GR, IT, PT. Source: AMECO (ESA 95); IMF Government Finance Statistics 2001; staff calculations.

21. EU8 countries tend to allocate a relatively larger share of spending toward “efficiency-oriented” programs, about the same on “basic functions” of the state, and less on income redistribution.33 Efficiency- oriented outlays ranged from 13% of GDP in Latvia to 22% of GDP in Hungary, compared to 17% of GDP in the EU15). All EU8 countries (except for Poland) allocated a larger share of total spending to this group than the EU15 (more than 42% of total spending in the Czech Republic, Hungary and Lithuania). Outlays on economic affairs were particularly high in Hungary and the Czech Republic, outlays on health were relatively large in Slovakia, Slovenia, and the Czech Republic (and low in the Baltic countries), and outlays on education very high in Slovenia (and low in the Czech and Slovak Republics). Spending on basic government functions amounted to 8-9% of GDP (except for Hungary at more than 12% of GDP)–below the EU15 average of 10% of GDP. This group of outlays accounted for more than 20% (EU15 average) of total spending in Hungary, Latvia, Lithuania and Slovakia, and less than 20% in the remaining EU8 countries. Finally, spending aimed at income redistribution ranged from 10% of GDP in Lithuania to 21% of GDP in Poland, similar to the EU15 average and the only EU8 country to devote a larger share of total spending to this purpose.

Table I.4. EU8 Functional Classification of Public Expenditure 2003 (% of GDP).

EU15 CZ EE HU LV LT PL SK SI General public services 7.0 7.4 3.2 8.2 5.3 4.1 7.0 5.2 8.5 Defense 1.6 1.8 1.8 1.3 1.3 1.5 1.0 1.8 1.4 Public order and safety 1.6 2.2 2.7 2.1 2.4 2.0 1.2 2.0 2.0 Economic affairs 4.6 11.7 3.8 5.7 3.7 4.2 3.2 5.1 3.5 Environment protection 0.7 2.0 0.6 0.8 0.4 0.5 0.6 0.7 0.5 Housing and community amenities 0.9 0.9 0.6 1.1 0.9 0.6 1.5 1.1 0.4 Health 6.2 6.5 4.1 5.7 3.3 4.3 3.0 2.3 6.8 Recreation, culture and religion 1.0 1.3 2.2 2.2 1.4 0.8 0.8 1.0 0.9 Education 5.7 4.9 6.4 6.1 6.2 6.0 6.3 4.3 5.9 Social Protection 19.1 14.6 10.4 17.0 10.8 10.1 19.9 15.7 18.3 Total Outlays 48.353.235.850.235.734.144.539.248.2

Source: Eurostat.

33 The “efficiency-oriented” expenditures comprise categories of economic affairs (including support programs and subsidies to mining, manufacturing, agriculture, energy and service industries), environment protection, health and education. “Basic functions” of the state and pure public goods include outlays on general public services, defense, public order and safety. These two broad categories essentially pursue “economic” objectives as opposed to programs aimed at income redistribution: social protection and housing and community amenities. 40

22. In a broader perspective, the level of general government expenditure in the EU8 countries is slightly above the average of the Europe and Central Asia region while below the levels observed in some of the most developed European countries (Figure I.9).

Figure I.9. General government, total expenditure (% GDP).

50

40

30

20

10

0 EU 8 EU15 Sub- North Africa Lat in America Saharan Africa Pacific & North Carabean America & America South Asia South Middle East Middle East Asia & Asia East Europe and Europe Central Asia Central Note: 2000-2004 average

Source: IMF World Economic Outlook Database; staff calculations.

Debt developments 23. Public debt levels in the EU8 have generally stabilized in recent years, although at high levels in some of the Visegrad countries (Figure I.1). In the Baltic countries, debt burdens have eased further to low levels (especially in Estonia) on account of buoyant output growth and strong fiscal positions. In Slovakia, debt levels have also come down from relative high levels a few years ago, and seem to have stabilized at just over 40% of GDP. In the Czech Republic, debt has shot up since the late 1990s on the back of sizeable fiscal deficits and assumption of contingent liabilities, but appears to have leveled off at around 35% of GDP. In Poland and Slovenia, debt has stabilized at just below 50% and 30% of GDP, respectively. Following a strong consolidation phase in 1998-2001, trends have reversed and the general government debt-to-GDP ratio has climbed back toward the critical 60% of GDP ratio.

24. In addition to generally improved fiscal balances, the broadly favorable debt developments in most EU8 countries have been aided by strong output growth, lower interest rates, and exchange rate appreciation. Decomposing debt dynamics into factors within and outside the government’s control, the latter “snowball effect” has generally been negative (reducing debt ratios) in recent years in the EU8 countries, with Poland during 2001-03 the main exception (Table I.5 and Appendix 1). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 41

Table I.5. Snow Ball Effect on Debt (% GDP).

1999 2000 2001 2002 2003 2004 EUR-15 0.6 -0.4 1.5 1.2 2.3 0.2 Czech Republic 0.4 0.1 -0.2 0.4 -0.3 -1.5 Estonia 0.2 -0.4 -0.3 -0.2 -0.1 -0.2 Latvia 0.0 -0.3 -0.2 -0.5 -0.6 -1.1 Lithuania 1.9 0.7 0.2 -0.1 -0.5 -0.9 Hungary n/a -2.6 -1.5 -1.9 -1.4 -0.4 Poland n/a -0.8 1.4 1.9 1.2 -0.9 Slovenia n/a 0.2 -0.5 -0.6 -0.1 -0.3 Slovakia n/a -0.5 0.3 -0.3 -1.2 -1.8

Source: AMECO (ESA 95), staff calculations.

25. On the other hand, “hidden” fiscal deficits have generally added to the debt burdens of EU8 countries. Hidden fiscal deficits in the EU8 amounted to 1.3% of GDP annually (unweighted average) in the period 2000-2003 (Figure I.10). Hidden deficits were particularly large in Slovakia (5% of GDP in 2002-2003), and the Czech Republic and Estonia (roughly 2% of GDP in 2000-03). However, in Estonia the high hidden deficit comes from large accumulation of financial assets (securities other than shares). This is quite typical for governments in surplus with relatively small debts. In the Czech Republic, the hidden deficit reflects in part the reclassification as government of the Czech Banking Consolidation Agency. Finally, the finding for Slovakia is influenced by large privatization inflows, which have been adjusted for in the measurement of the hidden deficit but which were not in fact used to reduce debt (but rather accumulate reserves to finance pension reform).

Figure I.10. EU8 Hidden Fiscal Deficit 2000-03 (% of GDP).

5.0 2000-2001

4.0 2002-2003

3.0

2.0

1.0

0.0

-1.0 CZ EE LV HU PL SK LT

Notes: simple average; in some countries the revaluation effect is calculated using central government foreign debt currency structure. Source: AMECO (ESA 95); EBRD; MOFs; CBs; Eurostat; and staff calculations.

Intergovernmental fiscal relations 26. In recent years, the management of public finances has been influenced by a process of decentralization whereby the budgetary autonomy of lower levels of government has been increasing.34 In 2003, both expenditure and revenue shares at the local levels of government were around 20% of general government totals, comparable to EU15 average levels (Table I.6). Local governments were relatively larger

34 Intergovernmental fiscal relations are discussed in more detail in Chapter II. 42 in the Czech Republic, Hungary, Latvia, and Poland. At the same time, the share of central government in general government revenue and expenditure was generally significantly higher in the EU8 countries than in the EU15, reflecting the lesser reliance on other elements of the general government (notably social security funds) in the former group of countries.

Table I.6. EU8 Expenditure and Revenue at Central and Local Government Level 2003.

Total expenditure Total revenue % of total % of total EU 15 Central government 44.4 42.4 Local government 19.7 20.5 CZ Central government 68.6 61.9 Local government 21.6 26.2 EE Central government 67.1 68.9 Local government 22.2 19.5 LV Central government 53.2 50.1 Local government 24.3 24.8 LT Central government 52.0 48.3 Local government 18.6 19.6 HU Central government 52.2 47.6 Local government 21.9 24.4 PL Central government 44.9 38.9 Local government 24.2 25.6 SI Central government 51.9 50.3 Local government 15.8 16.5 SK Central government 52.9 47.9 Local government 15.2 16.5

Notes: ESA 95. The general government sector is divided into four subsectors: (a) central government (S.1311); (b) state government (S.1312); (c) local government (S.1313); and (d) social security funds (S.1314). The subsector central government (S.1311) includes all administrative departments of the State and other central agencies whose competence extends normally over the whole economic territory, except for the administration of social security funds. Included in subsector S.1311 are those non-profit institutions which are controlled and mainly financed by central government and whose competence extends over the whole economic territory. The subsector local government (S.1313) includes those types of public administration whose competence extends to only a local part of the economic territory, apart from local agencies of social security funds. Included in subsector S.1313 are those non-profit institutions which are controlled and mainly financed by local governments and whose competence is restricted to the economic territories of the local governments. Source: Eurostat; and staff calculations.

27. Local governments in the EU8 have generally not been important sources of fiscal deficits in recent years (Figure I.11). In 2003, local government budgets in most EU8 countries – even those with the largest local governments – were broadly balanced. The situation was not significantly different in 2001-02, although local government deficits in these years tended to be slightly higher. This likely reflects to a large extent binding borrowing constraints on local governments. Nevertheless, the figures do not reveal any potential hidden deficits and accumulation of arrears or contingent liabilities.

Figure I.11. EU8 Balances at Various Levels of Government 2001-2003.

5

0

-5

-10 Central government Local government Social security funds -15 2003 2002 2001 2003 2002 2001 2003 2002 2001 2003 2002 2001 2003 2002 2001 2003 2002 2001 2003 2002 2001 2003 2002 2001 2003 2002 2001 EU 15 CZ EE LV LT HU PL SI SK

Source: Eurostat. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 43 2. Medium-Long Term Fiscal Prospects Fiscal objectives 2005-08

28. The Baltic countries and Slovenia already comply with the Maastrict fiscal criteria for Euro adoption, and the Visegrad countries plan to follow suit in 2006-08. According to the most recent convergence programs, Slovakia would reduce its general government deficit to below 3% of GDP by 2006, Poland and Hungary by 2007, and the Czech Republic by 2008 (Figure I.12). All countries are planning an improvement in their respective primary, cyclically adjusted balances by 2007: in the period 2005-2007, the Czech Republic would run the highest primary, cyclically adjusted deficit, while Hungary would have reached a primary surplus already in 2005. The medium fiscal plans for the Baltic countries and Slovenia envisage sustaining the fiscal deficit below 3% of GDP in the period 2004-07 and fulfilling other pre-conditions (inflation, interest rates) for joining the EMU in 2007. While all countries (except Hungary) project a worsening of their debt levels over the period, they would remain safely below the 60% of GDP threshold. In the Visegrad countries, these prospects, however, rely on significant fiscal adjustment in the period under review.

29. The envisaged fiscal consolidation would rely mainly on reduced expenditures. Revenue rations would edge up in Lithuania and Slovakia, while easing in the Czech Republic, Estonia, and Hungary. Expenditures would be cut back by about 2.5 percentage points of GDP in the Visegrad countries (notably the Czech Republic and Hungary) and by 0.6 percentage points of GDP in the Baltic countries (driven by Estonia).

Figure I.12. EU8 Fiscal Plans 2005-07 (% or % of GDP). Czech Republic Estonia Latvia Lithuania Total expenditure/ GDP Total expenditure/ GDP Total expenditure/ GDP Total expenditure/ GDP Total revenue/ GDP Total revenue/ GDP Total revenue/ GDP Total revenue/ GDP 50 GG balance (right scale) -2 50 GG balance (right scale) 2 45 GG balance (right scale) 0 40 GG balance (right scale) -1

45 45 40 35 40 -3 40 1 35 35 35 -1 30 -2 30 30 -4 30 0 25 25 25 25

20 -5 20 -1 20 -2 20 -3 2004 2005F 2006F 2007F 2004 2005F 2006F 2007F 2004 2005F 2006F 2007F 2004 2005F 2006F 2007F

Hungary Poland Slovenia Slovakia Total expenditure/ GDP Total expenditure/ GDP Total expenditure/ GDP Total expenditure/ GDP Total revenue/ GDP Total revenue/ GDP Total revenue/ GDP Total revenue/ GDP 55 GG balance (right scale) -1 55 GG balance (right scale) -1 55 GG balance (right scale) 0 45 GG balance (right scale) -1

50 50 -2 50 -2 40 45 45 -3 45 -1 -2 40 -3 40 -4 40 35

35 -4 35 -5 35 30 -2 -3 30 30 -6 30 -5 25 25 25 -7 25

20 -6 20 -8 20 -3 20 -4 2004 2005F 2006F 2007F 2004 2005F 2006F 2007F 2004 2005F 2006F 2007F 2004 2005F 2006F 2007F Note: HU, PL – plan including 100% cost of pension reform. Source: Eurostat (Convergence Programs November 2004). There are several risks underlying these medium-term fiscal plans: macro-risks, including weaker output growth (e.g. as a result of higher oil prices and failure to engineer a recovery in Western Europe); political cycle risk (pre-election spending); failure to make adequate fiscal room for co-financing of EU structural funds or the provision of generous topping up of direct payments to agriculture; realization of contingent liabilities, including related to PPPs; and in a somewhat longer perspective, population ageing and pressures on pension and health care expenditures as well as fiscal costs related to compliance with the EU social inclusion policy and Lisbon strategy. 44 Medium-long-term fiscal sustainability 30. Current fiscal positions in some of the EU8 countries are unlikely to be sustainable over the longer term. The simplest way of looking at fiscal sustainability is to compare current primary fiscal balances with those that would be needed to maintain current debt levels.35 These basic calculations show that current primary fiscal balances are lower than what would be required to stabilize current debt ratios in the Czech Republics, Lithuania and Poland. Only Estonia and Latvia have some room for lasting fiscal expansions if they wish to maintain their current debt burdens.36

Table I.7. EU8 Debt Stabilizing Primary Fiscal Balance 2004 (% of GDP).

Actual primary Primary balance Real intererst rate 4% 6% balance 2004 2005-2007 * CZ -1.1 -0.4 -1.8 -2.4 EE -0.2 -0.1 2.0 0.2 HU -2.2 -1.2 -1.1 0.4 LV -0.7 -0.5 0.0 -0.8 LT -0.9 -0.5 -1.5 -0.9 PL -1.1 -0.3 -4.2 -0.5 SK -1.2 -0.4 -1.1 -0.4 SI -0.8 -0.3 0.0 0.1 Notes: *Projections from Convergence Programs. The primary fiscal balances required to stabilize current debt levels were calculated according to the equation: ⎛ 1+ n ⎞ x = ⎜ −1⎟ *b ⎝ (1+ g)(1+ P ) ⎠

where n – real interest rate, g – real GDP growth, – inflation rate, b – current government debt as a percentage of GDP.Real growth rates assumed in calculations: 4% for CZ, HU, PL, SK, and SI; and 6% for the Baltic countries (based on the average real growth rate projected for 2005-2007 in Convergence Programs); Inflation rates used are average projections for 2005-07 from Convergence Programs. Source: Burnside (2004); Eurostat; and staff calculations.

31. Clearly the assessment of long-term fiscal sustainability is a much more complex issue than suggested by the simple exercise presented above. More sophisticated analysis provided by the European Commission (2005) that takes into account underlying demographic changes and expenditure dynamics confirms that the Czech Republic faces severe challenges, and that fiscal pressures will also mount in several other EU8 countries over the longer term on current policies. Age-related expenditures (pensions and health) are projected to rise rapidly in the Czech Republic and Slovenia, reaching 15% and nearly 20% of GDP in 2009, respectively, and over 25% of GDP in both countries by 2050 (Table I.8). Meanwhile, some of the countries that have undertaken comprehensive pension reforms can look forward to an easing burden of age-related expenditures (notably Estonia and Poland).

Table I.8. EU8 Medium-Long Term Spending Projections (% of GDP).

Age related expenditure Total non-age related Total revenues Pen si on s H eal th car e Edu cati on Others expenditure 2009 2050 2009 2050 2009 2050 2009 2050 2009 (co nst) 2009 2050 CZ 8.6 15.2 6.5 9.3 3.8 3.6 30.2 47.5 47.5 EE 6.4 3.7 4.6 4.6 26.2 37.5 36.9 HU 7.4 7.6 34.6 42.4 42.4 LV 5.0 5.2 4.1 4.5 5.9 5.8 20.4 34.8 33.3 LT 5.3 7.0 4.6 4.6 25.1 34.8 33.3 PL 7.1 4.5 4.4 3.5 31.8 42.5 42.5 SK 6.9 7.4 5.0 6.6 3.4 3.4 0.8 0.6 20.3 36.9 36.9 SI 12.8 18.2 6.8 9.6 26.1 46.3 46.3

Source: EC (2005).

35 Of course in some EU8 countries debt levels should be brought down to reduce vulnerability and the risk of short-term financing difficulties, while in others (notably the Baltic countries) there is room to expand public debt from a pure fiscal sustainability point of view. 36 The assessment of fiscal sustainability is a much more complex issue that goes beyond the simple analysis presented in this chapter. A thorough overview of theoretical and practical issues regarding sustainability of public finances can be found in Burnside (2005). More recent analysis of fiscal sustainability has focused on the issue of uncertainty: Barnhill and Kopits (2003) adopt a Value-at-Risk methodology to address the effects of macroeconomic volatility and contingent liabilities; IMF (2003) models stochastic processes for key variables and uses Monte Carlo simulation to compute the probability distribution of debt; Xu and Ghezzi (2003) model the probability of default using simulated debt dynamics; and Mendoza and Oviedo (2004) model the effects of stochastic revenue flows on natural borrowing limits faced by a government. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 45

32. Even assuming that EU8 countries achieve their medium-term budgetary targets (baseline scenario) and incorporating the full impact of legislated structural measures, long-run projections reveal a significant risk of unsustainable public finances (measured against the 60% of GDP debt reference value) in almost all of the EU8 countries (except for Estonia, Poland and Slovakia) (Figure I.13). The Czech Republic would breach the debt limit even before 2030.

33. The risk of unsustainable public finances clearly increases considerably if EU8 countries do not achieve the targeted medium-term fiscal consolidation (2004 budget scenario). In this case, only Estonia (and with a stretch, Poland) manage to stay within the critical debt limit. This is a real source of concern given the tendency for fiscal slippages in the past (discussed above).

Figure I.13. EU8 Long-Term Public Debt Projections (% of GDP). Czech Republic Estonia

500 100 Programme scenario ("baseline") 50 400 Re f e r e n c e v alu e 2004 budget scenario 0 300 -50 200 -100 Programme scenario ("baseline") 100 -150 Reference value 2004 budget scenario 0 -200 2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050

Latvia Lithuania

150 150 Programme scenario ("baseline") Programme scenario ("baseline") Reference value Reference value 2004 budget scenario 2004 budget scenario 100 100

50 50

0 0 2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050

Hungary Poland

200 100 Programme scenario ("baseline") Reference value 150 2004 budget scenario 50

100 0

50 -50 Programme scenario ("baseline") Re f e r e n c e v alu e 2004 budget scenario 0 -100 2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050

Slovenia Slovakia

250 200 Programme scenario ("baseline") Programme scenario ("baseline") 200 Reference value Reference value 150 2004 budget scenario 2004 budget scenario 150 100 100

50 50

0 0 2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050

Source: EC (2005). 46

34. Accordingly, most countries will need to undertake significant fiscal adjustment over the medium-long term in order to preserve sustainability of their public finances. “Sustainability gap” calculations suggest an adjustment need of up to 8 percentage points of GDP in the Czech Republic, even in the baseline program scenario illustrated above (Table I.9). The size of the required adjustment in this case reflects age-related expenditure projected to exceed 25% of GDP in 2050, including expenditure on pensions that are projected to reach 15% of GDP. In contrast, in Hungary, where expenditure on pensions is projected to be at comparable levels in 2009 and 2050, the implied adjustment need amounts to only 0.6 percentage points of GDP––much lower than in the Czech Republic despite the fact that Hungary currently has a much larger deficit and higher public debt. Hungary and Poland provide good examples of successful pension reforms that have gone a long way toward securing sustainability of their pension systems, although it would in part be achieved through substantially lower replacement rates from the PAYG systems that may not be wholly offset by pension entitlements from the private pension funds (World Bank, 2005). Significant adjustment would also be needed in Latvia and Lithuania, as well as in other EU8 countries (except Estonia) to the extent they fail to comply with their medium-term fiscal targets.

Table I.9. EU8 Sustainability Gap Indicators (% of GDP).

Programme scenario ("baseline") 2004 budget scenario S1 S2 S1 S2 CZ 4.3 8.0 7.0 10.7 EE -2.5 -1.8 -3.7 -2.9 HU -0.2 0.6 1.2 1.9 LV 0.8 2.1 1.0 2.3 LT 0.4 2.6 1.4 3.6 PL -2.9 -1.8 0.2 1.3 SK -0.1 1.2 1.8 3.2 Notes: S1 measures the difference between the current tax ratio and the tax ratio that – if implemented now and sustained – would ensure a debt level within 60% of GDP in 2050 (a positive sustainability gap indicates an adjustment need). S2 indicates the change needed in tax revenue as a share of GDP that guarantees the respect of the inter-temporal budget constraint of the government, i.e., that equates the flow of revenues and expenses over an infinitive horizon. Source: EC (2005).

35. It should be recalled that the purpose of such long-term debt projections is to signal possible imbalances on the basis of current policies and demographic trends. To the extent policies or demographics change, so will long-term projections. The assessments presented in this chapter are therefore of a highly tentative nature– –providing an overview of possible future developments they do not focus on country-specific issues and limitations. More in-depth country-specific analysis is required to draw firm policy conclusions.

3. Quality of Fiscal Policy 36. The Lisbon European Council of March 2000 called for the emphasis of public finances to be broadened from its focus on stability to include the contribution to sustainable growth, full employment, social cohesion and competitiveness. The quality of public finances in this context refers to the structure of taxation and public spending as well as mechanisms to ensure a high level of efficiency of public spending. This section explores the quality of fiscal adjustments in the EU8 and the impact of tax and spending structures on growth.37 Quality of fiscal adjustments 37. There is a growing consensus among economists and policy makers that the success of efforts to consolidate public finances depends crucially on the “quality” of fiscal adjustments. “High quality” fiscal adjustments are characterized by adjustments that emphasize expenditure cuts rather than increased revenues and on tackling those expenditures that are politically most sensitive such as transfers, subsidies, and wage expenditures (Buti and Sapir, 1998; Alesina and Perotti, 1995 and 1997; Perotti, 1998; and von Hagen, Hughes Hallett and Strauch, 2001). Success is defined in terms of the persistence of the consolidation effort.38

38. Most EU8 countries carried out fiscal adjustments at some point during the last 10 years, with nearly all episodes of fiscal contractions that have been undertaken since 1998 based on expenditure contractions

37 Annex 1 in Volume II summarizes the recent literature on fiscal policy and growth. 38 Annex 2 in Volume II provides an overview of the adjustment experience of EU member states during the Maastricht convergence process and after the start of the third stage of the EMU. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 47

(Table I.10).39 Moreover, nearly one-half (seven out of 15 episodes) of these contractions were based on expenditure restraint where the contribution of spending cuts was at least 70% (expenditure-dominated contractions), although these were hardly present in last four years.

Table I.10. EU8 Patterns of Large Fiscal Contractions 1998-2004.

1998 1999 2000 2001 2002 2003 2004 Czech Republic Estonia Latvia Lithuania Hungary*) Poland Slovenia Slovakia Legend: - fiscal contractions episodes - "expenditure-dominated" contractions episodes (contribution of spending >= 70%)

Notes: *for Hungary in 1999-2000 the effect is estimated due to lack of data. Calculation based on methodology proposed in Von Hagen (2004); fiscal contraction based on expenditure restraint is defined as a change of the expenditure/GDP ratio as a percent of the change in the deficit/GDP ratio exceeding 70%. Source: Eurostat; and staff calculations. 39. Fiscal contractions in the EU8 countries in the period under review relied mainly on cuts in gross fixed capital formation or other capital transfers (Czech Republic, 1999 and 2004; Estonia, 2003; and Latvia, 2003), social benefits other social transfers in kind (Lithuania, 2002; Latvia, 2001 and 2002; and Slovakia, 2003) and collective consumption. The fiscal consolidation in the Baltic countries in 2000 is particularly noteworthy as it was characterized by an average deficit reduction of about 4.4% of GDP, lasted two or more consecutive years, and tackled forcefully government consumption as well as social benefits and social transfers in kind. In contrast, the contraction episode in Hungary in 2004 relied mainly on revenue adjustment, with a contribution of tax increases of 60%.

40. Ireland and Slovakia provide good examples of high quality, sustainable fiscal adjustment (Box I.4 and Box I.5). In both countries, fiscal adjustment was pursued over an extended period of time and relied on scaling back unproductive but politically sensitive expenditures.

Box I.4. Case Study: Ireland In the first half of the 1980s Ireland suffered from very low economic growth accompanied by large fiscal deficits and mounting public debt reaching 112 percent of GDP in 1987.The government elected in February 1987 focused its attention mainly on the expenditures side of the budget: general government expenditures was reduced from 52 percent o GDP in 1987 to 43 percent of GDP in 1990, while at the same time revenues decreased by 3.5 percentage points of GDP. The cyclically adjusted primary balance improved by about 8 percent of GDP during 1985-6 to 1990, and debt was reduced to about 90 percent of GDP by the turn of the decade.Spending cuts were directed mainly to transfers (down by 2.6 percent of GDP), the government wage bill (down by 1.5 percent of GDP) and public investment (down by 1.5 percent of GDP). Public employment was reduced by 10 percent during the period 1987-1989 (mainly in the health sector) through a hiring freeze, early retirement schemes and voluntary redundancy schemes. Also, in 1987 a social partnership was concluded, limiting wage increase to about 1 percent below inflation (ex-post).Tax reforms concentrated on personal income tax cuts (the upper rate was cut by 9 percentage points to 56 percent, while the standard rate was reduced by three percentage points to 32 percent) and lower corporate taxes (CIT rate reduced from 47 to 43 percent).The fiscal adjustment was part of a broader policy package and took place on the backdrop of a favorable external environment. The fiscal adjustment was preceded by a devaluation of Irish currency and pegging to the German mark along with liberalization of capital flows. Enhanced credibility was associated with lower interest rates (real rates dropped by 3-4 percentage points) and a recovery in investment. Also, lower interest rates generated a positive wealth effect, which – along with lower taxes – boosted private consumption. Meanwhile, the external environment was characterized by declining international interest rates (following the 1987 stock market crash) and recovery in the world economy benefiting the export-oriented Irish economy.On this background, Ireland experienced a significant improvement in its economic situation. Already during the years of fiscal contraction (1987-89), growth gained strong momentum (reaching over 6 percent in 1989) and outpaced that of G7 countries by a large margin, especially during 1990-91.

39 An episode of fiscal adjustment is defined as one where the general government deficit (ESA95) decreased by at least by 1% of GDP (this differs from the literature on fiscal adjustment in OECD countries, which looks at primary structural or cyclically adjusted deficits and assumes slightly higher and more persistent contractions; these more rigorous definitions are mainly used for checking “expansionary” episodes, which is not our aim). 48

Box I.5. Case Study: Slovakia* In 1996, a large current account deficit emerged as domestic demand surged, including because of a very expansionary fiscal policy. These imbalances persisted into 1997. In an effort to reverse the deterioration in the external accounts, monetary policy was tightened, but its effectiveness in slowing domestic demand was undermined by a further loosening of fiscal policy. The associated high real interest rate environment created strains in the banking sector and led to increasing financing difficulties for the government. The new government in 1998 responded with a series of measures focused on the expenditure side of the budget (reductions in public investments, social welfare and subsidies). An ambitious fiscal plan introduced in 1999 was based on further expenditure cuts, including the postponement of some major investment projects and a freeze in civil service wages as well as on tax increases (VAT, excises, improving tax collection and administration, and introduction of temporary import surcharge) and significant adjustments in administered prices to bring them more in line with costs. The government also strengthened control over public enterprises and accelerated their privatization. Better control was exercised over activities of extra- budgetary funds and rising arrears in health and transport sectors. However, the new government also inherited a large amount of bad loans and government guarantees for domestic and foreign borrowing by state-owned enterprises, and revealed contingent liabilities drove the fiscal deficit in 2000 above 12 percent of GDP. Nevertheless, by 2002, the government has shrunk by about 10% of GDP compared to 1997. The 2002 government adopted a more ambitious and comprehensive reform program, aimed at both tax and expenditure reforms and improving the overall business climate. On the expenditure side, the key reforms were in the areas of welfare (measures to reduce abuse of social benefits; and a mixture of benefit cuts and increased linkages to job search efforts to improve incentives to work), health care (introduction of co-payments, limiting the basic benefits package, and drug policy), elementary and secondary education (incentives for self-governments to rationalize the number of schools and replace historical-base financing with per-student payments), pensions (increase of retirement ages to 62 years, tightened link between future benefits and contributions, and introduction of a second funded pillar from 2005), and subsidies (deregulating administrated prices thus removing energy subsidies and cutting back support to agriculture, mining and construction). At the same time, fiscal management was strengthened (including establishment of a Debt and Liquidity Management Agency and a Treasury), and many public services (like schools, social care, and roads) were decentralized.40 On the tax side, the reform of 2004 featured a single rate of 19 percent for VAT, personal and corporate income tax, along with a widening of tax bases by eliminating most exemptions. Some taxes, like on dividends, were abolished to prevent double taxation. As a result of these fiscal reforms and a strong recovery in growth from 2002, both fiscal expenditures and revenues were reduced by about 15% of GDP and 13% of GDP, respectively, from 1997 to 2004.41 Social transfers dropped by almost 6% of GDP, and subsidies were lower by 3% of GDP. From 2002, there were also significant reductions in interest payments (1.5% of GDP) thanks to lower interest rates and better debt and liquidity management. The reduction in revenues reflected lower direct taxes (4% of GDP), indirect taxes (3% of GDP), and social security contributions (1% of GDP). Such significant fiscal changes would not have been possible without strong political backing. Although the strategies of the two reformist governments (1998 and 2002) in this respect were slightly different, both profited from the prospects of EU membership, the challenge that united coalition and opposition parties on all major issues. The 1998 government of a wide political spectrum was able to build on the financial crisis in 1998-1999.42 The 2002 centre-right government was more coherent on economic issues, and came to office well prepared with all major reforms drafted and a main message of “do it fast, while support is maintained.” The reforms went much beyond the EU agenda, although the government argued that this was the only way to ensure a secure landing in euro zone.

*) Valuable comments were received from the Institute of Financial Policy of the Slovak Ministry of Finance. 40 Further, in 2003 the Ministry of Finance cut employment by 30%, with savings used to increase remuneration of the remaining employees. 41 Three (inconsistent) time series of general government statistics in ESA95 are available: data from the Statistical Office contains non-consolidated figures for 1993-2002 and 2004, and consolidates figures for 2003; data published by the EC (Spring 2005) are consolidated for 2003 and 2004; and finally the Ministry of Finance, for its own analytical purposes, uses consolidated figures for 2002-2004 (similar to those of the EC). When comparing 1997 and 2004 (or 2002), one has to estimate the size of transfers within the general government in 1997 and adjust the series accordingly. 42 The left was represented by the Party of the Democratic Left, the centre by the Party of Civic Understanding and the right by the Slovak Democratic Coalition and Hungarian Coalition. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 49

Figure A. GG expenditure and GDP growth Figure B. GG Revenue, Expenditure and Balance (% of GDP and %, respectively). (% of GDP). 15 Total revenue 75 GDP GG expenditures Total expenditure 10 65 Deficit 55 5 45 35 0 25

-5 15 5 -10 -5

-15 -15 Y95 Y96 Y97 Y98 Y99 Y00 Y01 Y02 Y03 Y04 Y93Y94Y95Y96Y97Y98Y99Y00Y01Y02Y03Y04

Figure C. Composition of GG Revenues (% of GDP). Figure D. Composition of GG Expenditures (% of GDP). Social contributions 40 40 Interest VAT Subsidies 35 Corporate income tax 35 Social welfare Personal income tax Public sector wages 30 30

25 25

20 20

15 15

10 10

5 5

0 0 Y93 Y94 Y95 Y96 Y97 Y98 Y99 Y00 Y01 Y02 Y03 Y04 Y93 Y94 Y95 Y96 Y97 Y98 Y99 Y00 Y01 Y02 Y03 Y04 Note: in Figure B, the deficit in 1993 reached 31% of GDP. Source: 1993-2001 unconsolidated data of Statistical Office of Slovakia; 2002-2004 consolidated data from Ministry of Finance.

Fiscal policy and growth 41. The literature on the relationship between fiscal policy and growth is inconclusive. Various studies have looked at both the tax and spending side of public finances, but in neither case have there been consistent findings as regards the role of the overall level or composition of tax and spending on growth. There is more consensus that efficiency is key, both in tax policy and collection and in spending across different areas of the budget.43 In the EU8 countries, a simple scatter chart suggests that there might be a negative (albeit weak) relation between the overall tax burden and growth, but this may of course well be spurious correlation (Figure I.14).44 Figure I.14. Tax Ratio and Output Growth (for all EU8, average 1995-2003).

EU8 : Growth versus tax burden 1995-2003 5 4 0 4 5 3 a Tx/GDP 0 3 5 2 -5 0 5 10 15 GDP per capita growth

43 Annex 3 in Volume II provides an overview of expenditure efficiency considerations. 44 Tax shares may not be the best measures of tax distortion, but high tax burdens tend to be correlated with high marginal tax rates (Widmalm 2001). 50

42. Results of an econometric panel analysis covering the period 1996-2004 show that both the tax burden and the structure of taxes also seems to matter for growth (Appendix 2). The total tax burden is negatively related to growth in the sample, although the finding is not robust (Table A.1 and A.2). Further, there is a robust negative relation between the share of direct taxes plus social security contributions (presumably more distortionary taxes) and economic growth, with an increase in the share of these taxes by one percentage point associated with 0.3 percentage point lower growth (Table ). Personal income taxes are found to have a positive, albeit fragile, relation to growth, but this may reflect that the effects on work and investment incentives are not properly captured by the ratio of personal income taxes to total tax revenue, or that higher labor taxes have not been shifted onto higher real wages. On the other hand, both corporate income taxes and social security contributions have a negative and robust impact on growth. Finally, indirect (presumably less distortionary) taxes appear to have a positive and robust correlation with economic growth.

43. At the same time, none of the variables reflecting expenditure structure were robustly correlated with growth (Table). Nevertheless, two variables – gross fixed capital formation and social benefits other than social transfers in kind – were robust in some combination of conditional variables. In these cases, gross fixed capital formation had a positive impact on growth while social benefits other than social transfers in kind were associated with lower growth.

44. These preliminary results warrant cautious interpretation because of the limitation of the methodology used and quality/availability of data. There is as yet no comprehensive analysis of the relationship between tax/ expenditure structure and growth in the NMS and our analysis is only a first, small step in that direction. Synthetic indicator of the composition of public spending 45. The European Commission has developed a synthetic indicator to evaluate the composition of public spending across member countries (EC, 2003). The methodology developed by the EC aims to categorize expenditure items in terms of their efficiency effects and impact on fostering long-term growth (Figure I.15). Each spending component has a range of 2 points between the best and the worst performer. All items enter the indicator with the same weight. The higher the sum of scores in all categories, the better is the composition of public spending relative to other Member States.

Figure I.15. Theoretical Efficiency of Various Categories of Government Expenditure.

Efficiency

Category 2 Category 3 Category 4

Expenditure/ GDP

Category 1

Source: EC (2003).

Source: Benedek, D., O. Lelkes, A. Scharle, and M. Szabó (2004). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 51

46. The EC has only calculated the indicator for EU15 countries, but we have supplemented this with our own calculations for the EU8 (Table I.11). Slovenia, Poland and – to a lesser extent – Slovakia and Hungary compare favorably to other EU member countries, reflecting good scores in education, health, and social protection (Slovenia), compensation of employees, social protection, and housing and community amenities (Poland), health, compensation of employees, and housing and community amenities (Slovakia), and public investment and social protection (Hungary). On the other hand, Latvia and the Czech Republic come out near the bottom of the table, with low scores in health, social protection and public investment (Latvia) and education, compensation of employees, and social protection (Czech Republic). These rankings do not appear to be correlated with the level of public spending, supporting the lack of evidence between spending and growth found above.

Table I.11. Composition of Public Spending in 2002 – Ranking.

Ranking Total expenditure (% of GDP) 1LU43.7 2SI48.1 3UK41.6 4PT*46.0 5FR*52.7 6PL*48.9 7FI49.8 8DE47.8 9 SK* 39.2 10 HU 52.6 11 LT 34.5 12 IT* 48.4 13 DK 56.2 14 AT 50.7 15 NL 47.8 16 BE 50.2 17 CZ 46.9 18 LV 35.8 19 GR* 50.0

Notes: The ranking does not include Cyprus, Estonia, Ireland, Malta, Spain and Sweden due to incomplete data. Calculations are based on GFS 2001 methodology (GG outlays by function and expenditure by type: education, health, social protection, housing and community amenities, compensation of employees, and interest) and ESA 95 (gross fixed capital formation and collective consumption). *FR - health and social protection figures for 2001; GR - education, health, social protection and housing and community amenities for 1999; compensation of employees and interest for 2000; IT - GFS data for 2000; PT - GFS data for 2001; SK - GFS data for 2003. Source: IMF GFS 2001-03; Eurostat; Polish Ministry of Finance; and staff calculations.

47. Clearly, the synthetic indicator has been designed only to benchmark good practices in public spending composition in member countries. Thus, it can only provide a highly suggestive signal regarding the quality of public spending in one member country relative to its peers. The indicator only provides ranking in relative terms and does not imply that higher-ranked countries are actually setting standards of good practice. Moreover, measuring only the level of spending on certain expenditure categories leaves out the issue of the output/outcome of these outlays and the indicator thus does not reflect the efficiency with which resources are used (for more detailed discussion on the issue of efficiency see Annex I.3 in Volume II). Data comparability and uncertain weighting of different expenditure component ratings are obviously also major shortcomings in this approach. Further, the indicator assesses public spending through the lens of its contribution to growth and as such does not capture other fiscal policy objectives such as stabilization and redistribution. The structure of public spending may result from social choices and more in-depth analysis is thus required to draw any policy conclusions. 52 APPENDIX 1 Decomposition of Changes in the Debt/GDP ratio 48. The “snowball” effect. Debt dynamics can be expressed as the change in the gross debt ratio resulting from the primary deficit, the “snowball” effect (contribution of interest and nominal growth) and the stock- flow adjustment:

D t - D t 1 PD t D t 1 i t y t SF t Y t Y t 1 = Y t + { Y t 1 • 1 + y t } + Y t Snowball effect where: Dt –general government gross debt, Yt –GDP at current market prices, yt –nominal GDP growth rate, SFt –stock-flow adjustment PDt – primary deficit, it- implicit interest rate, calculated as interest paid as % of gross debt at the end of period t-1.

49. The “hidden” fiscal deficit. Measuring “hidden” deficits entail first decomposing debt developments into the effects of initial debt, output growth, fiscal balance, and a residual (stock-flow adjustment - SF) which would include results from financial operations not related to deficit financing, privatization receipts, impact of exchange rate changes on foreign denominate debt, etc.

The change in debt-to-GDP ratio can be written as: g d d = b d + sf t + 1 t ( 1 +  )( 1 + g ) t , where

αε(1+ i) sf = d + p + hd (1+ π )(1+ g) t

d, b sf, hd, p – general government gross debt, GG deficit, stock-flow adjustments, hidden deficit, privatization receipts as % of GDP;

π1t –GDP deflator; gt –real GDP growth; i – nominal interest rate on debt; a - share of foreign currency debt in total debt; ε - depreciation of the domestic currency (indicated by increase in exchange rate as units of domestic currency per unit of foreign currency); and t- time dimension (variables without time dimension are t+1).

50. In general, these would be expected to more or less cancel out over time. However, large and persistent stock-flow adjustments (especially if they have a negative impact on debt developments) should give cause for concern, as they “may be the result of inappropriate recording of budgetary operations and can lead to large ex-post upward revisions of deficit levels” (EC, 2003). This may reflect “quasi-fiscal” activities or “bad subsidies that serve special interest groups,” financed through accumulation of building contingent liabilities (Kharas and Mishra, 1999). Accordingly, the SF can be split into three components: privatization flows, exchange rate effects and remaining statistical adjustments (so called “hidden deficits”). However, interpreting “hidden deficit” involves a careful scrutiny of its size and the components. Remaining statistical adjustments usually include45: differences between the accrual and cash bases of recording transactions (example PL - 0.5% of GDP over the period 2000 to 2004); differences in the gross and net recording of transactions with financial assets; (i.e. accumulation of currency and deposits with banks, securities other than shares bonds issued by non-government units), loans and shares; and other adjustments reclassified of some units from non-government sectors to government and vice- versa, or early debt reimbursements.

45 Based on EC (2005). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 53 APPENDIX 2

51. The analysis departs from the recent empirical growth literature (e.g. Kormendi and Meguire 1985; and Barro 1991). Previous studies have revealed that many of the structural and policy indicators found to affect growth are fragile and sensitive to small alterations in the conditioning information set (Levine and Renelt 1992; Folster and Henrekson 2001). We thus use a variant of the extreme-bounds analysis (Box I.6.).

Box I.6. EBA Methodology46 Extreme-bounds analysis involves the following steps (see e.g. Leamer 1983 and 1985): Imagine that there is a pool of variables (I and Z) that previously have been identified to be related to growth and one is interested to know whether the inclusion of a particular variable M is robust. Thus, one would estimate an equation of the form: Y = α + βmM+ βiI + βzZ + βdD + ε, where: Y is per capita GDP growth; I is a set of variables always included in the regression (e.g. the initial level of income, the investment rate, the secondary school enrolment rate, and the rate of population growth – following Levine and Renelt 1992); Z is subset of variables identified by past studies as potentially important explanatory variables of growth; usually up to three variables are taken from a pool of N variables available; and M is the variable of interest. Extreme-bounds testing involves varying the subset of Z-variables included in the regression to find the highest and lowest values for the coefficient on the variable of interest, (βm), that standard hypothesis test do not reject (at 0.05 significance level). Thus, the extreme upper bound is defined by the group of Z-variables that produce the maximum value of βm plus two standard deviations (βm + 2σm). A result is “robust” if βm remains significant and of the same sign at the extreme bounds. In contrast, if one finds a single regression for which the sign of the coefficient changes or becomes insignificant, the result is “fragile”. Thus, alteration in the conditioning information set may change the statistical inferences regarding the relationship between Y and M.

52. We considered the following basic I-variables in line with recent empirical studies: (i) the initial level of real GDP per capita; (ii) the investment-GDP ratio; (iii), the rate of population growth; (iv) the level of human capital (proxied by secondary school enrollment in line with common practice); and (v) the tax-to-GDP ratio (or in the expenditure analysis, the expenditure-to-GDP ratio). The latter was included in the base set to control for the overall tax (or spending) burden across countries and time while examining the role of the tax (or expenditure) structure (Widmalm, 2001). The Z variables included were the inflation rate (Stockman, 1981), the export-GDP ratio (measure of country openness), and the share of government consumption in GDP as indicator of the expenditure mix (or in the expenditure analysis, the tax-to-GDP ratio). The M variables of interest were in the case of the tax analysis: M – direct taxes and social security contributions; M1 – indirect taxes; M3 – social security contributions; M4 – personal income taxes; and M5 – corporate income taxes (all as a share of total taxes). In the expenditure analysis, the M variables were: M1 – subsidies; M2 – interest; M3 – social benefits other than social transfers in kind; M4 – collective consumption; M5 – social transfers in kind; and M6 – gross fixed capital formation (all as a share of total spending).

53. The empirical results are broadly in line with expectations (Table I.12.). In the first step, growth was regressed on the initial level of GDP, the investment share and the tax-to-GDP ratio (Model I).47 All coefficients have the expected signs and are significant. In the second step, population growth and secondary school enrollment were added as explanatory variables (Models II and III). The inclusion of these variables did not change the signs of the first three variables. Population growth has the “wrong” sign, but is not significant. School enrollment has the expected, positive and significant coefficient. However, the inclusion of this variable reduces the significance of the tax-to-GDP ratio. In the final step, the model is extended to include the Z variables used for the EBA analysis (Model IV). These variables are all of the expected sign and significant.

46 The EBA method has been criticized for “reverse data-mining” (Sala-i-Martin, 1997) and multi-collinearity (Levine and Renelt, 1992). 47 Results were similar for the expenditure analysis and are not shown here. The spending-growth elasticity was estimated at -0.36. 54

Table I.12. EU8 Panel Growth Regressions 1995-2004 (dependent variable: growth rate of per capita GDP).

Model IIIIIIIV Initial GDP per capita (Y0) -0.00*** -0.00*** -0.00*** -0.00*** (-2.9) (-2.5) (-3.8) (-5.9) Investment share of GDP (INV) 0.20*** 0.18*** 0.22*** 0.17** ( 3.0) (2.8) ( 2.9) (2.2) Tax-to-GDP ratio (TAX) -0.15** -0.13* -0.12* -0.15** (-1.8) (-1.6) (-1.4) (-1.7) Population growth (POP) -0.62 -- (-0.63) Secondary school enrolment ratio (HUM) 0.13*** 0.19*** (2.8) (4.0) Government consumption share of GDP -0.46*** (GOV) (-2.8) Openness (OPEN) 0.05** (1.8) Average inflation rate (INFL) -0.03*** (-2.1) Numbers of observations 80 80 61 59 R-sq (overall): 0.27 0.29 0.36 0.49

Note: Random effects estimation. ***(**)[*] stands for 1% (5%)[10%] significant

54. The initial level of GDP, the investment ratio, and the human capital variable are all robustly correlated with output growth (Table I.13.). Meanwhile, population growth and tax ratios are fragile, although the latter retains a negative sign.

Table I.13. Robustness of Base Variables.

M-variable Bound Bm t-ratio N R2 Z Robust Investment share min 0.17 2.18 59 0.48 OPEN,GOV,INF base 0.22 2.87 61 0.46 - yes max 0.28 3.39 59 0.39 OPEN, GOV Initial GDP min -0.00 -4.72 59 0.39 OPEN,GOV base -0.00 -3.78 61 0.36 - yes max -0.00 -3.51 61 0.38 INF Population min -1.83 -1.80 59 0.52 OPEN,GOV,INF growth base 0.50 0.42 61 0.36 - no max 1.49 0.92 59 0.33 OPEN Tax-to-GDP min -0.19 -1.58 59 0.36 OPEN,INF base -0.12 -1.35 61 0.36 - no max -0.07 -0.67 61 0.39 GOV Human capital min 0.13 2.80 61 0.40 GOV base 0.13 2.84 61 0.36 - yes max 0.19 3.95 59 0.48 OPEN,GOV,INF

Note: “base” corresponds to Model III above

55. The structure of taxes also seems to matter for growth. We find a robust negative relation between the share of direct taxes plus social security contributions (presumably more distortionary taxes) and economic growth, with an increase in the share of these taxes by one percentage point associated with 0.3 percentage point lower growth (Table I.14.).48 Personal income taxes are found to have a positive, albeit fragile, relation to growth, but this may reflect that the effects on work and investment incentives are not properly captured by the ratio of personal income taxes to total tax revenue, or that higher labor taxes have not been shifted onto higher real wages. On the other hand, both corporate income taxes and social security contributions have a negative and robust impact on growth. Finally, indirect (presumably less distortionary) taxes appear to have a positive and robust correlation with economic growth.

48 These results are based on Model I, but do not change fundamentally in the alternative specifications. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 55

Table I.14. Tax Structure and Output Growth (shares of total taxes).*

M-variable Bound Bm t-ratio N R2 Z Robust Direct taxes and social min -0.31 -3.84 78 0.40 OPEN security contributions base -0.30 -3.38 80 0.40 - yes (M) max -0.26 -2.59 80 0.42 GOV Indirect taxes (M1) Min 0.26 2.60 80 0.41 GOV base 0.30 3.39 80 0.40 - yes max 0.31 3.85 78 0.40 OPEN Direct taxes (M2) min 0.08 0.80 78 0.30 OPEN, INF base 0.09 0.99 80 0.26 - no max 0.16 1.79 78 0.27 OPEN, GOV Social security min -0.36 -4.26 78 0.46 OPEN, GOV contributions (M3) base -0.30 -3.61 80 0.40 - yes max -0.28 -3.76 80 0.40 INF

Personal income taxes min 0.08 0.94 78 0.26 OPEN, INF (M4) base 0.11 1.45 80 0.31 - no max 0.15 2.49 80 0.38 GOV, INF Corporate income min -0.22 -2.44 78 0.33 OPEN taxes (M5) base -0.22 -2.58 80 0.34 - yes max -0.19 -2.13 80 0.36 GOV, INF

*based on Model I 56. None of the variables reflecting expenditure structure were robustly correlated with growth (Table I.15.). Nevertheless, two variables – gross fixed capital formation (M6) and social benefits other than social transfers in kind (M3) – were robust in some combination of conditional variables. In these cases, gross fixed capital formation had a positive impact on growth while social benefits other than social transfers in kind were associated with lower growth.

57. Our findings should be interpreted with cautious given a number of problems that are commonly encountered in this type of cross-section regressions. The most important of these may be a potentially severe simultaneity problem, including potentially arising from business cycle effects and Wagner’s law (the tendency for government expenditure to be higher at higher levels of per capita GDP). As a result, the independent variable (like taxes or government spending as a share of GDP) is correlated with the error term in the growth regression, and this bias could impact regression results. While the possible simultaneity effect is an argument in favor of panel regressions with shorter time spans (as used by us), it at the same time limits the room for including potentially important lags of the explanatory variables. Further, there may be concerns about data quality, and measurement or misspecification errors may give rise to heteroskedasticity and misleading results.49 However, heteroscedasticity most often appears in a form where the error term is correlated with one of the independent variables or with the dependent variable, and this does not seem to be a major problem in our data.50

49 The linear ordinary least squares regression (OLS) makes the assumption that the variance of the error term in regression is constant (homoscedasticity). 50 Visual inspection of residuals plotted against fitted values show the distribution of the residuals in our pooled regression did not seem overly heteroscedastic. The more formal tests for heteroscedasticity (Breusch-Pagan / Cook-Weisberg test and White’s general test) also confirm that heteroskedasticity was not a problem in our pool regression. Not surprisingly, therefore, common corrections for heteroscedasticity such as the robust/White standard errors hardly change the results. 56

Table I.15. Expenditure Structure and Output Growth.

M-variable Bound βm t-ratio N R2 Z Robust Subsidies (M1) min -0.11 -0.25 52 0.39 INF base -0.01 -0.21 52 0.38 - no max 0.66 1.67 52 0.65 TAX, OPEN Interest (M2) min -0.36 -1.65 52 0.67 TAX,OPEN,INF base 0.01 -0.06 52 0.38 - no max 0.02 -0.06 52 0.39 INF Social benefits other min -0.54 -2.56 52 0.48 INF than transfers in kind base -0.48 -2.60 52 0.47 - no (M3) max -0.27 -1.45 52 0.67 TAX, OPEN INF Collective min -0.15 -1.24 52 0.61 OPEN, INF consumption (M4) base -0.09 -0.63 52 0.39 - no max 0.12 0.77 52 0.50 TAX, INF Social transfers in kind min 0.14 0.38 52 0.65 TAX, OPEN, INF (M5) base 0.29 0.62 52 0.39 - no max 0.29 0.62 52 0.39 - Gross fixed capital min 0.36 1.26 52 0.41 INF formation (M6) base 0.38 1.41 52 0.41 - no max 0.64 2.96 52 0.65 OPEN Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 57 References Barnhill, T.M., Kopits Jr., G. (2003). ”Assessing Fiscal Sustainability under Uncertainty.” Manuscript, George Washington University. Barro, R.J. (1991). “Economic Growth in a Cross-Section of Countries.” Quarterly Journal of Economics, 106(2): 407-443. Benedek, D., Lelkes, O., Scharle, A., Szabó, M. (2004). “The Structure of General Government Expenditure and Revenues in Hungary, 1991-2002.” Ministry of Finance Working Paper No.9. Burnside, C. (2004). “Some Tools for Fiscal Sustainability Analysis.” Duke University, November 2004 Burnside, C. [ed.] (2005). “Fiscal Sustainability in Theory and Practice.” The World Bank, August. Buti, M., Sapir, A. (1998). “Economic Policy in EMU: a Study by the European Commission Services.” Clarendon Press, Oxford, Oxford University Press, New York. European Commission (2003). Public finances in EMU - 2003. European Economy, 3/2003. European Commission (2005). Public finances in EMU - 2005. European Economy, 3/2005. Fölster, S., Henrekson, M. (2001). “Growth Effects of Government Expenditure and Taxation in Rich Countries.” European Economic Review, 45(8): 1501-1520. IMF (2003), “Sustainability Assessments-Review of Application and Methodological Refinements.” Manuscript, Policy Development and Review Department, International Monetary Fund, June. Kharas, H., Mishra, D. (1999), “Hidden Deficits and Currency Crises.” Mimeo. World Bank. Kormendi, R.C, Meguire, P.G., (1985). “Macroeconomic Determinants of Growth: Cross-country Evidence.” Journal of Monetary Economics, 16: 141-164. Leamer, E.E. (1983). “Let’s Take the Con out of Econometrics.” American Economic Review, 73(1), 31-43. Leamer, E.E. (1985). “Sensitivity Analyses Would Help.” American Economic Review, 75(3), 308-13. Levine, R., Renelt, D. (1992). “A Sensitivity Analysis of Cross-country Growth Regressions.” The American Economic Review, 82(4): 942-963. Mendoza, E.G., Oviedo, P.M. (2004). “Fiscal Solvency and Macroeconomic Uncertainty in Emerging Markets: The Tale of the Tormented Insurer.” Manuscript, University of Maryland, April. OECD (2004). “Enhancing the effectiveness of public spending: Experience in OECD countries”, WP No 380 Polackova Brixi H., Valenduc Ch., Zhicheng Li Swift (2004). “Tax Expenditures – Shedding Light on Government Spending through the Tax System. Lessons from Developed and Transition Economies”. World Bank. Sala-i-Martin, X. (1997). “I Just Ran Four Million Regressions.” NBER Working Paper No. 6252, National Bureau of Economic Research.

Ter-Minassian, T. (2005). “Fiscal Issues in Accesion Countries.” Mimeo. International Monetary Fund, Washington, D.C. Von Hagen, J., Hughes Hallet, A., Strauch, R (2001). „Budgetary Consolidation in EMU.” European Commission, Economic Papers No. 148. World Bank (2004). “EU-8 Quarterly Economic Report. Special Topic: Corporate Income Taxation and FDI in the EU-8.” The World Bank, Warsaw, October. World Bank (2005). “EU-8 Quarterly Economic Report. Special Topic: Sustainability of Pension Systems in the EU- 8.” The World Bank, Warsaw, October. Widmalm, F. (2001). “Tax Structure and Growth: Are Some Taxes Better Than Others?” Public Choice, 107(3-4): 199-219. Xu, D., Ghezzi, P. (2003). “From Fundamentals to Spreads: A Fair Spread Model for High Yield Emerging Markets Sovereigns.” Manuscript, Deutsche Bank, July. WB (forthcoming) Bulgaria Public Finance Policy Review, 58 II. INTERGOVERNMENTAL FISCAL RELATIONS51

1. Introduction 58. Subnational governments are an important part of the public sector in the EU8 countries. They provide basic public services in both the social sectors (education, health, and social assistance) and in infrastructure (water supply, sewerage and transport). They account for about one-quarter of government spending. How well they perform these functions has an important influence on the quality of both human and physical capital in these countries, and therefore on the prospects for economic growth. It also has equity implications, as the social services performed by municipal governments have important distributional effects. Further, subnational government operations have implications for macroeconomic stability, as independent fiscal actions by local governments could undermine central control over fiscal policy.

59. The period since the fall of the Berlin Wall and the breakup of the Soviet Union has seen dramatic changes in the political and economic institutions of the EU8 countries. All abolished the Communist party’s monopoly on political power, abandoned central planning as a device for allocating resources and setting targets for deliverables, and were prompted, in reaction to the high degree of centralization under the former regimes, to grant substantial power to subnational units of administration. The EU8 countries have spent the last 15 years devising a new system of intergovernmental arrangements that will respond to these changes. The aim of this chapter is to see how far along they have come and how well they now perform.

2. Evaluation Criteria 60. According to conventional economic theory, the primary aim of a system of local government finance is to promote efficiency in the allocation of resources (Box II.1). Theory argues that if the benefits of particular services are largely confined to local jurisdictions, welfare gains can be achieved by permitting the level and mix of such services to vary according to local preferences. Local consumers, confronted with the costs of alternative levels of service can express their preferences by voting for rival political candidates (i.e., voting with their hands) or moving to other jurisdictions (voting with their feet). In this respect, local governments can approximate the efficiencies of a market by “pricing” municipal services and relying on the local political process and household mobility to clear the market. On this basis, a successful system of intergovernmental relations would be one in which each level of government (1) performs only functions whose benefits are confined to its boundaries and (2) derives its revenues from sources (such as local taxes and fees) that confront its residents with the costs of these functions.

61. In practice, there are a number of limitations to this model. First, the benefits of public services can be difficult to define in geographical terms. While most of the benefits of spending on public health, for example, may be confined to a local jurisdiction, failures to control infectious diseases can have national or international implications.

62. Other aspects of the theory confront practical obstacles as well. As described in Box II.1, theory assigns the role of income redistribution to the central government. It is presumed that central governments will use the instruments under their control (e.g., progressive taxation and direct payments to low income households) to achieve a desired level of income equality, leaving local governments to focus only on the level and mix of services they provide. But spending on public services – particularly education and health – can be important instruments of distributional policy. Where these functions are assigned to local governments, their distributional implications cannot be ignored. Theory also assumes that macroeconomic stability is not a local government concern. But high levels of local taxation or widespread deficit spending could weaken the central government’s control over fiscal policy.

63. As a practical matter, the intergovernmental fiscal systems of the EU8 countries must be held to somewhat different standard: Transparency and predictability: The principal challenge confronting the EU8 countries was to devise a system for financing local government that would be perceived as fair and stable. The emergence of multi-party democracy in the EU8 required transparency in order to allay charges of partisan favoritism

51 Prepared by William Dillinger. Signe Zeikate provided research assistance and prepared the section on debt controls. Ken Davey and Deborah Wetzel were peer reviewers. An earlier version of this chapter was discussed in a regional workshop in Budapest in June 2005. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 59

and reduce the transaction costs that would result from continued reliance on bargaining. Predictability was required to provide local governments with a stable basis for budget planning. Equity: Because local governments assumed responsibility for social services, the new system had ensure that such services would be universally available, regardless of the strength of the local tax base. Macroeconomic control: The system had to ensure that local government fiscal behavior would not threaten macroeconomic stability. Central governments needed sufficient control over aggregate local taxing to conduct fiscal policy and sufficient control over local borrowing to ensure that aggregate deficit targets were achievable.

Box II.1. A Snapshot of the Economics of Expenditure Assignment Which functions should be assigned to which level of government? The literature on fiscal federalism provides a useful framework for thinking about this question. It begins by assigning three roles to the public sector as a whole: macroeconomic stabilization, income redistribution, and resource allocation (in the case of market failure). The model assigns the first two roles to the central government. The central government is assigned responsibility for stabilization on the grounds that local economies have no access to an independent monetary policy and are too open for effective countercyclical measures to be effective. The income redistribution function is also assigned to the central government, on the grounds that local attempts to address income disparities are likely to provoke inefficient migration: higher-income groups will move to low-tax areas and low-income groups to high-benefit areas (even with their populations in situ, local governments in poor regions would have little income to redistribute anyway). Subnational governments enter the picture only with respect to the third function: resource allocation. If the benefits of particular services are largely confined to local jurisdictions, residents can choose the level and mix of services that best matches their preferences (and pocketbooks). In this way, a system of intergovernmental finance can approximate the allocative efficiencies of a market in the allocation of local public goods.

Effectiveness: Finally, the systems needed to encourage effectiveness in the production of services under local control. The EU8 countries inherited over-scaled schools and hospitals and inefficiently run public utilities. Incentives for rationing services therefore had to be incorporated in the new system of intergovernmental finance.

3. The System Ex Ante 64. The current structure of intergovernmental relations in the EU8 countries represents a substantial departure from the system that preceded it. The former system had two key characteristics: a prominent role for state enterprises in the financing and delivery of public services and a system for allocating resources based on the Communist Party’s monopoly of political power.

65. The hallmark of the socialist-era economies was the state’s ownership of the means of production. While private ownership existed in some sectors in some countries (for example agriculture in Poland and small businesses in Hungary) it was the state that dominated both the economic and social sectors. State enterprises provided services directly to their employees. They built and supported hospitals, constructed and maintained housing, built and ran kindergartens and preschools and made donations toward public transportation and subnational extra-budgetary funds. But the majority of social and infrastructure services – education, health, and public utilities – were provided through separate sectoral ministries. Sectoral ministries’ policies were implemented through a hierarchy of subnational units of administration. As these units reported to a variety of sectoral ministries, they functioned not only as deconcentrated units of the individual ministries but, in some respects, as local governments.

66. The territorial units of administration derived revenue from four major revenue sources. The first was profits on enterprises. If an enterprise were directly subordinate to a municipality, that municipality would derive revenue directly from its profits. If the enterprise were subordinate to a higher tier of government, the municipality would receive revenue from profit tax levied on it (Berkowitz and Mitchneck, 1992). In addition, the local government would receive revenues from shares of the turnover tax (on consumer goods, foods and certain extractive and light industries) and from the personal income tax. As a result, a municipality’s revenues were partly dictated by the number and scale of enterprises within its territory, and by their ownership. Local budgets were also subject to a complex process of negotiation. Budgets could be supplemented from higher levels of government and by contributions from large enterprises within their 60 boundaries. Thus, the level of resources available depended upon the ability of the local party secretary to remain on good terms with the party bosses at higher levels and with the managers of state farms or factories (Kocan and Regulski, 1994).

67. Politically, the system had the trappings of Western democracies. At the republic level, there was a legislative body (congress of people’s deputies) elected by universal adult suffrage which (since it met infrequently) had a standing legislature. The chairman of the standing legislature functioned as head of state and oversaw the council of ministers, which acted as the executive branch of the government. A parallel structure existed at each subnational level. Each unit had its own council elected by universal suffrage. Like their counterparts at the national level, these met infrequently. Between sessions, each council delegated its authority to an executive committee whose chairman acted as chief executive and oversaw the functioning of the various departments of administration.

68. Effective political power was, however, exercised by Communist Party. The Party had its branches at the republic and subnational levels . The leadership of a local branch of the Party was, in theory, chosen by local party members. In practice, candidates were chosen by the local branch of the party, so that ultimately the existing party leadership designated the people who would be put on the ballot to elect it.

69. In the classic characterization, it was the Party that set policy and the government that implemented it. In practice, the relationship between the Party and the government was more intimate. The Party determined which candidates would be on ballot for local councils and appointed the key officials of the administration. And it was the Party secretary, not the exective the local government who had to lobby central planners for low plan targets and then find the resources to meet them (Zikel op. cit.). The result was a system that depended on bargaining and negotiation, most of which occurred within the Party structure.

4. Political and Organizational Reforms 70. All EU8 countries began to dismantle the Soviet era structure in the early 1990s. In political terms, this took the form of an end to the Communist Party’s monopoly on political power. As shown in Table II.1, national level multi-party Parliamentary elections Table II.1. Timetable of Post-Soviet Reforms. in Poland in 1989 were followed by similar elections Year of First Post Percent of GDP in Czechoslovakia, Hungary, and Lithuania (1990), Soviet-era Elections Privatized national Local year of percent Estonia, and Slovenia (1992) and Latvia (1993). parliame data -ntary 71. Changes in the political and organizational Czech Rep. 1990 1990 1996 75 structure of local government quickly followed. In Estonia 1992 1993 1997 67 Poland, the first Solidarity-led government introduced Hungary 1990 1990 1997 86 a Law on Local Self Government establishing the Latvia 1993 1989 1997 60 Lithuania 1990 1995 1997 70 existing gminy as the basic administrative unit of local Poland 1989 1990 1996 65 government (Barbone and Hicks, 1995). Multiparty Slovakia 1990 1990 1997 83 elections at the gmina level were held in May of 1990. Slovenia 1992 1994 1998 55 Subsequent legislation – an act on local revenues and Source: Freedom House (1998 and 2004): Nations in Transit. rules of financing (1990), a budget law (1991) and an act on taxes and local fees (1991) – designated the revenue bases of the gminy and separated their budgets from that of the central government.

72. Most of the other EU8 countries followed suit. In the former Czechoslovakia, the new Constitution and subsequent enabling legislation recognized the existing municipalities as self-governing units, guaranteed their autonomy and provided for local assemblies chosen by universal suffrage.52 The first post-Soviet elections in the former Czechoslovakia were held in 1990. In Estonia, similarly, the existing units of local government were recognized as legal entities, with independent budgets and councils elected by universal suffrage (Maeltsemees, 2000). Elections were held in 1993. In Latvia, the 1994 Law on Municipalities provided much of the legal structure and political arrangements. Local elections were held in 1989.

52 Czechoslovakia split into two separate republics in 1993. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 61

73. Organizational reforms in Hungary were more radical. The 1990 Law on Local Self- Government abolished the existing local governments and dramatically scaled back the responsibilities of the deconcentrated units of central administration. To replace the former local governments, citizens were granted the right to create autonomous units of self government. As a result, the former local governments broke up into discrete units, roughly doubling in number (Ebel et. al., 2000). The Law on Local Self- Government and related legislation established local governments as legal entities and established a system of local elections with directly elected councilors and mayors (mayors were indirectly elected in towns above 10,000 inhabitants in the period 1990-94 but are now directly elected).

74. Lithuania took an opposite track. In 1990, the Supreme Council of the Lithuanian Soviet Socialist Republic adopted the Law on the Foundation of Local Self-Government guaranteeing the autonomy of local units of self-government and providing for the election of local councilors on the basis of universal suffrage. But subsequent legislation substantially reduced the number of individual local units. The 1994 Law on Territorial Administrative Units and Their Boundaries replaced the former system of 581 rayons, cities, and villages with a system of ten counties (which function as deconcentrated units of central administration) and sixty municipalities (Beksta and Petkevicius, 2000).

75. Democratization did not stop at the lowest tier of subnational administration. In a second phase of reform in the late 1990s, Poland added an additional tier of elected subnational government by resurrecting a former intermediate tier of government, the powiats. The powiats, with directly elected councils which appoint an executive from among their members, have assumed most of the tasks that formerly belonged to the voivodships. Voivodships, nevertheless, remain fully self-governing authorities, with elected councils and chief executives. In 2001, the Czech Republic dissolved the former 77 deconcentrated units of central administration and transferred the majority of their functions to 14 newly-created regional governments, each with an assembly elected by universal suffrage and president elected from among its members. Slovakia has followed suit. In a series of reforms from 2002 to 2004, the 79 deconcentrated units of central administration were abolished and replaced by eight elected regional governments. These will, however, continue to operate alongside eight existing regional units of central administration.

76. In economic terms, the dismantling of the socialist-era structure took the form of privatization. This took different forms in different countries. In the Czech Republic, it began with the restitution of Communist-confiscated property – houses, farms, and shops – to their former owners or their descendents. Small enterprises were then sold through open auctions (as opposed to auctions restricted to workers and managers). Privatization of large scale enterprises began with the conversion of 100 large trusts into 330 independent enterprises. These were then privatized through a voucher program.53 By 1996, 75 percent of the Czech Republic’s GDP was in private hands (Table II.1). Poland’s initial privatization strategy envisioned sales to foreign investors, IPOs on the stock exchange, and insider (management and/or worker) takeovers. But all privatizations had to be approved by the firms’ managers and employees. The privatization of larger firms was therefore slow and controversial. By 1996, only 1,895 firms (out of a 1989 total of 8,853) had been privatized through these methods. Small privatizations – the sale of state-owned stores/shops and small firms to individuals or groups of private investors – proceeded relatively quickly. By the mid-late 1990s, all the economies of the Baltic and Visegrad countries were at least two-thirds privatized. Only Slovenia lagged (Freedom House, 1998).54

77. Privatization had two important implications for local governments. First, it eliminated an important provider of public services. Second, it drastically altered the revenue base. Profits from local-government- owned enterprises were no longer a source of revenue, and as central governments revised the tax structure to capture revenue from the growing private sector, the fundamental basis for tax sharing had to be altered.

53 There were two major waves of voucher privatization, the first in 1992-3 and the second in 1993-4. In each wave, each citizen received 1000 points, which could be exchanged for shares in privatized companies. Citizens were free to sell the shares they had been allocated through a special stock exchange created for this purpose. 54 Note that figures may underestimate the size of the private sector by undercounting informal economic activities. 62 5. Assignment of Functions 78. The new local government legislations assigned an extensive list of functions to the newly independent municipalities. In Poland, the gminy were made responsible for nurseries, kindergartens, pre-schools, primary and lower secondary education, social assistance to the elderly and handicapped, the homeless, and families in crisis, social housing, primary and preventative healthcare, theaters, museums, libraries, parks, public utilities (water, sewer, electricity, gas, central heating, telephones), waste collection and disposal, street cleaning, cemeteries, environment protection, local roads, public transportation, etc.53 In Hungary, local government tasks included the provision of kindergartens, primary education, health and social provision, looking after children and youth; support of scientific and artistic activities and sports, ensuring the enforcement of the rights of national and ethnic minorities; housing management, water resources planning and drainage (of rain water), canalization and sewerage, maintenance of public cemeteries, maintenance of local public roads and public areas, local mass transit, public sanitation and ensuring the cleanliness of the settlement, and providing for local fire protection and public security. In Latvia, under the Law on Municipalities of 1994 as amended, municipalities must provide residents with access to elementary (basic) education and general secondary education, kindergartens and schools for children at pre-school and school age; healthcare, social assistance to poor families and socially vulnerable persons, along with housing for orphans, the elderly and the homeless, public utilities (water and sewage, district heating, collection and transportation of household waste, collection and treatment of sewage), construction and maintenance of streets, roads and squares, street lighting and the collection and transportation of industrial wastes. They must also promote entrepreneurship within their jurisdictions, engage in unemployment reducing measures, ensure public order and combat alcoholism. This list is not exhaustive.

Figure II.1. Structure of Local Government Expenditures as Shares of GDP.

Slovak Republic 2000

Slovenia 2002

Lithuania 2002

Latvia 2002

Estonia 2001

Czech Republic 2002

Hungary 2002

Poland 2001

024681012141618 Administration Education Health Social Asst Housing Recreation Other

Education 79. Education is the largest single item of local government expenditure. It accounts for one-quarter to one-half of total government expenditure in the EU8 countries (Figure II.1).56 By the same token, local governments dominate the provision of education in the EU8. Local governments (including, in this case, intermediate levels of elected government) account for roughly two-thirds of total public sector spending on education in all but two of the EU8 countries (Table II.2).

55 Powiats later became responsible for upper-secondary and technical education, welfare homes, social assistance for the elderly, handicapped, homeless and families in crisis, unemployment assistance, protective and curative health, public health, culture and sport, environment protection, county roads, county planning, police and fire protection, etc. Voivodship responsibilities include university education, welfare homes, unemployment assistance, healthcare, sports and culture, environmental protection, natural disasters, regional roads, regional planning, tourism, police, fire protection, civil defense, electoral rolls and regulations. 56 The figures for Slovakia predate the decentralization of education in 2002. Figure II.2 is based on the IMF Government Finance Statistics Yearbook, 2004. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 63

80. Municipalities do not necessarily discharge this function directly. In Hungary, for example, only about one-half of the municipalities operate their own schools. Smaller municipalities tend to discharge this function through municipal associations or by contracting with outside Table II.2. Local Share of providers, including larger municipalities (Fiszbein, 2001). In Poland, intermediate Total Public Education tiers of government provide the more specialized higher levels of education. . Spending Gminas are only responsible for operating the six-year primary schools and the Country % first three years of secondary education. District-level governments (powiats) Czech Republic 65 (2002) are responsible for the last three years of secondary education. In the Czech Estonia (2002) 67 Republic similarly, municipal governments provide only primary education, with Hungary (2002) 61 regional governments assuming responsibility for secondary schooling. Slovakia Latvia (2002) 70 has recently followed suit (EC, 2003). In Slovenia, municipalities provide primary Lithuania 64 education, with the central government providing schooling at the secondary (2002) level. Poland (2001) 72 Slovakia.(2002) 27 Slovenia (2002) 21 81. The extent of municipal management control over education is limited. In Poland, locally-appointed school directors are theoretically free to make their own decisions about staffing levels and salaries. They are constrained, however, by legislation which sets out the framework for teachers’ salaries, teaching loads, and the rules for recruitment and promotion. Rules governing the salaries of teachers, for example, are established in the Teacher’s Charter and Ministry of Education regulations (Fiszbein op. cit.). Although a teacher may be dismissed when a school is totally or partially liquidated, gminas must provide six months severance pay and re-employ the teacher at the first opportunity. In most of the EU8 countries, collective bargaining agreements at the national level determine the extent of annual changes in teachers’ wages. Health 82. Local governments play a major role in managing healthcare in the EU8, but only a minor role in its financing. All of the EU8 countries now finance the majority of healthcare costs through compulsory insurance schemes. As discussed below, these are generally funded through payroll taxes (including employee- contributions), supplemented by central government budgetary contributions on behalf of pensioners and the unemployed. In most of the EU8 countries, HIFs provide funding directly to healthcare providers – primary physicians and secondary and tertiary healthcare facilities. As a result, subnational governments account for less than three percent of the total public sector health spending in the Baltics, the Czech Republic, Slovakia, and Slovenia (Table II.3).

83. Local governments, nevertheless, continue to play a major role in healthcare management, both as employers and as owners and operators of medical facilities. Primary healthcare is typically provided through municipally owned facilities. In some cases, Lithuania for example, the staff are municipal employees. In Hungary, municipalities have the option of employing primary physicians on the municipal payroll, but can also employ them as private contractors operating in municipally owned facilities (in which case they are paid according to the number of patients registered to them). Primary doctors can also work as private providers without a municipal contract, receiving capitation payments from the HIF. In the Czech Republic, most family doctors take the latter option, setting themselves up in private practices, renting facilities from the municipality, and billing the health insurance funds.

84. In Poland, primary healthcare has, since 1972, been provided by integrated healthcare units (ZOZs). These offer primary, outpatient, and secondary (non-specialized) treatment through a network of facilities in a defined geographical area (Girouard and Imai, 2000). Most are subordinate to the voivodships, although some larger municipalities operate facilities on behalf of the ZOZs in their territory. The 1994 policy document Transforming Primary HealthCare in Poland called for primary care to be separated out and become the responsibility of the gminy, and by 1998, one-third of the 3300 primary healthcare centers were administered by gminy. Since the introduction of a national health insurance system in 1999, primary care providers have been paid on the basis of a per enrollee capitation fee, although the majority continue to be public sector employees.57

57 The system was initially organized as sixteen separate regional funds, but in 2003, it was consolidated into a single national fund, with sixteen branches. 64

85. In most of the EU8 countries, responsibility for managing secondary and tertiary care is divided between municipal governments, intermediate tiers of government, and deconcentrated units of the central government. In Lithuania, responsibility for secondary and tertiary care is shared between municipalities and counties. In general, county governments own large Table II.3. Subnational Spending multi-specialist hospitals which provide secondary and tertiary inpatient on Health and Social Assistance and outpatient care to the acutely and chronically ill. Municipalities own (% of total public sector polyclinics providing outpatient specialty care and dispensaries. Some spending on these functions). municipalities also own multi-specialty hospitals providing secondary Health Social acute and chronic inpatient and outpatient care (not only to their own assistance Czech Rep. 2 6 residents but also to residents of neighboring jurisdictions). In Poland, Estonia 3 9 the majority of secondary and tertiary care is provided through the Hungary* 44 10 ZOZs, which (as noted earlier) are subordinated to the voivodships. Latvia 3 6 In the Czech Republic, hospital care is largely the responsibility of the Lithuania 1 9 newly-created regional governments. Prior to the recent abolition of the Poland* 93 6 districts (deconcentrated units of central government) larger hospitals Slovakia 0 1 were operated by these. Smaller hospitals (less than 200 beds), health Slovenia 1 1 Source: GFS . Data is for 2002 (except centers and polyclinics were owned by the municipalities. After the Poland-2001). Includes spending by abolition of the districts (2001) the majority of district facilities were intermediate tiers of elected transferred to the newly-created regions (Jaroš et. al., 2004). While the government. *Includes spending financed by majority of funding for secondary and tertiary care is provided by the regional health insurance funds. health insurance funds, municipalities continue to bear management responsibility for the facilities they own. Social Assistance 86. In all the EU8 countries, central governments provide the vast majority of social protection. This includes old age pensions, disability pensions, sickness benefits, and short term unemployment benefits.58 These are provided in the form of direct transfers to eligible individuals. Local governments, nevertheless, are responsible for a variety of assistance to other vulnerable segments of the population, including orphans, the homeless, those with physical or mental impairments, alcoholics, drug addicts, and particularly the long- term unemployed and their dependents.

87. Local governments have varying degrees of discretion in determining who is eligible for such benefits and what form they should take. In Latvia, for example, this has become increasingly circumscribed. Until 2002, the Law on Social Assistance authorized municipalities to provide a “poor family social assistance benefit” as well as a housing benefit – generally in the form of a subsidy for rent or utilities – and subsidies for food costs and medical assistance. Local governments enjoyed a high degree of autonomy in the way they administered these programs. The mayor (or in larger jurisdictions, the local social worker) had considerable discretion in determining who was eligible and often used this discretion to exclude the households they believed were “undeserving”. Benefits were more often provided in kind than in cash, taking the form of vouchers for the purchase of groceries at municipal stores, free lunches for the children of poor families, or part-time home care for the elderly or disabled. In 2002/2003, Latvia consolidated the various municipal benefits into a single Guaranteed Minimum Income (GMI). GMI benefits are now provided in cash and are allocated on the basis of explicit income criteria. To guard against applicants’ propensity to hide informal sources of income, the income test is supplemented by an estimate of the value of the applicants’ assets.

88. In Lithuania, similarly, social assistance (as of legislation enacted in 2003) is provided to families (and single persons living alone) with incomes falling below a government-determined minimum income level. The benefit is equal to 90 percent of the difference between the applicant’s actual income and the GMI. As in Latvia, the income test is supplemented by an estimate of assets. Although the benefit is funded from the government budget, it is administered by the municipalities, which retain some discretion over its application. Municipalities can, for example, require unemployed members of the household to participate in “socially useful activities.” They can substitute in-kind benefits for cash payments, and they can reject any application that appears to be based on false or incomplete information. The new law also permits municipalities to supplement the GMI benefit with additional benefits, in cash or in kind, at their own discretion. Estonia

58 Pension systems have been partially privatized in several countries in the region. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 65 has followed suit. Social assistance is guaranteed to persons whose income, net of housing expenses, falls below a government defined minimum level.59 Municipalities are entitled, however, to refuse subsistence benefits to any “person who has without justifiable reason, turned down suitable work offers or refused to participate in training programs authorized by the municipality” (Leetmaa and Vork, 2005).

89. Outside the Baltics, the pattern in much the same. In Poland, eligibility for social assistance is defined in income terms (although beneficiaries must also demonstrate a cause, such as mental or physical disability, alcoholism, and, increasingly, long-term unemployment.)60 The eligibility thresholds are changed every 3 years, on the basis of recommendations from the Institute for Labor and Social Issues. In Hungary, eligibility for social support is based on the level of the old age pension. Working-age applicants must have lost at least two-thirds of their ability to work and have a monthly income of less than 80 percent of the minimum old age pension. Benefits are equal to the difference between actual income and 80 percent of the minimum pension.61 Hungarian municipalities also provide a wide range of other forms of social assistance, including cash support to low-income families with dependent children, and medical benefits (free pharmaceuticals) and supplementary old-age pension to individuals meeting income eligibility criteria. In the Czech Republic, municipalities maintain homes for the elderly. The regional governments maintain facilities for handicapped children and adults and homes for abandoned children. Housing and Infrastructure 90. With the privatization of the housing stock, municipal responsibilities for housing management and maintenance have declined sharply in the EU8 countries. Municipalities have, on the other hand, assumed responsibility for water supply, sewage, district heating, solid waste management, and third tier roads (urban streets and rural feeder roads). These do not necessarily account for a large proportion of budgetary expenditure. Water, sewage, district heating and solid waste management are generally organized as separate, tariff-supported, enterprises.62 They do, nevertheless, constitute an important source of contingent liabilities (as these enterprises often have high levels of arrears) and a claim on scarce investment resources. Poor standards of original design and construction, high energy requirements, changing patterns of demand, and long periods of deferred maintenance have resulted in deteriorating and obsolete infrastructure networks (World Bank, 1998).

91. Generally, EU8 governments initiated the process of water company decentralization by transferring ownership of the assets and operational responsibilities from central or regional water companies to municipal governments. Laws governing tariffs and private sector participation were enacted much later, leaving the municipalities to confront deteriorating assets, declining government subsidies, and inadequate tariffs on their own. Once the regulatory framework was in place, municipalities were free to conduct their own reform initiatives. Some have responded by concluding operating or concession contracts with private operators (OECD, 2002).

92. In Poland, water supply was initially the responsibility of the 49 (pre-reform) voivodships. Each was responsible for water supply and wastewater discharge and controlled domestic and industrial tariffs. Legislation in 1990 authorized municipalities to organize water companies as municipal budgetary units or as commercial enterprises.63 Although some State-owned companies remain, the majority of utilities are now owned by the gminy. These are generally organized as budgetary enterprises or as commercial code companies.64 Where a single water and wastewater utility serves multiple municipalities, municipal associations are often organized. Although legislation permits private sector participation, most municipalities retain 100 percent ownership of the utilities operating on their territory (Szulinska, 2002).

59 In 2003, the subsistence level was 500 EEK (US$36) for the first member of the household and 400 EEK for each additional member of the household. For 2005, the benefit was increased to 750 EEK (about US$63) for the first member and EEK 600 for other family members. 60 Cash benefits are provided to single person households with income not higher than PLN 461 (US$138) monthly; and persons in families whose income per capita is not higher than PLN 316 monthly. 61 Unemployed persons whose income does not exceed 70 percent of the minimum pension are also eligible for social assistance. 62 Under GFS rules, revenues and expenditures of municipal enterprises are not recorded as budgetary expenditures, except to the extent they receive subsidies from the general budget. 63 A second law (in 1996) required water companies to be organized as commercial enterprises, but has not been universally implemented. 64 In 1999, there were 740 municipal water companies in Poland. Of these, 396 were organized as budgetary units and another 344 organized as commercial enterprises. 66

93. A similar process of decentralization took place in the Baltics. In Lithuania, prior to 1991, potable water supply and wastewater collection and treatment were managed by a state water company operating through 14 regional companies (Pietila and Spokas, 2004). After Lithuania regained its independence, responsibility for public water supply and sewage transferred from the state to municipalities. The assets and employees of the 14 regional companies were transferred to 45 newly created municipal water companies.65 During Estonia’s Soviet era, a single State Water Supply and Wastewater Board served all jurisdictions except Tallinn. In 1992, the Board was converted into a state-owned enterprise. In 1995, it was liquidated and its assets were transferred to the municipalities. Most became public-limited liability companies, wholly owned by the municipalities. The ownership of the Tallinn water company was transferred from the State to the city in 1991. In 1997 a limited liability company was formed, owned 100 percent by the city. In 2001, a majority of shares in the utility was sold to the private sector.

94. In Czechoslovakia, decentralization was begun under a Government Notice of 1991 and subsequent legislation in 1992. Nine State utilities (7 regional and 2 in the capital) were subdivided into 38 smaller state utilities (along the boundaries of district administrative units). This was followed by a complete transfer of ownership from the State to the municipalities. As no standard form of transfer was imposed, some municipalities set up independent water companies, while others formed joint companies serving several jurisdictions. The former was more common where municipalities possessed their own sources and networks; the latter, where they had supply/collection networks which crossed municipal boundaries. Private sector management is more common in the Czech Republic than in other EU8 countries. The French company Veolia, for example, operates the water supply systems serving Prague and much of northern and central Bohemia.

6. Revenues 95. To finance these expenditures, the EU8 countries had to make fundamental changes in the revenue structures of local governments in the post-soviet period. Functional decentralization required an increase in local government revenues. Privatization changed the nature of the tax base, and changed political conditions – hostility to the former centralized model and the emergence of multiparty democracy – increased pressure for local fiscal autonomy, while raising the risk that local governments would use that autonomy to thwart central government fiscal and equity objectives.

Figure II.2. Structure of Government Revenues as % of GDP.

Slovak Republic 2000

Slovenia 2002

Lithuania 2002 Taxes on Property Taxes on Income Estonia 2001 Other Taxes Latvia 2002 Grants Czech Republic 2002 Other Revenues Hungary 2002

Poland 2002

0% 2% 4% 6% 8% 10% 12% 14% 16%

Financing Discretionary Expenditures 96. The revenues of local governments can be grouped into two main categories. The first consists of discretionary funds, largely revenue sharing (and to a small degree local tax revenues and income from fees and charges). The second is funding for specific functions, principally education and health. In most of the EU8 countries, the specific purpose of this funding is to cover the full costs of the function to which it is assigned. There are two exceptions to this: Poland and Hungary. In both cases, the level of funding for education is determined independently of the costs of this function in the expectation that any shortfall will be funded from discretionary revenues.

65 The number of companies is smaller than the number of municipalities (60), because some water companies operate the water supply and sewerage systems of more than one municipality. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 67

Revenue Sharing 97. The primary source of discretionary revenues in all the EU8 countries is revenue sharing. As shown in Figure II.2, in all but one country, it is a share of the personal income tax (PIT).66 In all the EU8 countries, the shared taxes are administered by the central government at nationally uniform Table II.4. Personal rates. Fixed shares of the revenues are then assigned to local government. Income Tax Shared with Subnational 98. The percentage of the PIT that is shared with municipalities varies among Governments the EU8, ranging from 30 percent in the Czech Republic and Poland to 73 percent (2002). in Latvia and 94 percent in Slovakia (Table II.4). These variations are partly explained by the presence or absence of supplementary forms of central government funding. Czech Rep. 30% In the Czech Republic, shares of the PIT are supplemented by an equal percentage of Estonia 56% the revenues of the corporate income tax and the value added tax. Variations are also Hungary 40% explained by the rate of the tax and its consequent yields. Both Latvia and Slovakia Latvia 73% impose the PIT at a relatively low, flat rate (9 percent and 19 percent, respectively.) Lithuania 43% Poland 30% Large shares are associated with comparatively low yields. Variations are also explained Slovakia 94% by the nature of functions to be financed from tax sharing. For example, until 2002, Slovenia 35% Lithuanian municipalities received an average of 70 percent of the PIT.67 With the introduction of a capitation based transfer to fund teachers’ salaries, that share was reduced to a uniform 43.5 percent. It has since been increased to 47.4 percent (2005).

99. The stability of the tax sharing arrangements varies. In Latvia and Estonia, the municipal share is fixed in the Income Tax Law and can only be changed by amendments to these laws (although this has not prevented it from changing: in Estonia, the local share fell from 100 percent in 1991-93 to 52 percent in 1995, but it has remained stable at 56 percent since 1999). In Slovakia, the percentage share is also fixed in legislation, although the formula for distributing it is, at present, determined by government resolution. In the Czech Republic, Lithuania and Hungary, the municipal share of the PIT is specified only in the annual budget law and can be changed annually.

100. The distribution mechanisms for the PIT include a high degree of equalization. Distributing PIT shares solely on the basis of origin would result in wide disparities in per capita revenues among local jurisdictions. All the EU8 countries employ some form of equalization to reduce these variations. Three techniques are used. The first is to distribute the PIT on the basis of origin and provide additional funding to poorer jurisdictions from general central government revenues. In Estonia, for example, the local share of the PIT is distributed solely on the basis of origin. The central government then provides additional funding to poorer jurisdictions until they reach a fixed percentage of the national per capita average. In Slovenia, similarly, local governments receive 35 percent of the PIT, distributed on the basis of origin. The government then uses a complex formula to supplement the revenues of poorer jurisdictions. The fiscal equalization formula in Slovenia attempts to ensure an equal level of “relevant” spending’ in all jurisdictions. In principle, ‘relevant spending’ is defined as the amount needed to comply with local government tasks as provided in legislation. In practice, it is a nationally uniform per capita figure fixed annually by the Parliament. To calculate the amount of subsidy to a given municipality, this figure is first multiplied by the population of the municipality and four adjustment factors (favoring municipalities with lengthy road networks, large territories, and disproportionately large numbers of children and people over 65). The expected tax revenues of the jurisdiction (including revenue from the PIT and various local taxes) are then subtracted from this figure. Municipalities whose relevant expenditures exceed their projected revenues receive a subsidy covering the difference. Municipalities whose expenditure falls below estimated revenues are allowed to keep the surplus.

101. The second technique is to distribute the PIT entirely on an origin basis and then require richer jurisdictions to share part of their PIT revenues with poorer ones. This so-called fraternal equalization method is used in Lithuania and Latvia. In Lithuania, municipalities with per capita revenues over 165 percent of the national average must transfer all of the surplus to an equalization fund. The fund is then used to bring poorer jurisdictions to 65 percent of the national average. Any funds that remain once this target is achieved are transferred to municipalities on the basis of need indicators, the most important of which is school enrollment.

66 The exception is the Czech Republic, where the PIT, the VAT and the corporate income tax are all subject to sharing. Figure II.3 is based on the IMF Government Finance Statistics Yearbook, 2004. 67 The share varied from 30 percent in the biggest cities to 100 percent in the smallest ones. 68

102. Latvia uses a more complicated technique. The process begins with a calculation of expenditure needs. Aggregate expenditure needs are first calculated on the basis of total municipal expenditures in the preceding year, adjusted for inflation (the figure is further refined during negotiations with the Union of Local Governments). The projection of aggregate spending is then divided between republic cities and other local governments on a 45/55 basis. Within each of these two groups, the relative expenditure needs of each municipality are then calculated on the basis of six criteria: population; the number of children aged 0-6; the number of children aged 7-18; the population above retirement age; the number of children in orphanages; and the number of elderly in retirement homes. If a given municipality’s projected revenues exceed its projected expenditure needs by more than 10 percent, the municipality is subject to a revenue cap. Forty-five percent of the surplus is taken away and allocated to an equalization fund. If projected tax revenues are less than 90 percent of estimated expenditures needs, the gap is fully met, using proceeds from the equalization fund. Any shortfall between contributions into the equalization fund and payments out of it are to be covered by the central government.

103. The third technique is to distribute the PIT on a basis other than origin. The Czech Republic, for example, distributes the municipal share of the PIT (as well as the VAT and CIT) on a per capita basis, weighted to favor larger jurisdictions (a separate, more complicated, formula is used to distribute the PIT to a regional government). In Slovakia, similarly, the municipal share of the PIT is distributed on a per capita basis.

104. Poland uses a combination of the first and second methods. Municipalities are entitled to 27.6 percent of the PIT (powiats receive another one percent and voivodships 1.5 percent). The municipal share of the PIT is distributed on the basis of origin. Municipalities whose per capita revenues exceed 150 percent of the average are required to contribute to an equalization fund. These contributions, plus additional funding from the central budget equal to at least one percent of state budget receipts, are used to bring poorer jurisdictions up to 85 percent of the national average.68

105. Hungary uses a combination of the first and third methods. Ten percent of the PIT is distributed on an origin basis. This is subject to equalization: municipalities with per capita revenues less than 90 percent of the national average receive a supplement to bring them up to the average. The remainder of the municipalities’ share of the PIT is distributed according to various ‘normative criteria’. These include a flat HUF 2 million (approximately US $10,000) for each village, with an additional US $2.20 equivalent per capita to cover “administrative, communal, and sports related tasks” and US $10.50 per resident living outside the municipal administrative area. County governments also receive a flat US $1.3 million equivalent plus US $2.85 per capita from the subnational share of the PIT, and an additional US $77 per person served at county institutions (as described later, these normative grants are partly financed by central government general revenue).

106. All of these approaches produce a largely similar result: the disparities that would result from distributing the PIT on the basis of origin are substantially flattened. Using general government revenues to top up the PIT receipts of poorer jurisdictions has much the same effect as requiring richer municipalities to transfer some their PIT revenues to poorer ones, or distributing the PIT on a per capita basis (the degree of flattening depends, of course, on the specific parameters of the formula). Their more important difference may lie in their differing degrees of complexity. The Lithuanian system, which targets a fixed level of equalization of a single parameter (per capita revenues) is perhaps the most straightforward. The Hungarian system, with its multivariable distribution criteria, is perhaps the most complex.

Local Taxes 107. None of the EU8 countries assigns a major unrestricted tax base to local government – unrestricted in the sense that municipalities have the authority to set the rate and retain the revenue. The major taxes of the EU8 – the VAT, payroll taxes, excise taxes and the personal income tax – are imposed at centrally determined rates and retained or redistributed by the central government. As a result, local discretion over revenue is limited.

68 Revenues subject to equalization include not only the PIT but also the municipal share of the CIT (5 percent), as well as smaller locally administered taxes, whose revenues are estimated on the basis of standardized rates (OECD, 2001). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 69

108. The most widespread form of local taxation is the property tax. All of the EU8 countries impose a property tax for the benefit of local government (Table II.5). The nature of the tax base varies. Estonia taxes only land. In Poland, the real estate tax includes buildings and land (except agricultural and forest land which are subject to separate taxes). Lithuania has three forms of property taxation: one on buildings; one on leased land; and the third on land itself. The building tax is imposed only on buildings owned by enterprises. Lithuanian authorities are currently drafting plans for a residential property tax which would include both land and buildings.

109. Local discretion over property tax rates varies among countries. In Latvia, it is virtually nil. Local governments receive the revenue of the property tax, but the rate is fixed by the central government and the revenue is distributed on the same basis as the PIT transfer. In Poland, the Ministry of Finance sets a ceiling on the property tax rate, but local governments are free to fix the rate below that ceiling and to grant exemptions and deferments to individual taxpayers. In the Czech Republic, municipalities can set real estate tax rates within a bandwidth (0.3 percent to 4.5percent) (Oliveira and Martinez-Vazquez, 2001). In Hungary, similarly, municipalities may impose a building tax rate of up to 3 percent of market value, or a communal tax rate of up to 2 percent. 69

110. Regardless of local particulars, the property tax generates very little revenue in any of the EU8 countries (Table II.5). Property tax revenues are less than 0.5 percent of GDP in all the EU8 countries except Latvia and Poland. In Hungary and Slovenia, their contribution Table II.5. Property Taxes (% of GDP). is virtually nil. Even in Poland, the property tax constitutes only Selected Western 1.4 percent of GDP. To some degree, this is due to technical EU8 countries European Countries Czech Rep. 0.201 Denmark 1.18 constraints. At the beginning of the 1990s, none of the EU8 Estonia 0.444 France 3.48 countries had a well-functioning property market. The lack of Hungary 0.202 Germany 0.4 reliable market prices forced governments to value property Latvia 0.898 Netherlands 0.7 on the basis of arbitrary point values. Land was assigned an Lithuania 0.492 UK 1.8 arbitrary value per square meter depending on location, while Poland 1.369 buildings were assigned an arbitrary value per square meter Slovakia 0.386 according to use. These values were deliberately set low in order Slovenia 0.001 to avoid controversy. Valuations were also suppressed in order Source: IMF Government Finance Statistics Yearbook; Office of the Deputy Prime Minister (UK data). to avoid the particular problem involved in imposing a tax on assets that do not produce the income to pay it. In the recession that followed the Soviet collapse, the owners of privatized housing lacked incomes commensurate with the value of their newly acquired assets (Malme and Youngman, 2001).

111. Several of the EU8 countries have attempted to address the technical constraints. In 1993, Estonia became the first country in the region to introduce a tax-based on market value (on land only). It was followed by Latvia in 1998. Lithuania has introduced market elements in its area-based rates (on buildings and land) and has developed up-to-date land value maps in anticipation of a new Law on Land Taxation in 2005 (Malme and Youngman, op. cit.). Slovenia, similarly, has developed a property registration system and cadastre and is preparing for the introduction of market value-based tax on real estate in 2005. Poland has been debating to shift to market-based valuation for nearly a decade, although it has not yet done so.

112. However, there is little evidence that shifting the basis for valuation will have a dramatic effect on revenues. International experience suggests that the constraint on property tax revenues is ultimately more political than technical (Dillinger, 1995). Unlike the VAT or excise taxes, the property tax is imposed directly on the ultimate taxpayer, rather than being hidden in the form of higher prices. And unlike payroll and income taxes, it cannot be withheld at source. As a result, it tends to generate a level of political opposition that is disproportionate to its yield. Western European countries, which already value property on the basis of market prices, derive little revenue from it. The EU8 countries can expect the same.

69 The communal tax has two parts. The so-called ‘communal tax on individuals’ is payable by property owners, based on the floor space or market value of their property. The ‘communal tax on entrepreneurs’ is payable by any person who performs an economic activity in his own name, and is assessed on the basis of the number of staff. Municipalities are not permitted to impose more than one form of property tax on the same property. 70 Sector-Specific Financing 113. The second category of revenue consists of funding for specific services. All the EU8 countries have supplementary arrangements for financing two major items of municipal expenditure: education and health.

Education 114. The post-socialist-era systems of education financing reflect an ongoing tension among multiple objectives.70 The first is to ensure a basic level of education financing in all jurisdictions, regardless of the strength of their local tax bases. Thus, all the EU8 countries finance the largest component of education – teachers’ salaries – through some form of intergovernmental transfer. The second is to encourage efficiency in the use of education funding. All the EU8 countries, therefore, use some method to constrain the demand for transfers for central government support. One particular focus has been on spending on under-enrolled schools. The EU8 countries inherited an over-dimensioned system of primary and secondary education from the Soviet era. Schools were designed for an expanding population and one that was still to a large extent rural, but the number of school age children has dropped precipitously in the EU8 countries, and in rural areas more so than in urban ones. As a result, Ministries of Education confront an oversupply of school rooms and teachers. The education financing system of most EU8 countries now reflects attempts to address this problem using fiscal instruments.

115. At one extreme are two countries – Estonia and Slovenia – that rely solely on central government administrative procedures to ration the level of local spending on education. In Estonia, the level of funding for teachers’ salaries is based on the number of teaching positions authorized by the Ministry of Education and salary levels are agreed upon during annual negotiations with the teachers’ unions. The government finances these costs through earmarked transfers to the municipalities’ budgets. Slovenian local governments are only slightly less constrained. Although the občine (municipalities) are the legal owners of public elementary schools, the rules of job classification, which determine the number and type of posts in a school, are subject to approval by the Minister for Education who also lays down the criteria of financing (which are applied without variations to the whole country).

116. At the other extreme are countries that distribute funding for education on a per pupil (capitation) basis and allow local governments considerable discretion over how these funds are used. In theory, this approach has several advantages. It ensures a minimum level of education financing in all jurisdictions while allowing the central government to ration the level of such spending through its control over the per-capita amount. At the same time, it permits local governments to find the most efficient means of providing education within this overall spending envelope. In particular, it imposes efficiency measures on jurisdictions with under-enrolled schools. Under a capitation-based formula, falling enrollment will cause a drop in school funding, forcing local government to close schools they can no longer afford.

117. Five of the EU8 – Poland, Hungary, Lithuania, the Czech Republic and Slovakia – now employ some form of capitation-based financing for primary education. None of them employs a single, nationally uniform amount, however. In all five countries, capitation rates are adjusted to reflect ostensible differences in the costs of providing education. In Poland, for example, rural schools receive a 33 percent supplement over the basic per capita amount. Towns with populations under 5,000 receive an 18 percent supplement.71 Hungary also supplements its standard per capita amount (equal to about US $440 per student in grades 1-8 and US $570 per student in grades 9-13) with additional funding (US $58) for primary education in villages with populations of 3,000-3,500 and those with fewer than 3,000 inhabitants (US $116).72 In Lithuania, the capitation amount is determined according to a complex formula that differentiates among both school characteristics and grade levels. For students in grades 1-4, rural elementary schools receive roughly twice as much funding per student as those in municipal capitals and big cities. Schools with fewer than 150 children receive 50 percent more than those in the 150-299 student bracket and nearly twice as much as those with 300 or more students. In the Czech Republic, capitation figures distinguish among different levels and forms of education as well as among regions. Regional variations are intended to reflect variations in labor costs, and therefore favor, rather than discriminate against, Prague.

70 See Chapter III for an extensive discussion of higher education financing. 71 To encourage gminas to hire qualified staff, the formula continues to distinguish among teachers with different levels of qualifications, based on five centrally determined pay categories. 72 The Hungarian education transfers are not intended to cover the full costs of education, nor are they earmarked for that purpose. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 71

118. These differentials have been used, in part, to soften the impact of the switch to capitation based financing in rural areas. Additional transitional arrangements have had to be introduced in order to smooth the adjustment process. In Poland, for example, the initial weights reflected teacher characteristics so that places with unusually high wage levels did not experience extreme cuts (this provision has since been phased out). The transition legislation also specified that no gmina’s subvention could be changed by more than ten percent in a given year, adjusted for inflation.

119. Two issues threaten the success of capitation based financing. The first is the inability of local governments to dismiss staff. While falling enrollments trigger a drop in funding, local governments often lack the legal authority or political will to make corresponding cuts in staff. In Poland, for example, school directors are, in theory, free to make their own decisions about staffing levels. However, regulations constrain dismissals: although a teacher may be dismissed when a school is totally or partially liquidated, a gmina must provide six month’s severance pay and re-employ the teacher at the first opportunity. Similar constraints on downsizing exist in Lithuania and Hungary. Political constraints appear to be particularly acute in municipalities where downsizing implies the closure of entire schools. The Czech Republic has addressed this problem by assigning the authority to establish or eliminate schools to the newly created regional governments (Hendrichova et al., 2001). While government funding for education is distributed among the regional governments on an (adjusted) per capita basis, each region is able to devise its own system of allocating funds to individual municipalities (EC 2002 and 2003).

120. The second threat comes from the unwillingness or inability of central governments to increase the level of capitation transfers to reflect centrally-mandated increases in costs. The principal determinant of costs – the wage level – is largely determined by the central government in the EU8 countries. In Lithuania, the national civil service laws sets out a structure of pay scales for municipal employees, based on grade, years of employment, and – in the case of teachers – class size and number of classes taught. The pay structure is expressed as a multiple of the so-called basic wage. As a result, government changes in the basic wage trigger automatic increases in salaries. In Poland, similarly, teachers’ salaries are determined on the basis of the Teacher’s Charter and annual ministerial regulations on the remuneration of teachers (Fiszbein op. cit.). Legislation in some of the EU8 countries requires the government to increase capitation grants in line with annual salary adjustments. The legislation enabling Slovakia’ recent shift to capitation based financing, for example, requires the capitation amount to be based on annual salary negotiations (EC, 2003). On the other hand, other countries have no provision linking education financing to increases in wages. In Poland, the level of central funding is determined as a fixed percentage (currently 12.8 percent) of the total projected government expenditure. In Hungary, the level is determined through annual adjustments in the normative grant distribution formula.

Health 121. The EU8 countries have also attempted to use the health financing system as a tool to achieve efficiency.73 All eight of the EU accession countries have created health insurance funds (HIFs) to finance the majority of primary, secondary, and tertiary healthcare costs. The HIFs are funded from payroll taxes, supplemented by central government payments on behalf of certain categories of non-contributors: pensioners; workers on maternity or sick leave; and the registered unemployed.

122. Health funds were generally created to isolate healthcare spending from the vicissitudes of the annual government budgeting process. As such, they supply a guaranteed source of funding for health, albeit one that is subject to fluctuations in the payroll tax base and in the rate of the payroll tax. In principle, they also function as a ceiling on healthcare spending. With few exceptions, spending on primary, secondary, and tertiary healthcare is not allowed to exceed the revenues of the HIF. (Latvia is one such exception: there, the health financing system explicitly contemplates annual subsides from the central government budget).74 In this way, the HIFs should impose a hard budget constraint on overall spending in the sector.

73 See Chapter IV for an extensive discussion of health financing issues. 74 Hungary is also an exception, although not an intentional one. In Hungary, the national health fund has been in deficit since its inception. The national government has responded by introducing an additional tax – the “hypothecated health care tax” and expanding the base and removing caps on the payroll tax. 72

123. Attempts to use fiscal instruments to ration the supply of health services on a more micro level have had mixed results. One of the major sources of inefficiency in the socialist model of healthcare was its excessive use of secondary and tertiary facilities. Primary physicians were trained in a narrow range of fields, requiring referrals to more expensive secondary and tertiary facilities. Performance incentives were also weak, including because compensation to medical staff was not linked to the volume of services they provided.

124. To varying degrees the EU8 countries have addressed the excessive use of secondary and tertiary facilities by introducing the so-called family medicine model of primary care. Under this approach, patients register with an individual doctor of their choice who becomes the primary point of contact in the healthcare system. These doctors receive training in a broad range of primary healthcare fields, limiting the need for referrals. To encourage physicians to register patients, the HIFs pay providers a standard rate for each patient on their roster. To encourage physicians to actually serve these clients, they often provide additional funding on a fee-for-service basis.

125. Performance incentives have proven incompatible with cost containment, however, due to the tendency of providers to respond to fee-for-service forms of compensation by increasing the volume of services they ostensibly provide. Slovakia, for example, introduced a capitation based system for primary care system in 1994. To encourage performance, it replaced it with a mixed system (60 percent capitation, 40 percent fee-for-service) in the following year. The resulting explosion in the number of claims forced the government to revert to an exclusively capitation-based system in 1998. It has now reintroduced a fee-for service element but limited it to specific preventive procedures (EOHCS, 2000 and 2004).

126. The Czech Republic arrived at a similar system by a different route. The Czech Republic initially attempted to control primary care costs by fixing an overall spending ceiling for primary care and allocating it among individual physicians on a fee for service basis. Physicians responded by inflating the number of treatments they claimed. Because the spending ceiling was fixed, more services meant smaller reimbursements per treatment. This prompted physicians to further inflate the number of treatments they claimed. The Czech Republic eventually brought an end to this vicious circle by introducing a strict capitation-based system in 1997. Under the current system, primary physicians receive a fixed amount per registered patient, differentiated by age group ( 0-4 year olds, for example, are worth 4.2 times as much as patients in the 20-24 year old range). Hungary, similarly, now reimburses primary care solely on a on a capitation basis, using a points system that distinguishes among patients in five different age groups. In Lithuania, the HIF reimburses health facilities on the basis of the number of patients registered with it. Each facility, in turn, is allowed to reimburse individual physicians on a fee-for-service or capitation basis.75

127. Rationing secondary and tertiary care has been even more difficult. To varying degrees, all the EU8 countries have experimented with the German points system. Under this system, each treatment eligible for reimbursement is assigned a number of points, based on its complexity and cost. A total budget (in monetary terms) for eligible healthcare is then agreed (in the German case) between individual health insurance funds and regional physicians’ associations. Individual physicians report the number of points they have accumulated to the physicians association, which calculates the value of a point by dividing the agreed budget by the total number of points reported. Individual physicians are reimbursed according to the number of points they have billed and the value of an individual point. In effect, this system encourages the physicians’ association to ration the supply of health services, as more points for one physician means lower reimbursements for all.

128. Hungary experimented with this approach during the 1990s. Between 1993 and 1997, separate points systems were introduced for outpatient services, acute inpatient services and chronic inpatient services. The amount of funding for each category of service was fixed in the yearly budget of the HIF. Points were added up nationally and the monetary value of one point was calculated by dividing the predetermined budget by the total number of points. But the Hungarian HIF proved less adept at restraining performance inflation than the German physicians associations. Claims increased and the value of a point

75 Salaries, however, remain the predominant form of compensation. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 73 dropped. In 2000, the HIF abandoned this approach. For outpatient care, the value of a point is now fixed in advance. To control the supply of services, fiscal penalties are imposed on individual providers. As of 2004, providers receive 100 percent reimbursements only for the level of services provided in the previous year – or more precisely, for 98 percent of that level. The excess, up to five percent, is reimbursed at 60 percent; between five and ten percent, at 30 percent; and above that level, at only ten percent. Individual providers therefore have a strong incentive to turn away patients once their quota has been filled (the HIF contracts for inpatient services on the basis of fixed numbers of beds for specific types of services).

129. Poland, similarly, now fixes the level of reimbursement per treatment in advance. To ration supply, the number of cases to be treated (under separate diagnostic-related groups) is also agreed in advance, and specified in the annual contracts between the sickness funds and individual healthcare providers.

130. The Czech Republic has largely abandoned the points system altogether. The Czech Republic experimented with a points system during 1993-97. A total of 4,500 distinct procedures were reimbursable, with points based on the amount of time required to carry out a given procedure. Hospitals also billed points for each day spent by patients in the hospital. Under the Czech system of regional sickness funds, each fund calculated the value of a point separately, by dividing its revenues from contributions by the total number of points billed to it. But, as in Hungary, the volume of claims exploded. In 1997, the Ministry of Health published a new list of procedures with new point numbers. This met with criticism both from providers, who believed that the new point numbers were insufficient to cover the costs of services, and from insurers, who argued that their revenues from contributions would be insufficient to pay the providers even the stipulated amounts. The points system was therefore largely abandoned. Sickness funds now reimburse individual hospitals on the basis of aggregate budgets, which are largely based on expenditure in the preceding year. The points system is merely used to determine whether each hospital has delivered an equivalent volume of services.

7. Debt Controls 131. At the beginning of the nineties, many of the EU8 countries (e.g. Hungary, Latvia, Slovakia, and Slovenia) permitted local governments to operate freely in the capital markets. Lacking experience in the preparation of bankable projects (and having a weakness for high-risk profit making ventures) many later defaulted. To prevent excessive borrowing in the future, most of the EU8 countries have introduced various forms of debt controls. Ex-ante quantitative restrictions 132. The majority of EU8 countries impose some form of ex ante quantitative ceiling on municipal borrowing (Table II.6). Estonia, Lithuania, Poland, Slovakia and Slovenia all place ceilings on the stock of debt as a percent of recurrent revenues (including revenue sharing): Estonia, Poland, and Slovakia set the ceiling at 60 percent; in Lithuania, the ceiling is 35 percent (except for the two major cities of Vilnius Kaunas, where the ceiling is higher); and in Slovenia the ceiling is ten percent. To assist in complying with Maastricht provisions, Poland also links growth in the stock of municipal debt to ceilings on the stock of debt in the public sector as a whole. Once the consolidated public debt exceeds 50 percent of GDP, growth in the stock of municipal debt cannot exceed the central government’s planned deficit, expressed as a percent of revenue. Local borrowing is entirely forbidden if the aggregate stock of public sector debt exceeds 60 percent of GDP.

133. The majority of EU8 countries also impose ceilings on debt service as a percent of recurrent revenues. These range from 10 percent (Slovenia) to 25 percent (Slovakia). Hungary also imposes a ceiling on debt service: there, the denominator excludes revenue sharing, and debt service is limited to 70% of own source revenues. Ex ante administrative controls 134. Several of the EU8 countries also impose case by case administrative controls on local borrowing. In Slovenia and Latvia, local governments are eligible to borrow only after receiving permission from the Ministry of Finance and the Local Government Borrowing and Guarantees Board, respectively. In Lithuania and the Slovak Republic such permission is required if the proposed debt would exceed the quantitative ceilings. In Poland, local government borrowing must be approved by the Regional Accounting Office, unless loans are taken from the bank where the local government holds its main account. 74 Restrictions of certain forms of debt 135. Several of the EU8 countries restrict certain forms of borrowing. In Poland and Slovenia, for example, local governments are not permitted to borrow externally. In all but two of the countries, legislation explicitly requires that long term loans be used for capital investment financing. In Estonia, local governments are allowed to take long term loans only for those capital investment projects which are included in the Local Development Plan.76

Table II.6. Summary of Municipal Debt Regulations. Country Limitation on Limitation on Flow of Requirement for Requirement Municipal Stock of Debt Debt Higher to use Long Insolvency (Limits on Debt Administrative Term Procedures (Limits on Total Service Payments) Body’s Permission Borrowing Outstanding to Borrow only for Debt) Capital Investments Czech Republic No No No No No 70% of the local governments’ own- Hungary No source revenues No Yes Yes 60 % of planned recurrent 20% of recurrent revenues, planned revenues, excluding excluding earmarked Yes No, but under Estonia earmarked grants grants No consideration

Latvia No No Yes Yes Yes 35% (for Vilnius and Kaunas 50%) of recurrent revenues, excluding 20% of recurrent earmarked grants revenues, excluding and short term earmarked grants and Yes, if above the Yes Lithuania loans. short term loans. set limits No Yes, unless loan is obtained from the 60% of recurrent bank that holds revenues.* 15% of recurrent the municipality’s Poland revenues major account. No No 60% of the previous year’s 25% of the previous Yes, recurrent year’s recurrent if above the set Yes Slovakia revenues revenues annual limit Yes 10% of the previous year’s 10% of the previous recurrent year’s recurrent Yes Slovenia revenues revenues Yes No * also subject to Maastricht ceilings

Source: Local Government and Public Service Reform Initiative. 2004. Local Government Borrowing: Risks and Rewards; and country sources.

Bankruptcy 136. To discourage excessive local government borrowing, EU8 countries do not rely on ex ante controls alone. Several of them impose penalties on municipalities that fail to meet their debt obligations. These include the loss of budget autonomy, the forced sale of assets, and the dismissal of the local council. In principle, this should discourage other municipalities from following suit in the future.

76 Local governments are not allowed to borrow in foreign currency in Poland. However, there is a Ministry of Finance Ordinance that determines that loans in another currency maybe incurred if they are issued by multilateral financial institutions or have an investment grade rating by an international rating agency. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 75

137. Three of the EU8 countries – Hungary, Latvia, and Slovakia – currently have municipal bankruptcy laws governing such circumstances. Hungary relies on the court system to administer bankruptcy proceedings (Box II.2). Remedies available to the court include the sale of assets and (with the concurrence of Parliament) the dismissal of the council, and the calling of new elections. In Latvia, the process is overseen by the Ministry of Finance, which appoints a financial supervisor to assist the municipality to prepare a stabilization plan, specifying budget cuts and refinancing needs. If the municipality fails to agree on a plan, the Cabinet of Ministers has the right to impose one. At the request of the Minister of Finance, the supervisor can also control all municipal expenditures and sign the municipality’s payment orders.

Box II.2. Hungary’s Municipal Bankruptcy Procedure Under the terms of Hungary’s local insolvency law, bankruptcy proceedings can be initiated—by a creditor or by the local government itself—when a debt is 60 or more days overdue. Cases are submitted to a court, which, after examination, can either order the start of debt settlement procedures and designate a financial trustee, or decline the debt settlement procedure if it determines that the local government can meet its obligations from existing cash flows and assets. Once a bankruptcy case has been approved by the court, the local government has fifteen days to appoint a “crisis committee” and 30 days to prepare an emergency budget (by law, the budget may only finance priority functions as specified in legislation). During this period, the court-appointed financial trustee must endorse all invoices prior to payment made by the municipality.

The municipality and the creditor have 210 days, once the bankruptcy procedure begins, to agree on a mutually acceptable resolution of the debt. If this deadline is not met, the court-appointed trustee prepares an asset and debt adjustment plan. On this basis the court can proceed to auction the local government’s assets. The court may also request Parliament to dissolve the local government and call for new elections. Criminal and civil proceedings can also be initiated.

The law has a successful record of implementation, although it has been used sparingly. Since 1996, there have been eleven municipal bankruptcy cases approved by court decision in Hungary. Nine cases resulted in a voluntary debt settlement agreement between the parties. In two cases the courts had to liquidate assets and liabilities based on the recommendations of the trustee.

138. Slovakia, similarly, relies on the Finance Ministry to oversee bankruptcy procedures. Under the Slovak regime, local governments are obliged to introduce a recovery regime once any financial obligation falls more than 60 days overdue. Municipalities have 120 days to demonstrate an improvement in their financial condition. If they fail to meet this deadline, the Ministry of Finance has the authority to impose forced administration, which includes the authority to approve all financial transactions of the municipality and to request the local government to adopt measures to increase revenues or decrease expenditures. If the municipality fails to do so, the administration may impose such measures unilaterally.

139. To date, these measures have been used sparingly. As noted in Box II.2, there have been only eleven municipal bankruptcy bases submitted to the courts in Hungary, and in Latvia seventeen municipalities have been subjected to the financial stabilization process.

8. Directions for Further Reform 140. Over the long term, the sustainability of all eight of these intergovernmental fiscal arrangements depends upon the ability of central governments to keep their agreements with local governments. In monopolizing the major sources of tax revenues, central governments have arrogated to themselves the power to decide how large a share of total public sector revenues local governments should receive, and where it should be spent. In the process, they have made themselves vulnerable to the claim that the level of funding which they have provided to local governments is insufficient.

141. In the case of sector specific financing, this has one clear implication. When the central government increases the costs of providing the service, it must increase the level of financing. Thus (as noted earlier) centrally-negotiated increases in teachers wages must be matched by an increase in capitation grants. 76

Box II.3: Calculating the Formula Grant in England and Wales Like their counterparts on the European continent, British local governments have limited taxing authority. They are confined to a property tax, which itself is limited to residential properties. The majority of local revenues are therefore derived from intergovernmental transfers, the most important of which is a so-called formula grant. In principle, the formula grant is calculated on the basis of revenue gaps. Expenditure needs are first calculated on the basis of national norms for education (preschool, primary, secondary, and adult education), social services, police, fire, highways, environment, and “capital finance” (the norms are based on biennial spending reviews and annual negotiations with the local government association). The norms are then adjusted according to the characteristics of individual jurisdictions—high poverty levels, high prevailing wages, and low population density all result in higher figures—yielding an estimate of each jurisdiction’s expenditure needs. Estimated revenues from the residential property tax are then subtracted from this figure, yielding an estimate of each jurisdiction’s revenue gap (to avoid discouraging tax effort, the revenue estimates are based on the average national rate rather than the rate actually imposed in each jurisdiction). The central government is under no obligation to fully fill these gaps, however. The level of Government spending on the formula grant is determined in the course of annual budget preparation. As the allocation normally falls short of the aggregate revenue gap, available resources are rationed by fixing a cap on the percentage increase allocated to any one jurisdiction. Source: Local Government Association (2005) RSG Settleme4nt 2004/05: a Guide.

142. The implications for funding discretionary expenditures are less clear. In the absence of a central government mandate, local governments are under no obligation to provide a specific level or mix of discretionary services. Thus, in principle, they are in no position to claim that they are underfunded. At present, the level of discretionary funding – i.e., the percentage of the PIT to be shared with local governments – is determined through annual negotiations between the central government and associations representing local governments. Such negotiations could easily break down.

143. Experience in Western Europe suggests that there are two solutions to this problem. The first is to rely on increasingly sophisticated estimates of the costs of providing the services on the discretionary list – in effect, to treat discretionary services as if they were mandatory. The British have perhaps the most elaborately developed version of this approach (Box II.2). It is echoed in Latvia’s intergovernmental transfer formula, which distinguishes the costs of a variety of specific services, as well as the revenue potential of individual jurisdictions.

144. The alternative would be to reduce the central governments’ role in financing discretionary expenditures and allow the local governments to decide for themselves on the appropriate level of spending, on the basis of their constituents’ willingness to pay. This would require the assignment of a broad based tax to the local level.77 The PIT is an obvious candidate. This approach has been attempted in some Scandinavian countries. Danish local governments, for example, are permitted to impose a surcharge on the national personal income tax at a locally-determined rate (the rate must be flat, rather than progressive). In 2004, the combined county and local tax rate ranged from 28.5-35.6 percent, with an average of 32.6 percent.78

145. The possibility of a Scandinavian-style PIT surcharge has been discussed in a number of EU8 countries, including Hungary, Poland, and Slovakia. It is also cited in the European Charter of Local Self-Government and actively promoted by the Council of Europe. However, it has few supporters among the EU8. At present, central governments prefer to retain near-total control over tax policy, and local governments prefer to avoid the onus of taxation.

9. Conclusions 146. Measured against the four criteria defined at the beginning of this paper, the EU8 countries have made substantial progress in the reform of intergovernmental relations.

77 The incidence of such a tax would have to be limited to the local jurisdiction, in order to prevent local governments from “exporting” the burden of their expenditures onto other jurisdictions. 78 The Danish system, nevertheless, includes a revenue equalization feature, based on the fraternal equalization principle. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 77 Transparency and stability 147. The allocation of central government budget resources to local governments is no longer a matter of closed-door negotiations within a Party hierarchy. The major source of discretionary expenditure financing – revenue sharing – is determined as a fixed share of specified taxes, distributed according to formulas. Education financing, similarly, is also largely formula based (except in Estonia and Slovenia, where the central government determines the funding directly). Primary healthcare is financed on a capitation basis in most of the EU8 countries.

148. Some of the EU8 countries perform better against the transparency and stability criterion than others. Hungary’s structure is perhaps the most problematic of the eight. As noted earlier, while some discretionary funding is provided from a fixed share of the personal income tax, the majority is funded from annual appropriations from general government revenues, raising uncertainty about the annual level of funding. These funds, moreover, are distributed on the basis of over forty variables, undermining its transparency. In this respect, Hungary would benefit from the examples of the Czech Republic or Slovakia, both of which determine the aggregate amount of revenue sharing on the basis of fixed shares of specified taxes, and distribute the funds on a per capita basis. Equity 149. The EU8 countries also manage to achieve a high degree of redistribution in the financing of social services. Primary and secondary education is financed from central government revenues with funds distributed on the basis of enrollment. Primary healthcare, similarly, is financed from pooled funds (in this case, HIF contributions) and distributed on a per-patient basis. Social assistance is an exception. Unlike education or health, social assistance is financed from the discretionary revenues in most of the EU8 countries.79 But discretionary funding is, itself, highly equalized. Through the various combinations of fraternal equalization schemes, top-ups, and equalizing distributional formulas, all eight countries substantially reduce the disparities that would arise from distributing discretionary funding on the basis of origin. This permits poorer jurisdictions to spend more on social assistance than would otherwise be the case. In fact it could be argued that discretionary funding is over-equalized. While there is a strong case for equalizing spending on social assistance, the case for equalizing spending on other functions financed from discretionary revenues – road maintenance, community amenities, or general administration – is not as strong. Macroeconomic control 150. All the EU8 systems are highly effective in asserting central control over tax policy. As in the rest of the EU, the vast majority of subnational revenues is derived from intergovernmental transfers. A local government’s tax authority is largely limited to the highly sensitive property tax. Local discretion over property tax rates is limited by rate caps. All except the Czech Republic also impose some form of control over municipal debt. Effectiveness 151. Attempts to use the system of intergovernmental relations to encourage greater efficiency in the production of public services have been less successful. While the majority of the EU8 countries now finance education on a capitation basis, this has not been sufficient to prompt the closure of under-enrolled classrooms or schools. In the health sector, efforts to encourage primary providers to increase the volume of services they provide have been thwarted by over-billing. Efforts to ration secondary and tertiary care using variants of the German points system have run aground for similar reasons.

152. Overall, the EU8 countries have made remarkable progress in reforming their systems of intergovernmental fiscal relations. The eight countries now have much more in common with the longer- standing members of the Union than they do with the socialist-era systems they inherited. However, the process is not completed. In the same way that Western European countries continue to adjust their systems to reflect ongoing changes in political and economic circumstances, the countries of the EU8 are likely to continue to adjust their systems for a long time to come.

79 Lithuania and Estonia are exceptions. In both cases, social assistance is funded directly by the central government. 78 References Barbone, Luca and J. Hicks (1995), “Local and Intergovernmental Finances in Poland: an Evolving Agenda,” in Bird, Richard, and Wallich, “Decentralization of the Socialist State.”

Beksta, Aruna and A. Petkevicius (2000), “Local Government in Lithuania,” in Horvath, Tamas (ed. 2000), “Decentralization, Experiments and Reforms.”

Berkowitz, Daniel and B. Mitchneck (1992), “Fiscal Decentralization in the Soviet Economy” in “Comparative Economic Studies” Vol 34.

Dillinger, William (1995), “Urban Property Tax Reform: Guidelines and Recommendations.”

Ebel, Robert, et.al. (2000), “Sorting out Government Roles and Responsibilities in the Hungarian Transition,” World Bank Institute Working Papers # 21025.

European Commission (2002), “The Education System in the Czech Republic 2001/2002,” Eurydice Education Data Base.

European Commission (2003), “Structure of Education, Vocational Training and Adult Education Systems in Europe: Czech Republic,” Eurydice Education Data Base.

European Commission (2003), “The Education System in Slovakia 2001/2002”, Eurydice Education Data Base.

European Observatory on Health Care Systems (2000a), “Health Care Systems in Transition: Lithuania.”

European Observatory on Health Care Systems (2000b), “Health Care Systems in Transition: Czech Republic.”

European Observatory on Health Care Systems (2004a), “Health Care Systems in Transition: Slovakia.”

European Observatory on Health Care Systems (2004b), “Health Care Systems in Transition: Hungary.”

Fiszbein, Ariel (ed. 2001), “Decentralizing Education in Transitional Societies”, in World Bank, T. Levitas, and J. Herczynski: “Decentralization, Local Governments, and Education Reform in Post-communist Poland.”

Freedom House (1998), “Nations in Transit.”

Girouard, Nathalie and Y. Imai (2000), “The Health Care System in Poland,” OECD Economics Department Working Paper No. 257.

Hegedus, Jozsef (2002), “Decentralization and Structural Adjustment in Hungary.”

Hendrichova, Jana, et. al. (2001), “Czech Republic,” in Fiszbein Ariel (ed) op. cit.

IBRD (2003), “Latvia: Beyond Territorial Reform.”

IBRD (2002), “Lithuania: Issues in Municipal Finance.”

IMF (2004),“Government Financme Statistics Yearbook.”

Jaroš, Jan et.al. (2004), “Decentralization and Governance of Healthcare in the Czech Republic in the 1990s ,” LGI.

Youngman, Joan and J. Malme (2004), “The Property Tax in a New Environment: Lessons from International Tax Reform Effort.”, Lincoln Land Institute.

Leetmaa, Reelika and A. Vork (2005), “Social Insurance and Social Assistance in Estonia.” Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 79 Maeltsemees, Sulvlev (2000), “Local Government in Estonia,” in Horvath, Tomas (ed. 2000): “Decentralization, Experiments and Reforms.”

Malme, Jane, and J. Youngman (2001), “The Development of Property Taxation in Economies in Transition,” World Bank Institute.

OECD (2001a), “Fiscal Design Across Levels of Government”, Hungary Country Report.

OECD (2001b), “Fiscal Design Across levels of Government”, Poland Country Report.

OECD/DANCEE PROGRAMME (2002), “Models of Water Utility Reform in the Central and Eastern European Countries.”

Oliveira, Joao, and Martinez-Vazquez (2001), “Czech Republic: Intergovernmental Fiscal Relations in the Transition.”

Open Society Institute (2002), “Local Government and Public Service Reform Initiative and Energy Regulators Regional Association,” Training Material No. 1 Sector Background (June).

Pietila, Pekka, and R. Spokas (2004), “Comparison of Water Services Development in Finland and Lithuania”, in European Water Association: European Water Management Online.

Regulski, Jerzy, and W. Kocan (1994), “From Communism toward Democracy: Local Government Reform in Poland,” in Bennett, Robert (ed.): „Local Government and Market Decentralization: Experiences in the Industrial Developing and Former Eastern Bloc Countries.”

Regulski, Jerzy (2003), “Local Government Reform in Poland, an Insiders’ Story”, in Open Society Institute: “Local Government and Public Service Reform Initiative. “

Szulinska, Justyna, (2002), “Water Supply Systems in Poland – The Safety Aspects.”

World Bank (1998), “Living by Serving: Infrastructure Sector Strategy.”

World Bank (2003), “Poland: Toward a Fiscal Framework for Growth – A Public Expenditure and Institutional Review.”

World Bank (2004), “Implementation Completion Report, Latvia Welfare Reform Project.”

Zickel, Raymond (1991), “The Soviet Union: a Country Study.” 80

III. FINANCING HIGHER EDUCATION80 1. Introduction 153. While historically, EU8 universities were elitist in nature, there were also inherited differences among these countries. Before 1945, in the CEE countries, higher education systems had broadly evolved along the lines of the Austro-Hungarian and German models; in the Baltics, the model was closer to the Soviet system. Under communism, universities were, at least notionally, treated as homogenous and equal. State financing was organized centrally and on a historical basis. While higher education was free of charge, access to higher education was, for the most part, in the hands of the universities, and institutions were able to control admissions through their own entrance examinations, a practice which was opaque and frequently inequitable.81

154. In the Communist era in Central Europe, the proportion of the working-age population that had attained tertiary education was between 11 and 15 percent, with at most, modest increases since the 1960s. In the Baltic countries, the proportion was larger. Everywhere, tuition was free. Although this involved a concentration of a large amount of government expenditure on a relatively small minority of the population, this was not wholly inequitable, since, in centrally planned economies, the State was virtually the sole employer of graduates, whose jobs may have had more power and prestige attached to them, but whose pay scales were marginally (if any) higher than those for skilled workers.

155. With the coming of a market economy and substantially greater opportunities for university graduates, both in local private sectors and now potentially elsewhere in the EU, the private benefits from university education have increased considerably both in terms of higher income and lower rates of unemployment. This has been reflected in sharply increased demand for tertiary education in all post-socialist countries. In some countries, such as Poland, Hungary, Latvia, and Estonia, the growth in demand has been met partly by a rapid increase in private provision. In other countries, public institutions have increased their enrollment.

156. At the same time, university programs began to change. Prior to transition, higher education was biased toward programs in engineering and the physical sciences, with ministries of science and industry often setting enrollment quotas. Deregulation and increased autonomy of universities made higher education more sensitive to student choice and labor market developments, rather than the dictates of central planning.82 By 1995, there was a marked shift in enrollments everywhere, away from science and engineering courses and toward the social sciences, which had been neglected under socialism.

157. Throughout the 1990s, the difficulty of managing centrally a growing and more diversified higher education system became increasingly clear. Inevitably, enlarged participation began to raise questions about the sustainability, equity, and quality of the systems. Accordingly, new legal and quality assurance frameworks were developed everywhere to accommodate the need for more flexible institutions with new forms of governance and management. In most countries, private provision of higher education services emerged in response to the need for system expansion. Some countries developed a school leaving examination (the Matura system), which began to act as a national entry examination for the universities, thus introducing greater intellectual equity of access. By the later part of the decade, the Bologna process, which aims to establish a European Higher Education Area (with mobility of students and teachers, convergence towards a common framework of qualifications and cycles of study, and measures to encourage lifelong learning by 2010), was providing an incentive for even greater flexibility in the organization of degree programs.83

80 Prepared by Mary Canning, Martin Godfrey and Dorota Holzer-Zelazewska. Peer reviewers were Nicholas Barr and Jamil Salmi. Further helpful comments were provided by Thomas Laursen, Andrew Burns and Bernard Funck, and Andrej Salner. The authors also wish to thank the following individuals for data and for the appended country case studies: Josef Benes (Czech Republic); Thomas Havady (Hungary); Marina Meksa (Latvia); Andrej Salner (Slovakia), Eva Marjetic and Dusa Marjetic (Slovenia); and senior staff in the Ministries of Education of Lithuania and of the Slovak Republic. An earlier version of this chapter was discussed in a regional workshop in Warsaw in June 2005. Throughout the study, the terms “higher” and “tertiary” education are used interchangeably in line with UNESCO definitions, which distinguish tertiary (ISCED 5 and 6) from post-secondary non-tertiary (ISCED 4). 81 On occasion, there was political interference with admissions practices. 82 Outcomes determined by student choice can conflict with labor market demand if subsidies distort the relative costs of different types of courses, and boost private in comparison with social rates of return. 83 See Eurydice (2004a) for further discussion of the Bologna Process. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 81

158. By the middle of the 1990s, new forms of allocating finances to universities had emerged: block grants (the Czech Republic); normative financing, based on the number of students and on norms for research and maintenance (Hungary); and performance-related financing (Estonia), to name but a few. For example, by 1999 Hungary had merged many of its universities to achieve economies of scale and to improve quality, had developed an innovative income-contingent student loan scheme, and had begun to encourage inter-institutional competition through the application of rigorous investment evaluation criteria to all institutional investments. The economic reality of the need to attract more private financing into the system, to offset the limitations in state funding, had also become clear.

159. As tertiary enrollment rates increase, questions are arising about the proper extent of public funding at this level. In general, there is a case for continued public funding, but the taxpayer cannot be expected to pay the whole of the cost because: it is unaffordable, colliding with competing fiscal imperatives and to the likely detriment of quality; and it is unfair, because it is regressive – most of the beneficiaries of subsidized higher education come from families with incomes higher than those of the average taxpayers who provide the subsidies (Barr, 2004) – and because the returns are largely private.

160. Upfront charges (i.e. payable at the outset of or during the period of study) are also unfair because they deter students from poorer backgrounds. Thus, a system of deferred tuition fees (i.e. combined with a well-designed system of student loans), supported by a system of scholarships or grants targeted to those who need them, is an essential part of any strategy for expanding quantity, improving quality and achieving equitable access.

161. Developing an efficient and equitable financing system is not easy. Some new EU countries, however, have gone further in this direction than others, and one purpose of this paper is to review this experience. Because of the social benefits from higher education, as well as the political impossibility of eliminating all tuition subsidies, universities are likely to remain very substantial recipients of direct government grants. A mass system and increasing institutional differentiation require a wider variety of public allocation mechanisms. The range of mechanisms includes traditional line item budgeting, block grants, formula funding, and competitive and performance-based funding. Funds can also be channeled directly to students through scholarships, vouchers, and mortgage and income-contingent student loans. The rest of this chapter will discuss the growing experience with a variety of financing mechanisms in the EU8 countries, drawing on detailed country case studies, and will seek to develop some useful lessons of experience, mindful that each country will continue to develop its own solution based on national priorities.

162. Section 2 provides an overview of current higher education systems in the EU8 in a comparative perspective. Section 3 suggests some directions for further reform initiatives. Section 4 concludes.

2. Higher Education Systems in a Comparative Perspective

Scale and impact of Higher Education 163. EU8 countries have, as already emphasized, joined the headlong rush towards mass higher education (Figure III.1). The fastest rate of increase in enrollment rate over the whole 1990/1-2002/3 period was shown by Hungary, followed by the three Baltic countries and Poland. Expansion in Finland, France and Ireland, which started the period with relatively high enrollment rates, has lagged behind that in most of the EU8. 82

Figure III.1. Gross Enrollment Rates in Tertiary Education (EU8 & selected EU15 countries, 1990/91 – 2002/2003).

% 90 Finland

80 Latvia

Lithuania

70 Slovenia

Estonia 60 Poland

France 50 Hungary

Ireland 40 Czech Republic Slovakia 30

20

10

2 8 90/9191/9 92/9393/9494/9595/9696/9797/9 98/9999/0000/0101/0202/03

Source: Gross enrollment rates are from UNESCO Institute for Statistics, Global Education Digest 2005; attainment rates in column 3 are from EUROSTAT Labor Force Survey database; other data from EUROSTAT and UNESCO databases and OECD, Education at a Glance 2004.

164. By 2002, average enrollment in higher education in the EU8 countries was equivalent to well over one-half of the relevant age group: rates in the Baltic countries, Poland and Slovenia exceeded 60 percent, while the Czech Republic, Slovakia and (in spite of its fast expansion), Hungary were lagging. In many countries the number of part-time students has increased disproportionately: in Slovakia, for instance, they now account for a third of total enrollment. In 2002, the number of tertiary graduates per 1,000 people in their twenties was slightly higher in the EU8 than in the EU15 countries, with Lithuania and Poland leading the way, and the Czech Republic and Slovakia lagging. On average, the proportion of the 25-64 population with higher education is still larger in the EU15 countries than in the EU8, but Estonia and Lithuania have unusually highly educated populations. Higher education is slightly more feminized in the EU8 group, with the Baltics showing particularly large proportions of women among their students. Student/teacher ratios are similar on average in the two groups, but there are wide variations between the EU8 countries, from Estonia with very few students per teacher to Slovenia with very many. The proportion of students enrolled in science and technology courses (most of them more expensive than humanities courses), is smaller, on average, in the EU8 than in the EU15, but the two countries with the lowest enrollment rates, the Czech Republic and Slovakia, are ahead of the rest in this respect. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 83

Table III.1. Higher Education Indicators (EU15 and EU8 countries). Gross Tertiary % of 25- Females Student/ Students enrollmen graduates 64 as % of teacher enrolled in t rate, during populatio total, ratio, science, 2002/03 2002 per n with 2002 2002 mathematics, 1,000 higher computing, populatio education, engineering, n aged 2004 manufacturin 20-29 g & construction as % of total, 2002 EU15 58% 45 53% 16 27% of which: France 56% 68* 24% 55% 16 Ireland 50% 71 28% 55% 15 34% Finland 86% 58* 34% 54% 16 37% EU8 57% 48 16% 57% 17 23% Czech 36% 26 12% 51% 12 32% Republic Estonia 66% 41 31% 61% 9 21% Latvia 73% 59 20% 62% 21 17% Lithuania 72% 63 25% 60% 11 26% Hungary 51% 39 17% 55% 15 18% Poland 60% 75 15% 58% 20 21% Slovenia 68% 48 20% 58% 32 21% Slovakia 32% 31 13% 52% 12 28%

Notes: Females as % of total and % of 25-64 population with HE are weighted average for EU15 and EU8 (other averages unweighted); Denmark and Portugal missing from EU15 average for student/ teacher ratios. The gross enrollment rate (as a % of the population in the five-year age group following on from the secondary school leaving age) counts as enrolled all students (full-time and part-time) in tertiary education (ISCED 5 and 6), whatever their ages, as opposed to the net enrollment rate which counts as enrolled only those students in the specified age range. France, Ireland and Finland are arbitrarily chosen for inclusion in this and other Tables and Figures as examples of EU15 countries with different higher-education traditions. * 2001 Source: See Figure III.1.

165. Higher education institutions are important for the EU8 countries’ efforts to insert themselves into the global knowledge economy. On average, about 14 percent of 25-34 year olds in the EU8 labor force in 2003 consisted of tertiary graduates who were employed as professionals or technicians – a higher proportion than in the case of older age groups, although below the 18 percent average in the EU15 (Table III.2). Slovenia had a particularly high percentage in this category – comparable to Ireland. Most EU8 countries have been increasing the total number of graduates in these occupations at a faster rate than the EU15 average. Moreover, these are not old-style scientists and technologists: two thirds of them (not much lower than the EU15 average) were working in knowledge-intensive services. 84

Table III.2. Tertiary Graduates in Science and Technology Employed as Professionals or Technicians (2003).

As % of the labor force aged: Annual average rate % of professionals/ of growth, 2000-03 technicians in knowledge intensive services age group: 25-34 35-44 45-64 25-64 25-64

EU15 18.0%a 15.8% a 15.3% a 2.5% 68.9% a of which France 23.7% 14.9% 14.4% 2.6% 67.5% Ireland 20.1% 16.5% 14.5% 6.7% 75.8% Finland 26.1% 23.7% 20.2% ... 63.5% b b b EU8 13.5% 11.8% 12.0% … 65.0% a c Czech Republic 9.9% 11.3% 9.4% 3.1% 64.5% Estonia 14.7% 14.4% 16.7% -0.2% 66.2% Latvia 11.6% 9.5% 10.5% ... 58.0% Lithuania 14.2% 12.9% 13.4% ... 63.6% Hungary 13.6% 12.9% 13.7% 5.8% 70.4% Poland 14.9% 11.2% 9.5% 4.8% … Slovenia 19.0% 13.8% 13.1% 7.0% 57.8% Slovakia 10.0% 8.4% 10.0% 5.6% 62.3%

Notes: a weighted; b unweighted; c excluding Poland. Source: Eurostat (2004): Statistics in Focus: Science and Technology (November).

166. As might be expected, the earnings premium for higher education in the labor market is considerable, but seems to be higher in countries where it still has some scarcity value (the Czech Republic, Hungary) and lower in countries with a high proportion of graduates in their population and high enrollment rates (Lithuania) (Table III.3). Unemployment rates (similar on average in both groups of countries), vary between countries but are invariably lower for higher education graduates than for those with lower levels of education. Again, the impact of higher education (compared with those of upper secondary graduates) on unemployment rates tends to be greater in countries where fewer people have acquired it: among 15-39 year olds, for instance, the impact is the largest in the Czech Republic, and the smallest in Estonia. Except in Lithuania, female tertiary graduates in this age group have higher unemployment rates than males, but in several countries (France, Estonia, Lithuania, and Slovenia), women in this age group gain more in employability from higher education than do men. Overall, the private rate of return to higher education is still high in the EU8 and, given the pattern of subsidization discussed below, likely to be higher than the social rate of return.84

Table III.3. Higher Education and the Labor Market (EU8 and EU15 countries).

Earnings premium, Unemployment rate, 15+, Unemployment rate, 15-39, 25-64 year olds – 2004 q2 2004 q2 HE over upper secondary, 2002 Higher Upper 2ndary Higher Upper 2ndary education education Both Females Both Females Both Females Both Females Both Females sexes sexes sexes sexes sexes EU 15 43% 46% of which: France 50% 46% 6% 6% 8% 10% 7% 7% 10% 13% Ireland 49% 61% 2% 2% 4% 4% 3% 3% 5% 4% Finland 50% 46% 5% 5% 10% 10% 6% 7% 12% 13% EU 8 5% 5% 11% 12% 6% 7% 13% 15% Czech 79% 70% 2% 2% 7% 9% 2% 3% 9% 11% Republic Estonia 6% 6% 11% 10% 8% 9% 13% 14% Latvia 5% 5% 12% 11% 6% 6% 12% 13% Lithuania 46% 52% 7% 7% 13% 14% 7% 5% 13% 15% Hungary 110% 79% 2% 3% 5% 6% 3% 4% 7% 8% Poland 61% 7% 8% 20% 22% 10% 11% 24% 27% Slovenia 74% 66% 3% 3% 6% 7% 4% 5% 8% 10% Slovakia 6% 6% 17% 18% 8% 10% 19% 21%

Note: Latvia unemployment rates for Q1-2004; Estonia and Latvia unemployment rates for 15-39 year old females with higher education are estimated. Greece, Luxembourg, and Austria are missing from the earnings premium average. Source: OECD (2004); and EUROSTAT Labor Force Survey database.

84 Both private and social rates of return are based on earnings as a measure of benefit, but after-tax in the case of private and before- tax in the case of social (in principle, social benefits should include indirect benefits not captured by earnings). On the cost side, private rates of return are based on costs to the individual, and social returns based on costs including government subsidies. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 85 Financing Higher Education 167. Most EU8 countries have managed to protect the percentage of GDP that is devoted to public expenditure on education (slightly higher on average than in the EU15), but with variations between countries (the Czech Republic and Slovakia are below the average, though not far below the share achieved by Ireland) (Table III.4). Real public expenditure on education has increased faster in the EU8 than in the EU15 since 1995, with Lithuania leading the way (exceeding Ireland’s rate of increase), and the Czech Republic and Slovakia again lagging behind. The two groups spend the same proportion of GDP on higher education (1.1 percent) but with variations: no EU8 country approaches the 2.1 per cent of GDP devoted by Finland for this purpose. There is less variation in the proportion of the education budget going to higher education – from 17 to 23 percent: Ireland and Finland allocate much larger shares to higher education than any EU8 country.

168. The challenge facing the EU8 countries in integrating into the EU higher education system is shown by the fact that, while on average they spend a much higher proportion of GDP per head per student in public tertiary institutions than do EU15 countries, and a much higher multiple of expenditure per primary school student, the number of Euros spent per tertiary student, adjusted for differences in purchasing power, is much lower than in the EU15 group (42 percent lower on average). Such cross-country comparisons between countries with different salary levels, even when adjusted for purchasing-power differences, are fraught with difficulties, but the implication that an increase in unit expenditure will be needed for integration purposes is probably correct. Tertiary institutions in the EU8 spend a lower percentage of their budgets on research and development than do their counterparts in the EU15, but even if this is taken into account, the contrast in unit educational spending remains (OECD, 2004). Significantly, expenditure per head seems to vary inversely with enrollment rates, being lowest in Lithuania and Poland and highest in Hungary and the Czech Republic, which suggests at least the possibility of a trade-off between accessibility and quality of higher education from the student’s point of view.

Table III.4. Public Financing of Higher Education (EU8 and EU15 countries, 2001).

Total Change in Total Tertiary Expend- Ratio of Expendi- public total real public as % of iture per tertiary to ture per expendi- public expendi- total student in primary ex- student in ture on expendi- ture on public public penditure public education ture on tertiary education tertiary per tertiary as % of education, education expendi- institutions student in institutions GDP 1995-2001 as % of ture (EUR PPS) public as % of GDP institutions GDP per head

EU15 5.1% 14% 1.1% 21% 8426 1.9 38% of which: France 5.8% 12% 1.0% 18% 8043 1.8 33% Ireland 4.4% 45% 1.2% 29% 9282 2.7 35% Finland 6.2% 16% 2.1% 33% 9069 2.1 39% EU8 5.3% 25% 1.1% 20% 4877 2.8 48% Czech Republic 4.2% -1% 0.8% 19% 5431 3.3 40% Estonia 5.5% 31% 1.1% 20% 5143 3.1 59% Latvia 5.8% 29% 1.0% 17% Lithuania 5.9% 52% 1.3% 23% 3274 1.7a 39% Hungary 5.2% 21% 1.1% 22% 6942 3.0 60% Poland 5.6% 40% 1.1% 19% 3582 1.6 38% Slovenia 6.1% 1.3% 22% Slovakia 4.0% 1% 0.8% 21% 4891 4.2 49%

Note: a ratio of tertiary to general (primary + secondary) expenditure per student. EU8 averages are unweighted. Source: Eurostat website.

169. Although higher education students tend to be disproportionately from more prosperous families, there is a tendency in European countries for the public to support them through financial aid.85 Financial support to students is strongly biased toward higher education (Table III.5). In general, this bias was higher

85 Defined by Eurostat as “transfers from the public sector to students in the form of grants, loans and child allowances, contingent on their status as students.” 86 in the EU15 countries than in the EU8: while both groups spent a comparable proportion of their education budget on financial aid to students, the NMS allocated a much larger share to primary/secondary schools than did the existing members. However, there were wide variations between EU8 countries. Poland stands out as a country that spent hardly any of its education budget for this purpose. Estonia spent a higher proportion on financial aid to students at lower levels than at the tertiary level. Latvia (where about 16 percent of publicly funded students receive grants awarded on the basis of academic merit), Slovenia, and Hungary were especially generous to their higher-education students, but the biggest contrast between primary/secondary and tertiary patterns of spending in this respect was in Slovakia, where as much as 10 percent of full-time higher education students received scholarships in 2003. Within the EU15, the examples of France, Ireland and Finland show the wide variation in the models used by different countries.

Table III.5. Financial Aid to Students (EU8 and EU15 countries, 2001).

Financial aid to students as % of total public expenditure on education All levels Primary & secondary schools Higher education institutions EU15 5.0% 2.8% 13.8% of which: France 3.9% 3.4% 8.4% Ireland 5.8% 3.3% 11.9% Finland 8.1% 3.4% 18.2% EU8 6.6% 5.8% 12.8% Czech Republic 5.6% 5.9% 7.9% Estonia 5.4% 6.3% 2.8% Latvia 9.2% 7.1% 23.5% Lithuania 6.2% 5.5% 11.9% Hungary 10.5% 10.1% 19.5% Poland 0.5% 0.5% 0.4% Slovenia 11.6% 8.7% 25.6% Slovakia 3.6% 2.0% 10.5%

Note: Table includes EUROSTAT estimates. Source: EUROSTAT database.

170. In most EU15 countries, higher education is funded almost entirely from public sources – the UK is the only country with more than 25 percent private funding (including fees and donations); Italy, the Netherlands, and Spain the only others with more than 20 percent. Data are available for only three EU8 countries: in Hungary, private funding accounted for 22 percent of total expenditure on tertiary education in 2001, while the Czech Republic (with 15 percent private funding), and Slovakia (with only 7 percent), are more in line with the EU15 norm.

171. Only four of the eight countries – Estonia, Hungary, Latvia, and Poland – have more than 14 percent of their university students in private institutions; in the other four countries, the proportions are negligible (Figure III.2). In Latvia, the number of private higher education institutions increased tenfold to 20 between 1991 and 2004, but on average they are much smaller than their 36 public counterparts.

Figure III.2. Share of University Students in Public and Private Institutions (EU8 countries, 2002).

Slovenia

Slovakia

Poland

Lithuania Public

Private Latvia

Hungary

Estonia

Czech Republic

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Source: UNICEF: TransMONEE database. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 87

172. The small role of fees in most EU8 countries does not mean that parents and students avoid paying for higher education. For one thing, many countries have instituted a dual system, maintaining fee-free higher education for regularly admitted state-supported students, while adding a special fee-paying track for those who fail to gain such admission. Institutions have also imposed and increased user charges for formerly heavily- subsidized board and residence facilities, which are borne by students or their parents, as are the costs of living for those who do not make use of such facilities. Also, although a significant proportion of educational budgets still goes to stipends, maintenance grants or other types of financial aid to students, they have been eroded by inflation to virtually nominal levels in many EU8 countries (Johnstone, 2004).

173. There are large differences in the higher education expenses borne by parents and students in public institutions, between the more and the less fortunate categories of student in some of the EU8 countries that have introduced a dual-fee-track system (Table III.6). In Poland, although the constitution guarantees free higher education to all who achieve entry level qualifications, over 50 percent of all students pay some form of tuition fees (when the numbers of extra-mural, part-time or evening students who can be charged tuition fees are added to those attending private HEIs) (World Bank and EIB, 2004). In Hungary, state-funded full-time students pay no fees and receive subsidies for books and living costs, while those who are not state-funded (for various reasons, including below-average scores in entrance examinations) pay tuition fees and all their own expenses. In Latvia, state-financed places are rationed, and those who fail to get one (again on the basis of academic merit), have to pay tuition fees. Latvia has significantly increased the proportion of students in the fee-paying track, from 32 percent of the total in 1995/6 to 77 percent in 2004/5, and tuition fees accounted for 31 percent of the income of public higher education institutions in 2002.

174. Among other countries with a dual-track system, higher education institutions in Slovakia are allowed to charge fees only for a limited range of services, and full-time students do not pay tuition fees (unless the study time exceeds the standard length). However, many institutions collect fees illegally from part-time students. In the Czech Republic, students in public higher education institutions do not pay fees for study programs provided in the Czech language, unless they have exceeded the standard length of study by more than one year, or are enrolled in more than one program at the same time. Tertiary professional schools (ISCED 5B) do, however, charge tuition fees (ranging from €90 to €1,100 per year, depending on the ownership of the school: regional; church; or private). In Hungary, students in public universities and colleges who have achieved a minimum level in their secondary school-leaving examination do not pay tuition fees in most fields of study. In Lithuania, only half of the full-time students in universities and 20 percent in post-secondary colleges pay fees; the rest are totally funded from the state budget. Fees are set at a nominal level and are equivalent to less than 10 percent of the national budget expenditure per full- time university student (and less than 6 percent in the case of colleges). Exemption from fees is awarded to students based on their academic performance. Full-cost fees are paid only by extra-mural, evening, and postgraduate students. In Slovenia, all full-time undergraduates are financed by the state: only part-time undergraduate and all postgraduate students pay tuition fees.

175. The dual-track system means that there are two classes of students in higher education institutions – those whose instructional expenses are nominal, and those who pay fees equivalent to several thousands of dollars. If student living expenses are included, the cost of putting a fee-paying student through higher education looks prohibitive.86 Those who go to private institutions can pay even higher tuition fees – ranging from $3,500 to $8,000 in Poland, and from $1,400 to $16,700 in Latvia in 1998/99 and 2003/04, respectively (adjusted for differences in purchasing power).

86 Even in the case of non-fee-payers, student/parent expenses are considerable. 88

Table III.6. Higher Education Expenses Borne by Parents and Students (first degree, public institutions: Poland, Hungary and Latvia; US$ PPP).

Poland 1998/9 Hungary 2000/1 Latvia 2003/4 Moderate High State- Non- Low Moderate High funded state- students funded students Instructional expenses Tuition fee 0 3696 0 2400 0 1666 10095 Other fees 0 0 50 530 71 48 95 Books etc. 271 271 0 … 24 33 52 Sub-total 271 3967 50 2930 95 1747 10242 Student living expenses Lodging 1359 1630 750 1800 0 952 11428 Food 2446 3261 750 1200 2000 2857 8571 Transportation 163 163 300 180 743 1714 4571 Other personal expenses 434 516 600 1800 2286 2857 6743 Sub-total 4402 5571 2400 4980 5449 8380 32213 Total cost to parents & 4673 9538 2450 7910 5544 10127 42455 student

Notes: In Poland, “moderate” = living in a dormitory or shared apartment; “high” = extra-mural, part-time or evening student, living as an independent adult. In Latvia, “low” = enrolled in a state-financed slot, living with parents; “moderate” = moderate public tuition fee, living in a dormitory; “high” = high public tuition fee, living as an independent adult. Source: The International Comparative Higher Education Finance and Accountability Project, Graduate School of Education, State University of New York.

176. An anomaly associated with the dual-fee-track system is that it tends to penalize students from disadvantaged families. Those who obtain fee-free, state-subsidized places are disproportionately from privileged backgrounds (which have contributed to their academic success); poorer students, who are less successful in entrance examinations and cannot afford the alternative fee-paying track, are excluded from higher education. Throughout the OECD, participation rates of 18-24 year-olds in higher education vary strongly and directly with parents’ levels of education (Blöndal et al., 2002). In Poland, for example, there is a 17-percentage-point difference between the education enrollment rates of 19-24 year olds of the richest and the poorest consumption quintile (Table III.7). This probably understates the influence of differences in family backgrounds, since students living away from home in bed-sitting rooms are counted as one-person households (with presumably a relatively low budget).

Table III.7. Poland: Enrollment Rate by Age Group and Household Consumption Quintile (2003).

Age group: 3-6 7-10 11-15 16-18 19-25 & quintile: I 4.6% 94.7% 99.6% 90.9% 36.2% II 3.8% 95.0% 99.7% 96.1% 46.6% III 4.2% 95.2% 99.9% 97.2% 52.6% IV 2.0% 94.1% 99.4% 97.9% 55.0% V 3.3% 96.4% 100.0% 97.3% 52.6% Total 4.1% 94.9% 99.7% 94.7% 47.4%

Source: Household Budget Survey database.

177. The influence of social class on access to higher education may be even stronger in the EU8 than in other industrialized countries (Figure III.3). In Poland and Hungary, for instance, a 20-35 year-old in 1998 whose father had a professional occupation was four times as likely as a comparable child of a manual worker to have gained access to higher education, whereas in the USA, the ratio between the chances of the two categories was only two. Moreover, there is strong preliminary evidence for the hypothesis that the post-communist transformation brought an increase in inequality in access to tertiary education (Mateju et. al., 2004). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 89

Figure III.3. Chances of Obtaining Higher Education by Family Background (indicated by father’s occupation: the Czech Republic, Poland, Hungary, Finland, and the USA, 1998; ratios for individuals aged 20-35). 4.5

4.0 Professional vs. manual 3.5 Professional vs. nonmanual Nonmanual vs. manual

3.0

2.5

2.0

1.5

1.0

0.5

0.0 CZR Poland Hungary Finland USA Source: Mateju et. al. (2004).

178. In order to facilitate entry to higher education for students who need financial support to pay fees and/or maintenance costs, several EU8 countries have instituted loan systems: In Latvia, students enrolled in private as well as public higher education institutions are eligible for two types of loans, one covering tuition fees, the other living expenses. The number of loans in 2004 was over 55 thousand compared with less than seven thousand six years earlier. The interest rate is kept to 5 percent by a government subsidy covering the difference from the rate charged by the commercial bank which administers the program. There is a year-long grace period after graduation, and those who get specified jobs in the public sector have their loan repaid by their employer at a rate of 10 percent per year. Other concessions include a 30 percent reduction in debt for every child born or adopted. In addition to the government’s program, some commercial banks have started issuing guaranteed loans to students, sometimes in cooperation with higher education institutions. In Lithuania, there is an official student loan scheme, but it covers only a very small proportion of full- time students, probably less than five percent. In Poland, the coverage in 2002/03 was about 11 percent of all students. In April 2001, Hungary introduced an income contingent student loan scheme (whereby a graduate only makes a repayment if his or her income exceeds a designated level), thus providing an important instrument for the support of further institutional and policy change in higher education. By September 2001, all higher education students enrolling in Hungary were eligible to apply for loans, and in the intervening years interest in this loan scheme has been impressive. In Slovakia, there has also been a proposal to extend the system of student loans to cover tuition fees, repayable after a student finishes studying and reaches at least the minimum wage, in addition to the loans for maintenance that are already available.

179. EU8 countries are also experimenting with the use of formulae to determine the amount of public funding to be allocated to each higher education institution: In Slovenia, state funding of undergraduate studies in any institution has two elements: basic annual funds (a lump sum proportionate to the previous year’s allocation, which in 2004 was 80 percent of the annual funds allocated in 2003); and standard annual funds which depend on the number of students and the number of graduates with differential weighting of subject areas. Public co-financing of postgraduate studies and funding of research are distributed by tender. In Slovakia, the allocation for wages and salaries is based on the number of students and the number of graduates (varying between fields in line with differences in student/ teacher ratios and faculty 90

structure), and on the results of research. Allocations for research partly reflect an institution’s success in obtaining research grants, and the success rate of doctoral students in qualifying examinations. State funding for goods and services, scholarships for graduate students, capital expenditure, student stipends, dining, lodging, sport, cultural and religious activities, and development and reform of higher education are allocated separately. The Czech Republic is planning to introduce a new funding mechanism, with a formula based at least partly on output rather than input indicators, supplemented by an agreement on the longer term plans of each institution. This is expected to lead to a division of labor between different types of higher education institutions/faculties, with some basing their study programs mainly on top-quality research, some combining such research with teaching, and others focusing primarily on teaching. In Hungary, the allocation to each higher education institution is based on the number of full-time- equivalent students (and their fields of study), the number of lecturers/ researchers, and of state- financed PhD students (and their fields of study), the amount of research support in the previous year, and the number of non-teaching staff and maintenance norms for such staff and students. Efforts are being made to move towards a more results-based allocation mechanism, which would encourage institutions to use resources more efficiently. Latvia has had a formula funding system since 2002, with allocations based on the number of approved full-time places and the unit cost. The unit cost for each subject is calculated as a multiple of the “base cost” per student, which reflects wages, social insurance costs, costs of utilities, etc., and rose from 440 lats in 2002 to 696 lats in 2005. The indices used for each subject are being gradually adjusted to reflect differences in costs from year to year. Research is funded partly by tender and partly by grants to institutions: each institution receives an amount equivalent to at least 1 percent of the funds it is allocated for undergraduate and postgraduate studies, plus funds for research infrastructure based on the number of approved student places and faculty posts. In Lithuania, experimental “student’s basket” systems were introduced for colleges and universities in 2004, with allocations partly based on the number of students and type of program. This was a welcome step towards transparency, but the logic of the formulae was not allowed to prevail: rectors at risk of losing resources successfully lobbied parliament to change the results. Poland in 2004 switched back to using the funding algorithm that had been introduced in the early 1990s, designed to promote the expansion of enrollment and to encourage HEIs to employ greater numbers of professors with doctorates. The new version of the algorithm addresses issues of quality, and includes criteria that are based on the National Accreditation Commission evaluations. The formula takes into account variations in unit cost between programs, numbers of students (including PhD students), and numbers of faculty in both research and teaching with PhDs. Governance 180. A modern system of tertiary education, comprising many autonomous institutions with differing but complementary missions and goals, requires flexible governance on a national level and an enhanced policy function of the Ministry of Education. Autonomous higher education institutions should be set within a framework of accountability, openness and segregation of the regulatory and operational management functions.

181. Throughout the EU8 countries, changes in financing arrangements have, as expected, been accompanied by changes in the governance of higher education institutions. In most EU8 countries, the lower tier of tertiary institutions (ISCED 5) are subject to greater government control than universities, many of which match their counterparts in the EU15 in the extent of their autonomy (Table III.8). The Czech Republic, Latvia and Poland appear to be the countries in which universities have the greatest freedom, in contrast to Lithuania and Slovenia in particular. The most widely available freedom is to employ and dismiss academic staff. Also, most universities can set their academic structure and course content, and spend their budgets to achieve set objectives. Less widely available is the authority to set salaries, decide on the size of student enrollment, own buildings and equipment, borrow funds, and decide the level of tuition fees (allowed in Hungary alone). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 91

Table III.8. Extent of Autonomy Enjoyed by Universities (selected EU15 and EU8 countries).

Institutions are free to: 1 2 3 4 5 6 7 8 Own their Borrow Spend Set Employ & Set Decide size Decide buildings & funds budgets academic dismiss salariesb of student level of equipment to achieve structure/ academic enrollment tuition objectives course staffa fees content EU15 Netherlands ● ● ● ○ ● ● ● ○ Ireland ● ○ ● ● ● ○ ● ○ UK ● ○ ● ● ● ● ○ ○ Denmark ○ ● ● ○ ● ○ ● ○ Sweden ○ ○ ● ● ● ● ○ Finland ○ ● ○ ● ● ○ Austria ○ ● ● ● ● EU8 Czech Rep ● ● ● ● ● ● ● Estonia Latvia ● ● ● ● ● ● ● ○ Lithuania ○ ○ ● ● ● ○ ○ ○ Hungary ○ ○ ● ● ● ● ○ ● Poland ● ● ● ● ● ○ ● ○ Slovenia ○ ○ ● ● ○ ○ ○ Slovakia ● ● ○ ● ● ● ● Institutions have autonomy. ○ Institutions have some autonomy, but limited. a “Employ and dismiss academic staff” (column 5) and “Set salaries” (column 6) include cases where any legal requirements for minimum qualifications and minimum salaries have to be met. b “Decide size of student enrolment” (column 7) includes cases where some departments or study fields have limits on the number of students able to enroll. c Public universities. Source: Interviews for EU8; OECD (2003) for EU15.

182. An important aspect of university governance is the arrangement for appointing leaders.In all the EU8 countries for which information is available, leaders of universities are still elected by bodies consisting of representatives of academic staff and, in some cases, students (Table III.9). Only in Hungary and Latvia are such appointments subject to governmental approval, and in all cases, they are renewable for at least one more term. There is no mechanical connection between the method of selecting a university leader and the propensity to reform, but academics tend to elect rectors who will look after them rather than cause them trouble. The trend in the EU15 is clearly towards a redefinition of the functions and composition of university Boards or Councils, with a greater management role for a mix of academics and outside people. Such Boards then take over the function of appointing university leaders from shortlists that emerge from a wide-ranging search. A move in this direction would help EU8 universities to implement the financial reforms that will be needed in the challenging times ahead. 92

Table III.9. Appointment of Leaders of Universities (selected EU15 and EU8 countries).

Process for election or appointment Government Typically Renewable position? has to appointed for approve? how many years? Countries where leaders are usually ELECTED by: Lithuania Senate (academic staff). No. 4-5 Yes, for one more Finland Academic staff & heads of separate No. 5 round. institutes. Yes. France Board or Council. No. 5 Czech Academic Senate (academic staff No. 3 No. Republic and students representatives). Yes, for two consecutive periods, Hungary Senate (academic staff and Yes. 4 with possibility of Latvia students). Yes. 5 later reelection. Constitutional Meeting (academic Yes, for one more Poland staff 60%, other staff 20%, and No. 3 round. students 20%). Yes, for one more Slovakia Academic Senate or Electoral No. 4 round. College. Slovenia No. 4 Yes, for a maximum Academic Senate. of two consecutive periods. All higher education faculty, faculty Yes, for one more assistant, researchers employed by round. the university, and students 20% of all votes. Yes. Countries where leaders are usually APPOINTED by: Ireland Governing Body (approximately No. 10 No. 50% external). Netherlands Supervisory Board: 5 external No. 4 Yes. members appointed by Minister. Sweden Government, on recommendation Yes. 6 Yes, for two periods of mainly external Governing Board, of three years. which first consults students & UK employers. No. 7 Yes. Governing Body, of which majority are external members.

Countries where reforms have been implemented in 2003: Austria Formerly elected by University No. 4 Yes. Assembly (75% staff, 25% students). From 2003 appointed by University Council, made up of external members, from a shortlist of 3 Denmark candidates nominated by Senate. No. 4 Yes. Formerly elected by academic staff (50%), other staff (25%), and students (25%) From July 2003, appointed by Board with majority of external members.

Source: for EU8, interviews/questionnaires; for EU15 countries, OECD (2003). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 93 3. Directions for Further Reform 183. The need for further reforms in higher education finance in the EU8 countries arise from a combination of factors: the trend throughout Europe towards mass participation in education at this level, reflected in the enrollment rates in Table III.1 above; the pressure for improving the quality and increasing the relevance of higher education in the EU8 countries, associated with membership of the EU and participation in the Bologna process and the European Credit Transfer System (Eurydice op. cit.), and with globalization, which probably implies a need to increase expenditure per student towards EU15 levels; the tight overall fiscal situation faced by EU governments trying to limit the size of their budget deficits; concerns about equity given that private rates of return on higher education are likely to be much greater than the social rate of return (Psacharopoulos and Patrinos, 2002) and that higher education students are still disproportionately from privileged backgrounds.87

184. As has already emerged from the discussion of the current situation, a useful starting point for programs of further reform of higher education finance in EU8 countries would be a thorough review of the extent and incidence of subsidies to higher education. Subsidies take several forms: tuition fees at less than full cost-recovery levels, or exemptions from tuition fees; stipends, scholarships or maintenance grants, and subsidized loans; subsidized food and lodging; and subsidized books and other learning materials.

185. As already emphasized, the imposition of fees is a recognition of the private benefit to recipients of higher education, but the level at which tuition fees are set is clearly a political as well as an economic issue. In some countries, to impose any tuition fees at all would require a constitutional change.88 Generally, the flawed tradition in Europe of regarding higher education as entirely a public good is still strong, and there is a general refusal to acknowledge the limitations on public revenue caused by rising costs and enrollments. In many such countries in the EU15, controversy about the possibility of introducing fees still rages. In Ireland, an attempt to reintroduce tuition fees in 2003 was abandoned after public protests rendered it politically impossible. In 2004, the OECD noted the need for greater investment in tertiary education to improve Ireland’s competitiveness, including some cost-sharing arrangement with users. In 2005, while the Minister of Education broadly accepted the OECD recommendations, undergraduate tuition fees are still not politically feasible in Ireland. The UK introduced tuition fees some years ago in conjunction with a student loan scheme and has recently increased their level. However, this process has been fraught with difficulty and is a perpetual subject of acrimonious debate in the media. In May 2005, the Slovak parliament refused to ratify a government proposal to introduce a new system of tuition fees (equivalent to between 5 and 30 percent of unit expenditure) at specific levels to be determined by the institutions.

186. In the EU8, however, the pressures outlined above make it almost inevitable that the average level of fees in tertiary institutions will be increased, to a level which covers a significant proportion of total costs. Should such fees be fixed and uniform, or variable (by institution and by specialization)? Barr (2004) argues strongly in favor of variable fees on various grounds: variable fees bring in more resources; they promote efficiency, for instance guiding students to choose cheaper ways of acquiring the skills they want, or to choose courses at a cheaper local college instead of a more expensive and more prestigious university – to the benefit of both student and taxpayer; they are an incentive for institutions to improve quality (and hence charge higher fees), and to increase efficiency (and maybe attract more students by charging lower fees); they give institutions control over their own income stream instead of being dependent on what the government is currently able to allocate; and they are more equitable (as part of a package), introducing higher charges for those who want to pay them, helping poor people to pay those charges through redistributive policies, and ensuring some correspondence between what individuals pay, and what they get (instead of, as Barr puts it, “taxing beer to subsidize champagne”).

87 Work on promoting equitable access must, of course, start well before higher education, given the critical importance of early child development (Feinstein 2003). 88 The German constitutional court recently declared that tuition fees in higher education do not violate constitutional guarantees. 94

187. An increase in average fee levels would allow the inequitable dual-track system to give way to a single-track system, under which all students pay tuition fees, and to a distribution of the various subsidies that are based on need, rather than academic ability. Institutions could use their own resources to offer a few competitive scholarships to their best applicants, but state subsidies to students (most appropriately handled by social service ministries rather than educational institutions) would be based only on means-tests. This raises the question of the feasibility of means-testing in current EU8 country conditions. The technical pre-conditions for successful means-testing include “pervasive and generally workable income tax systems with high degrees of voluntary compliance that capture most sources of taxable income, and allow reasonably reliable calculation and monitoring of family means, or need” (Johnstone, 2004). As part of the process of integration with the EU, the EU8 countries are moving towards such systems. For example, Slovakia was proposing to extend means-tested social stipends in scope and amount, to cover at least a third of full-time students. Progress will vary from country to country, but meanwhile, existing relatively informal systems to estimate means or need can be built on, “with the onus of demonstrating financial need placed upon the family that is claiming it, and with clear penalties for misrepresentation.”

188. Means-tested grants to higher education students are, at best, a blunt instrument and need to be backed up by other measures to promote equity in access. Barr (2004) suggests a means-tested stipend for children above the minimum school-leaving age to encourage them to complete school, and financial incentives to universities to widen participation (which could be embodied in the funding formula). He also calls for action to inform school children about the benefits of higher education and to raise their aspirations: “the saddest impediment to access is someone who has never even thought of going to university.”

189. While students from disadvantaged families could thus be subsidized, the extension of fees to all students and an increase in their average level would imply a need for a wide-ranging student loan system. This could take the form of a loan, either repayable in the conventional way or (as implemented by Sweden and the UK, among EU15 countries), on an income-contingent basis, or a graduate tax whereby a student incurs an obligation to pay a higher rate of income tax.89 Barr (2004) argues in favor of income- contingent loans charging “unsubsidized” interest rates (broadly in line with the government’s cost of borrowing). Income-contingent repayments are calculated as a given percentage of the borrower’s earnings until the borrower has repaid: low-income earners make low repayments and those with low lifetime earnings never fully repay. Those who take out loans effectively get their higher education free at the point of use, repaying only later if and when they can afford to do so. Thus, even the most risk- averse students (often assumed to come disproportionately from low-income backgrounds), may be willing to borrow in order to finance their studies. Interest rate subsidies add a lot to the costs of a loan scheme, divert funding from quality improvement to student support, and are deeply regressive (mainly benefiting successful professionals in mid-career whose loan repayments are switched off early because of the subsidy).

190. Whatever the exact form of the loan system, Johnstone (2004) suggests several technical pre- conditions for success: ways to keep track of people’s movements, including a wide-ranging postal system with “skip-tracing” capabilities, and official and enforced employee identification; systems of withholding at the point of wage and salary payment that can facilitate student loan repayments; and effective systems of government guarantees, and primary and secondary private capital markets, that together allow private savings to supplement public revenue (thus providing a real alternative to dependence on tax revenue).

The first two pre-conditions pose fewer problems for EU8 countries than the third: very few countries (apart from the USA) have private capital markets for student loan notes. However, the absence of appropriate

89 In a progressive tax system, higher education graduates who achieve higher incomes pay higher taxes. It is sometimes suggested that, in this way, they repay the cost of their higher education, so fees plus loans are not necessary. As Barr (2004) points out, this is to some extent a fallacy: two people with the same income, one with and one without higher education, pay the same amount of tax, but if part of the tax payment of the graduate is covering the cost of higher education, he/she is contributing less than his or her less educated counterpart to the school, healthcare systems, and other tax-funded services. Emigration of graduates is another concern. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 95 capital markets is not an insurmountable obstacle to the creation of successful loan schemes. It is possible to attract private resources to such schemes from multinational banks. It is useful, also, to distinguish between the fiscal costs of a loan scheme (i.e. money that is lent out but never repaid), and its cash flow costs (i.e. money which is lent out and subsequently repaid). Cash flow costs matter for the EU Stability and Growth Pact (and this is why it is important to bring in private resources, if possible), but the fiscal costs of a loan scheme matter much more. The lower the fiscal costs of a loan scheme, the less critical is private finance.

191. Only four EU8 countries allow a significant role for private tertiary institutions. Their role is likely to increase, particularly if admission to public universities and colleges is held back by fiscal constraints. The ownership of higher education institutions should not be an important issue; they can be public, private or mixed, as long as there is a funding and regulatory framework which operates smoothly across the entire sector. Students should be eligible for government subsidies whether they are attending a public or a private institution. At the same time, it is the role of the government to set and monitor quality standards for all tertiary institutions, regardless of ownership.

192. A potentially important source of extra-budgetary resources for tertiary institutions is paid teaching, research, and consultancy. Arrangements have to be set up for channeling the income from such activities to the institutions, rather than merely to individual faculty members. The danger of increasing inter- departmental disparities in resources available also has to be recognized. Departments of economics, business studies, foreign languages, and information technology are likely to be the main beneficiaries; other arts and science departments may not be able to attract as many resources in this way. In Latvia, the proportion of expenditure on higher education financed from sources other than the state budget and tuition fees (i.e. from scientific contracts, international agreements and commercial services), rose from 1.9 percent in 1995 to 11.5 percent in 2003.

193. The criteria for determining the amount of public money to be allocated to each institution are also in need of further reform. In general, as already described above, the desire to get away from politicized and opaque negotiation between the government and each institution is inspiring a move towards formula funding, whereby the amount allocated is determined by a transparent formula. The simplest formula for funding, adopted in several countries, is to base it on the number of students enrolled – “money follows the student.” This can lead to distortions, such as expanding enrollment beyond the optimum level, delays in student completion, and a bias towards programs that are popular and/or low-cost. There is growing interest, accordingly, in more complex, performance budgeting or results-based funding.

194. Results-based funding has to strike a balance between manageability and comprehensiveness. Indicators used have to be easy to calculate, difficult to manipulate, reliable as a guide to an institution’s value added, and not subject to statistical “noise” (Thorn et. al., 2004). Several EU15 countries have experimented with systems of this kind. In the UK, allocations are based on two separate sets of indicators: a teaching grant, based on number of students, length of the courses, size of the institution, location, level of specialization, number of disadvantaged students, etc.; and a research grant, based on the number of qualified researchers, published research documents, the number of research students, the quality of research assessed by peers, external research income etc. The Danish “taximeter” model bases the amount institutions receive for their teaching activities on the number of students who pass their examinations. The tariff per passed examination varies according to the field of study, with three components: costs of education and equipment; joint costs (e.g. administration, buildings); and expenses for experimental sciences and practical training (e.g. in medicine and physics). In the Netherlands, a macro-budget for the university sector is fixed and then allocated to each institution on the basis of the number of students who 1 completed their program multiplied by the normal study duration (4 /2 years) plus the number of students who dropped out multiplied by a notional study duration (of only 1.35 years). This number is calculated for each field of study and then multiplied by a fixed amount per student (e.g. €4.477 for engineering and €3,461 for other programs in 2001). All such systems have the advantage of transparency in funding criteria, but run the risk of unintended consequences, for instance: overemphasis on publication in the case of the UK, artificial boosting of pass rates in Denmark, and the premature dismissal of struggling students in the Netherlands. None of them as yet has found a way of rewarding quality of teaching.

195. In general, further experimentation with more sophisticated, results-based formulae can be expected in the EU8 countries over the next few years, to the benefit of transparency of funding and efficiency in 96 higher education. Experience worldwide has shown that a certain variety in public funding mechanisms is desirable in higher education for several reasons (World Bank, 2005a). First, such variety provides the government with multiple ways to support investment in evolving public priorities. For example, financial awards can be modified to encourage institutions to invest in important public goods or to support public objectives such as the survival of areas of study which might not have immediate benefits to the labor market in the short term, but are vital for long-term development of a country. Second, variety in funding mechanisms provides the government with multiple processes that can stimulate favorable organizational behaviors at both the central and institutional levels (e.g. collaboration, transparency, accountability, stakeholder inclusion, etc.). Third, certain funding mechanisms can be used by governments to reward institutions for quality outputs, or to signal to stakeholders the quality of an institution’s outputs. Fourth, variety in funding mechanisms prevents budget instability at the institutional level which can occur when governments tie too large a proportion of institutional budgets to only one sort of resource-distribution procedure.

196. It is in the government’s interest to develop stable, more sustainable public institutions, infusing public higher education with a variety of effective and efficient funding mechanisms is one way to support that objective. There is no ideal mix, and each country must determine on its own which financing mechanisms and what combination of mechanisms work best in the given context. Innovations should be prudently and gradually introduced to a higher education system to provide opportunities to assess the impact of the new mechanism, to learn lessons from the initial implementation, and – based on such lessons – to make adjustments to the operational aspects of implementation before scaling up. Ultimately, however, oversight will have to be exercised by the ministry of education or some other central body to ensure that programs and mechanisms do not conflict or generate perverse incentives.

197. Finally, further governance reforms would also help to increase efficiency. Part of the motive for delegation of authority to autonomous institutions is to allow them to manage their use of resources. Accordingly, budgets for public universities in European countries are increasingly lump-sum: it is then up to management to decide on the allocation of the funds between line items, mid-year interchange between line items and, if necessary, carrying forward of expenditures from one year to another. There are dangers in this system, particularly if university leaders are elected by faculty members rather than appointed by a Board or Council with significant representation from the outside world. Links between the university and the private sector are desirable, not only to encourage labor market relevance but also to attract private funds through endowments and other mechanisms. Thus, it is likely to be efficient to increase the financial autonomy of higher education institutions, but only if institutional management systems are reformed to make them more sensitive to the public interest, rather than to that of the faculty and other politically powerful constituencies.

4. Conclusions 198. The EU8 countries have enthusiastically embraced mass higher education, and are rapidly closing the gap between themselves and the EU15 in enrollment rates and the incidence of higher education in the population of working age. This expansion is helping these countries to insert themselves into the global knowledge economy, as a high proportion of their graduates get jobs in the knowledge-intensive services sector. The earnings premium for higher education tends to be higher in the EU8 than in the EU15, and the private rate of return to higher education still looks to be high.

199. Expansion brings problems, however. Essentially, there is a conflict between the need for high- quality mass education and fiscal constraints. Most EU8 countries have managed to protect the percentage of GDP that is spent on higher education, and on average they spend a much larger proportion of GDP per head on each student in public tertiary institutions than do EU15 countries. However, the number of Euros spent per student, adjusted for differences in purchasing power, is much lower than in the EU15 group. The trade-off between accessibility and quality of higher education, from the student’s point of view, is becoming increasingly obvious.

200. The EU8 countries have a similar approach to funding higher education in public institutions to that of most of the EU15 – they spend public money on financial aid to students, and try to avoid charging them tuition fees. However, many countries have instituted a back-door dual-track system, maintaining fee-free higher education for regularly admitted state-supported students, while adding a special fee-paying track for those who fail to gain such admission. This tends to penalize students from disadvantaged families, Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 97 who are less successful in entrance examinations, and cannot afford the alternative fee-paying track (or the private institutions that are growing in importance).

201. This combination of trajectories is clearly untenable. The trend towards “massification” of higher education cannot be combined with the improvement in quality that is needed in order to be competitive within the EU and to increase participation in the global knowledge economy, unless the amount of money available to institutions increases. At the same time, government budgets are severely constrained, and it is difficult to justify an increase in public allocations to higher education which yields high private rates of return to people who are disproportionately from privileged backgrounds.

202. An increase in the average level of fees in EU8 public tertiary institutions, to a level which covers a significant proportion of total costs, thus looks inevitable. There is a strong case, on the grounds of revenue maximization, efficiency, autonomy, and equity, for making such fees variable, rather than fixed and uniform. This would allow the inequitable dual-track system to give way to a single-track system, under which all students pay tuition fees, and to a distribution of state subsidies based primarily on need, rather than on academic ability.

203. It would also imply the need for a wide-ranging student loan system, preferably extending income- contingent loans and charging unsubsidized interest rates, and other measures to promote equity in access at all levels of education. The combination of variable fees, needs-based grants, and loans would also help to increase the relevance of the specializations chosen by students to the labor market reality.

204. Governments should not pay too much attention to the public/private distinction. They should rather focus on the quality and relevance of the graduates of tertiary education institutions, regardless of the sources of funding, and put in place a quality assurance system to monitor and enhance the performance of tertiary education institutions of all kinds. Loans and government subsidies could then also be available to students in private universities which meet national eligibility criteria for access to additional public funding by participating in evaluation/accreditation exercises.

205. Further reforms in the criteria for determining the amount of public money to be allocated to each tertiary institution would also help to increase transparency of funding and efficiency. Having started with fairly simple “money-follows-the-student” formulae, experimentation with more sophisticated results-based formulae would be useful in meeting the required policy goals. Rather than immediate implementation of a single mechanism, experience suggests that variety in funding mechanisms is desirable, on grounds of flexibility, multiplicity of objectives, quality and stability. Further governance reforms, particularly in institutional management and procedures for appointing leaders of autonomous tertiary institutions, are likely to be needed to ensure that lump sums allocated through such mechanisms are used in a way that is sensitive to the public interest.

206. Decreased reliance on government funding and increased reliance on price incentives would not mean the elimination of a role for governments in relation to higher education. They would still be an important source of funds, organize and oversee student loan schemes, and be responsible for the promotion of equitable access. They would also have to ensure that quality assurance systems are in place, and would be able to design and use formula-funding schemes to achieve national objectives that go beyond those of the immediate market (e.g. by channeling additional resources to music, drama, engineering etc.). and to modify excessive competition between institutions. What is needed, rather than detailed interference in academic processes, is a combination of standard-setting and financing systems designed to ensure high- quality outcomes. The rest could be left to higher education institutions, autonomous but accountable in their governance arrangements. 98 References Barr, Nicholas (2004), “Higher Education Funding, Oxford Review of Economic Policy.” 20.2. Behrman, Jere R. (1997), “Simple analytical considerations for skill development for international competitiveness”, Chapter 1 in Godfrey (1997). Blöndal, Sveinbjörn, Simon Field and Nathalie Girouard (2002), “Investment in Human Capital through Post- Compulsory Education and Training: Selected Efficiency and Equity Aspects”, Economics Department Working Paper No. 333, ECO/WKP (2002)19, OECD, Paris. Büchel, Felix, Andries de Grip and Antje Mertens (2003), “Overeducation in Europe: Current Issues in Theory and Policy”, Cheltenham, Edward Elgar. Eurydice (2004a), “Focus on the Structure of Higher Education in Europe 2003/04”, Brussels. Feinstein, Leon (2003), “Inequality in the Early Cognitive Development of British Children in the 1970”, Cohort, Economica, 70.277, 73-98. Godfrey, Martin (ed. 1997), “Skill Development for International Competitiveness”, Cheltenham, Edward Elgar. Hopper, Richard (2005a), “unpublished manuscript”, World Bank, Washington DC. Johnstone, D. Bruce (2004), “The Economics and Politics of Cost Sharing in Higher Education: Comparative Perspectives”, Economics of Education Review, chapter 23, pages 403-410. Lucas, Robert E. (1993), “Making a Miracle”, Econometrica, 61.2, pages 251-272. Matějů, Petr, Blanka Řeháková, and Natalie Simonová (2004), “Transition for Secondary to Tertiary Education: Communist and Post-Communist Patterns of Inequality”, Paper presented at the meeting of the Research Committee on Social Stratification and Mobility of the International Sociological Association, Neuchâtel, 7-9 May. Murphy, K., A. Schleifer, and R. Vishny (1991), “The Allocation of Talent: Implications for Growth”, Quarterly Journal of Economics, pages 503-530. OECD, “Reviews of National Policies for Education. Estonia; Latvia (2001) and Lithuania (2002)”, Paris. OECD (2003), “Education Policy Analysis”, Paris. OECD (2004), “Education at a Glance 2004”, Paris. Psacharopoulos, George, and Harry Patrinos (2002), “Returns to Investment in Education: a Further Update”, World Bank Policy Research Working Paper 2881, September, Washington D.C. Romer, Paul M. (1994), “The Origins of Endogenous Growth”, Journal of Economic Perspectives, 8.1, Winter, 3- 22. Sen, Amartya (1999), “Development as Freedom”, Oxford University Press. Sianesi, Barbara, and John Van Reenen (2002), “The Returns to Education - A Review of the Empirical Macro- Economic Literature”, Institute for Fiscal Studies, London. Thorn, Kristian, Lauritz Holm-Nielsen and Jette Samuel Jeppesen (2004), “Approaches to Results-Based Funding in Tertiary Education: Identifying Finance Reform Options for Chile”, World Bank Policy Research Working Paper 3436, October, Washington D.C. World Bank (2002), “Constructing Knowledge Societies: New Challenges for Tertiary Education”, Washington D.C. World Bank/European Investment Bank (2004), “Tertiary Education in Poland”, Warsaw.

World Bank (2005b), “unpublished proceedings of the World Bank/Korean Government International Forum on Tertiary Education Financing Reform in Seoul”, (Jamil Salmi, et al.).

Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 99

IV. CONTROLLING HEALTH EXPENDITURES90 1. Introduction 207. Following the transition from central planning toward market-based economies, the formerly communist states of Central and Eastern Europe introduced a number of reforms in the finance, management and organization of the health sector. Reforms in these countries have been synonymous with rapid dismantling of the state apparatus and with restoration of property and ownership rights, and have accompanied or followed broader structural changes in governance, authority relationships and ownership resulting from a combination of social, political and ideological forces. In almost all cases, the first set of reforms identified systemic problems in the centrally planned socialist health systems and sought to bring about widespread changes. Most importantly, in all EU8 countries the healthcare reforms removed the state monopoly on healthcare by privatization and decentralization, and started the process of establishing an economic relationship between the healthcare provider and the patient through changes in healthcare financing. Almost all countries adopted the Bismarckian social health insurance model, some earlier than others, and in all cases one or more Social Health Insurance Agencies started operating independently of the state budget. In most countries, the next set of reforms also touched the pharmaceutical sector, and brought about changes in the health financing structure in an effort to address the manifold problems brought about by the insurance system.

208. The health sector reforms in the EU8 countries have generally been successful in safeguarding resources for the health sector. Total spending on health in the most EU8 countries is not out of line when compared with other European countries and high-performing middle income countries (Figure IV.1). In Estonia and Latvia, though health spending levels are on the low side relative to their per capita GDP.

Figure IV.1. GDP and Health Expenditures, Selected EU, OECD and Other Countries (2002 or latest available year).

12

o 10

Slovenia 8 Hungary PolandCzech Rep Lithuania

D 6 Slovak Rep GP) xpenditure (% f Estonia Latvia a 4 ethE l 2 Total H

0 0 10000 20000 30000 40000 50000 60000 GDP per capita

New member state EU-15; High-income OECD*; Middle-income high-performing; *High-income OECD countries are Switzerland and Norway. Source: IMF World Economic Outlook Database; World Bank SIMA (Reproduced from Mitra: Towards Sustainable Social Sector Expenditures in the New Member States of the European Union – Keynote Presentation, Chatham House, London, June 22-23, 2005).

90 Prepared by a team headed by Mukesh Chawla (principal author) and including Sarbani Chakraborty and Jan Bultman (background paper on Slovenia), Panagiota Panopoulou (background paper on Slovakia), Pia Helene Schneider (background paper on comparator countries), Marzena Kulis (background paper on Poland), Peter Pazitny (background paper on Hungary and the Czech Republic), Inga Cabe (background paper on Latvia), and Janina Kumpiene (background paper on Lithuania). Peer reviewers were Paolo Belli (The World Bank), Andrzej Rys (former Deputy Minister of Health, Poland) and Imre Hollo (European Investment Bank). The authors are grateful to senior staff in the Ministries of Health and Finance of the EU8 countries for their support in the preparation and their review of the country-specific papers, especially Pawel Sztwiertnia, Deputy Minister of Health, Poland, Rinalds Mucins, State Secretary of Health, Latvia, Laila Ruskule, Deputy State Secretary, Ministry of Health, Latvia, Jan Kralik, Director, Ministry of Health, Slovak Republic, Dorjan Marusic, State Secretary, Ministry of Health, Slovenia, and Rimantas Sadzius, Deputy Minister of Health, Lithuania, Detailed comments and suggestions at various stages of preparation were provided by Armin Fidler, Thomas Laursen, John Langenbrunner, Maris Jesse, Christine Blades and Anton Marcincin. An earlier version of the chapter was discussed at a regional workshop in Riga in June 2005. Akshay Sethi and Shweta Jain provided excellent research assistance. 100

209. The health status of the people of the eight new member states of the European Union is at levels commensurate with their levels of health spending and income, but is generally poorer compared to the health status of the people of the EU15 countries (Figure IV.2). Life expectancy in the EU8 countries varies between 70 years in Estonia and 77 years in Slovenia, and is lower than the average life expectancy of 79 years in EU15 countries. The difference between the EU15 and EU8 average life expectancy has decreased in recent years, but most EU8 countries have some 5 years shorter life expectancies (including disability adjusted life expectancies), than the EU15 average. Gender differences in life expectancy are also higher in the EU8 relative to EU15 averages. Infant death rates in the EU8 countries vary between 3.9 in the Czech Republic to 9.4 in Latvia, but for most countries are higher than the EU15 average of 4.6. The incidence of tuberculosis varies from 11 cases per 100,000 in the Czech Republic to 74 in Lithuania, much higher than the EU15 average of nine per 100,000.

Figure IV.2. Expected Years Spent in Poor health, Selected EU, OECD and Other Countries (2002 or latest available year). r o

o 11,0 p

n r i i 10,0 b s h r a a

e

s 9,0 y

e l e a g

a 8,0 m r

rt e v ot a 7,0

t d l e a

e 6,0 c hhf e p 5,0 Ex t 0 10 000 20 000 30 000 40 000 50 000 GDP Per Capita

New member states EU-15; High-income OECD; Middle-income high-performing; *High-income OECD countries are USA, Australia, New Zealand, Canada, Switzerland and Norway. Source: IMF World Economic Outlook database; WHO Statistical Information System. Reproduced from Mitra: Towards Sustainable Social Sector Expenditures in the New Member States of the European Union – Keynote Presentation, Chatham House, London, June 22-23, 2005).

210. A comparison of the death rates from main causes between countries gives broad indications of how far the observed mortality rates might be reduced. Standardized Death Rates (SDRs) from all causes range from 795 in Slovenia to 1114 in Latvia, significantly higher than the EU15 average of 640. Likewise, SDRs from circulatory system disorders, cerebrovascular disorders, ischemic heart diseases and cancers are higher in EU8 countries compared to the EU-15 average.

211. Within the EU8 countries, health status indicators in Slovenia are generally better compared to other countries, while the health status indicators in Hungary and the Baltic countries are on the lower side. The number of clinically diagnosed cases of AIDS is the highest in Latvia (2.5 per 100,000), followed by Estonia (0.7). SDRs from all causes are over 1,000 in the Baltic countries and Hungary, which also have the highest SDRs due to diseases of the circulatory system as well as cerebro-vascular and ischemic heart diseases. Death rates due to lung and cervical cancer are highest in Hungary and Poland, while Lithuania has the highest reported deaths due to cancer of the cervix. Latvia has the lowest rate among the EU8 countries for malignant neoplasms, but is still about 7 percent higher than the EU15 average. Hungary has the highest rate of almost all types of cancers among the EU8 countries, with figures for cervical cancer and trachea-bronchus-lung cancer almost triple and double respectively the EU15 averages. Since most causes of death due to cardiovascular diseases and cancer are influenced by societal, collective, and individual behaviors and lifestyles, health risks, diseases and premature deaths can potentially be reduced by a wide variety of health promotion and preventive measures. Further, observed mortality can be reduced significantly in the EU8 countries if the healthcare systems in these countries and other determinants of health are more effective in addressing health problems that account for prematurely high levels of mortality. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 101

Table IV.1. EU8 – Standardized Death Rates (per 100,000); different causes (2003).

Czech EU-15 Republic Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia average

All Causes 899.6 1090.58 1047.97 1113.62 1008.26 891.55 971.49 795.49 639.88 Circulatory System 461.88 560.35 508.3 593.02 519.78 413.89 527.71 295.29 236.32 Cerebro- vascular 132.37 154.06 134.59 206.23 117.37 98.57 88.18 78.76 59.05 Ischemic heart diseases 176.09 323 232.66 291.58 327.75 125.78 283.48 94.37 92.89 TB 0.68 6.1 2.41 8.7 9.45 2.33 1.19 1.05 8.65 Alcohol- Related Causes 89.74 174.29 149.55 160.22 176.98 88.95 92.8 111.42 61.28 Smoking- Related Causes 380.91 541.74 491.02 566.77 518.51 306.79 443.02 251.08 220.78 Malignant Neoplasms 234.22 200.6 263.81 193.4 193.57 216.67 213.32 203.66 180.5 Trachea Bronchus 45.27 40.43 66.49 36.85 36.2 53.22 38.11 41.23 37.05 Cancer of the Cervix 6.05 6.67 7.16 6.76 10.64 8.41 6.58 4.11 2.35 Infectious & Parasitic Disease 2.55 8.43 3.98 13.32 13.23 6.18 3.81 4.31 8.38 Respiratory System Diseases 42.35 36.26 41.42 29.32 39.1 37.62 55.2 62.05 48.31 Digestive System Diseases 38.5 42.82 79.94 38.07 41.99 36.68 52.86 53.33 30.81 Liver Diseases & Cirrhosis 16.66 21.72 53.53 14 20.98 12.98 26.55 31.31 12.62

Source: WHO (2005), European Health for All Database (January).

212. There is no doubt that the health sector reforms in EU8 countries succeeded in securing spending levels commensurate with their levels of development, and ensuring health outcomes at levels commensurate with spending on health. However, these efforts have addressed only one part of the problem, leaving a number of issues unresolved. First, commitments in the health sector are higher than actual spending, particularly in the Visegrad countries, resulting in huge and growing indebtedness in the health system. Only Estonia and Latvia, among the Baltic States, and Slovenia in Central Europe have managed to exercise adequate expenditure control. However, longer-term sustainability of health care systems is a key concern in all EU8 countries.

213. Second, while health financing reforms have generally succeeded in safeguarding allocations to the health sector and protecting it from exogenous shocks, the accompanying reforms in efficiency have been half-hearted at best, and have not resulted in cost-containment. The health insurance system in many countries continues to function like centralized collection and reimbursement agencies, and many of the tasks and responsibilities usually identified with an insurance system are lacking. On the delivery side, the shift from the more resource intensive inpatient care to less expensive outpatient care has not been accomplished, and the culture of over-hospitalization and seeking specialized care persists. Recent gains from the reduction in length of hospital stays per episode, and increases in bed occupancy rates in individual 102 facilities have been negated by slow progress in addressing the over-supply of hospital infrastructure. On the financing side, not all healthcare systems have been able to find appropriate mechanisms of reimbursement so as to motivate providers to deliver better quality of services, and to produce them in an efficient and cost- effective manner. On the management side, even though most state-owned hospitals enjoy a fair degree of de-jure autonomy, few efforts have been made to improve administrative capacity and managerial acumen within hospitals, which continue to function as before.

214. Third, health sector reforms in most countries have avoided the contentious issue of the scope of services covered by the public system. The bases for determining the scope of services covered by the health insurance system are laid down in the Constitutions of most countries, and are generally interpreted to imply universal coverage and free-of-charge access to healthcare services through the means of compulsory health insurance built upon the principles of solidarity and the right of health protection for every citizen. This also forms the basis of peoples’ expectations regarding the level of health services which they expect the social health insurance system to finance, and the state-run health system to provide. In most countries, the move from general tax to health insurance financing did not significantly reduce the scope of services provided free of charge to the patients, nor has it been accompanied by efforts to unambiguously define the scope of services to be covered by health insurance. In order to ascertain the appropriate use of available and mobilized resources for the health sector – including non-public sources such as private out-of-pocket payments and private insurance premiums – and in order to structure the system to effectively use these resources, it is necessary to provide more specificity with respect to the scope of publicly covered services. Without such specificity, it is also difficult to ascertain the role that might be played by private supplemental insurance and other non-public resources.

215. Fourth, health sector reforms in most EU8 countries have had little or no impact on the pervasiveness of informal payments (defined as payments in cash or kind that the recipients of such payments are not authorized to receive under the conditions of their contract or under the statutes of the governing bodies of the parent organizations) from patients to providers of healthcare, which constitute a financial burden especially for the poor, and have a negative impact on equity in healthcare financing. Informal payments undermine the impact of health reforms, siphon funds away from the health system, and negatively affect the quality of care for those who cannot or do not pay. Reliable estimates of the nature and extent of informal payments are generally not available, but the problem is known to be particularly severe in Poland and Slovakia, and widespread in many other countries as well.

216. Fifth, health sector reforms have generally not been accompanied by an emphasis on improving the quality of care. Most EU8 countries do not have a well-functioning quality control that regularly incorporates evidence-based medicine in the production, delivery, and financing of healthcare services, and institutional mechanisms to review the quality of care at the health facility level are generally lacking. Clinical protocols for major diseases are often outdated and non-compatible with evidence-based medicine and cost-effectiveness analysis, and there is anecdotal evidence that quality of care is generally low. The existing quality improvement systems are fragmented and are not integrated across different elements of the healthcare system, including professional self-regulation, purchasing by health insurance funds, training of health professionals, and management of individual healthcare practices. As a result, the existing quality control institutions do not achieve the objective of continuous quality improvement.

217. Thus, while health sector reforms in the EU8 countries in the past decade have involved deep structural changes, they have generally been less successful in securing sustainability of health care financing, improving efficiency, enhancing equity in healthcare financing and delivery, and managing clinical quality of health services. Total health expenditures have increased in almost all countries, especially in recent years, and with revenues not keeping pace, huge debts have accumulated in the health sector. Population aging will add further strain on health care systems in the region. Efficiency gains have been few and far-between, and with the dynamic nature of technology and demographic changes increasing the complexity of health services and the health marketplace, further reforms are becoming even more difficult. The main objectives of this chapter, therefore, are to take stock of recent trends in health expenditure aggregates in the public sector and to identify specific areas of reform consistent with the objectives of consolidating the fiscal situation in these countries without adversely affecting the production, delivery and utilization of health services. The rest of this chapter is organized as follows: Section 2 discusses trends and structure of health expenditures in the EU8. Section 3 assesses the nature Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 103 and extent of indebtedness in the health sectors. Section 4 discusses the key expenditure areas. Section 5 presents population ageing and proliferation of medical technology as the future spending pressure points in the health sector. Section 6 focuses on the management of health expenditures. Finally, Section 7 concludes.

2. Health Expenditures 218. Healthcare spending in the EU8 countries varies greatly, from 8.1 percent of GDP in Slovenia to 5.4 percent of GDP in Estonia (Figure IV.3). Overall, the three Baltic countries spend less than 6 percent of GDP on health, the Visegrad-4 countries (henceforth V4) spend between 6.2 percent and 7.5 percent of GDP, and Slovenia spends a little over 8 percent of GDP. On average, the EU8 countries spend 6.6 percent of GDP on health, relative to 8.1 percent in EU15 countries.

Figure IV.3. Health Expenditures in EU8 Countries, Percent of GDP, 2003.

10 8 DP f 6 o G

n 4 e c

r 2 e pt 0 a ia ic ry ia d i ia ia 0 5 8 n l a k n tv n n 3 1 e b g a la a a o U v u n v o L u t D U E lo p u lo P h s C E S e H S it E E R L O h c e z C Public expenditures Private expenditures

Source: Country Background Papers; OECD 2004.

219. Health expenditures in the EU8 countries are mainly financed from public sources, but the share of public financing of health expenditures has been declining in recent years. On average, public financing accounts for 77 percent of total health spending in EU8 countries, which is comparable to the EU15. The proportion of public financing of health expenditures is highest in the Czech Republic (93 percent) and lowest in Latvia (52 percent). The decline in public expenditures has gone hand in hand with an increasing volume of out-of-pocket payments in the health sector. Thus, the share of out-of-pocket spending on health in Slovakia increased from 8 percent in 1998 to over 13 percent in 2003, following the introduction of modest co-payments for health services. In Poland, out-of-pocket financing of health expenditures increased from 25 percent to 30 percent of total spending on health between 1997 and 2003. Likewise, in Latvia, the share of out-of-pocket payments increased in just three years, from 39 percent in 1998 to 47.5 percent in 2001, and has remained at that level since (Figure IV.4).

Figure IV.4. Public and Private Expenditures on Health in Latvia.

70 60.7 59.7 56.8 55.8 60 52.5 52.5 52.5 50 e 47.5 ct 47.5 47.5 40 per n 44.2 43.2 39.3 40.3 30 20 1997 1998 1999 2000 2001 2002 2003

Public Expenditures on Health Private Expenditures on Health

Source: Country Background Papers. 104

220. Social health insurance is the dominant form of public financing of the health system in EU8 countries, even though many also directly transfer funds from the state budget to the social insurance system. Latvia is the only exception among the EU8, where statutory healthcare resources consist partly of income tax collected at the central level (28.4 percent of income tax revenue is earmarked for healthcare), partly of subsidies from general revenues (also financed by tax revenues at the central level) and partly of payments from patients and private insurers. In all other countries, the social health insurance system is at the core of health financing, with varying degrees of budgetary support. With the exception of the Czech Republic and Slovakia, all EU8 countries have a single monopsonist agent who purchases health services on behalf of the insurees. The public financing system in all EU8 countries provides generous coverage of health services, in most cases with little or no financial participation from the patient at the point of service (Table IV.2).

Table IV.2. EU8 – Scope of Services Covered by Social Health Insurance.

Country Population Services Covered Formal Co-payments (million) Czech 10.2 Preventive services, diagnostic procedures, Dental services, some drugs Republic ambulatory and hospital curative care, including and medical aids. rehabilitation and care of the chronically ill, drugs and medical devices, medical transportation services, and spas. Estonia 1.31 All, with some exclusions such as cosmetic GP home visits, outpatient surgery, alternative therapies and opticians’ prescription drugs, outpatient services, and limited coverage of adult dental specialist care, inpatient care. care. Hungary 10.11 Almost all primary, secondary and tertiary care Dental treatments, services services. without referral, and extra ‘hotel’ spaces of hospital services, chronic care and treatment in a sanatorium. Latvia 2.31 Emergency care, treatment of acute and chronic All patients receiving statutory diseases, prevention and treatment of sexually benefits participate with co- transmitted and contagious diseases, maternity payments. Pediatric, natal and care, immunization programs and drugs. emergency treatments are exempted from copayments. Lithuania 3.41 All services covered. Drugs and some medical aids for ambulatory treatment and spa services. Poland 38.47 All, with some exclusions such as cosmetic Drugs, some medical aids, surgery, alternative therapies and opticians’ dental care. services. Slovakia 5.39 All, with some exclusions such as cosmetic Introduced in June 2003 at all surgery, alternative therapies and opticians’ levels of care except for services. emergency care, preventive care and health services for children under the age of six years. Slovenia 1.97 Full coverage of essential services, defined to All inpatient and outpatient include services for children and adolescents, care not covered under family planning and obstetric care, preventive essential services as defined in care, diagnosis and treatment of infectious the previous cell, and drugs diseases (including HIV), treatment and (though people may take rehabilitation of a range of diseases including complementary insurance to cancer, muscular and nervous diseases, mental cover these co-payments). diseases and disability, emergency care (including Several population groups are transport), nursing care visits and home care, exempt from co-payments. donation and transplantation of tissues and organs, and long-term nursing care.

221. Health expenditures in almost all EU8 countries are dominated by inpatient care and pharmaceuticals, as the following examples show, and account for roughly 40 percent and 30 percent of total health expenditures respectively. Primary and specialist outpatient care, long-term care, and administrative expenses comprise the balance. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 105

222. The general trend over time has been that of rising expenditures on pharmaceuticals, (marginally) falling expenditures on inpatient care and (marginally) increasing expenditures on outpatient care. In Poland, the share of inpatient care in total public expenditures on health has fallen from 43.3 percent in 1999 to 40.6 percent in 2003. The share of outpatient care expenditures has remained more or less constant at 23 percent, while the share of expenditure on drugs increased from 13.7 percent to 19.6 percent (Figure IV.5). In Slovakia, a large share of public resources is devoted to hospital care as compared to primary and secondary level outpatient care. In 2002, inpatient care costs represented 40 percent of total health costs, as compared to 7 percent of total costs for primary care and 3 percent of total costs for secondary outpatient care (Table IV.3). Likewise, health expenditures in Hungary are dominated by expenditures on inpatient care, and on subsidies to pharmaceuticals and medical equipment. Expenditures on inpatient care have fallen marginally from 41.9 percent of total public expenditures on health in 1999 to 41 percent in 2003, and are expected to be around 40 percent in 2004. Likewise, expenditures on pharmaceuticals have fallen marginally from 31.8 percent of total public expenditures on health to 30.4 percent in 2003, and are expected to be 29.3 percent in 2004 (Figure IV.6). In Latvia, detailed information is available only for statutory healthcare resources. Including related pharmaceutical costs, inpatient services account for 59.4 percent of the total statutory healthcare resources, outpatient care for 39.6 percent, and emergency medical care for 6.7 percent.

Figure IV.5: Structure of Health Expenditures in Poland, 1999-2003.

100% 80% 60% 40% 20% 0% 1999 2000 2001 2002 2003 Hospitals Outpatient care Drugs Public health Adm inis tration Long-term care Other

Source: Poland Ministry of Health (2004).

Table IV.3. Health Sector Costs in Slovakia (1998-2003; percent of total).

1998 1999 2000 2001 2002 2003 Primary outpatient care 7.4 7.5 7.3 7.0 6.8 6.6 Secondary outpatient care 2.6 3.1 2.9 3.0 2.9 2.9 Inpatient care 44.8 42.7 40.2 39.9 40.3 38.9 Drugs and medical devices 28.2 32.1 31.9 32.3 32.3 32.5 Others 8.8 7.0 10.7 10.9 11.1 13.0 MOH Expenditure 8.2 7.5 7.0 7.0 6.4 6.1 Total costs (SKK billion) 57.1 58.5 64.6 70.5 74.6 78.5

Note: Figures for 2002 and 2003 are estimates. Source: Zajac and Pažitný (2002). 106

Figure IV.6. Structure of Public Expenditures on Health in Hungary.

100% 80% 60% 40% 20% 0% 1999 2000 2001 2002 2003 2004e

Inpatient Care Subsidization of Drugs and Medical equipment Outpatient Specialist care Outpatient Primary Care (including dental) . Others

Source: Country Background Papers.

Figure IV.7. Structure of Public Expenditures on Health in Latvia.

100%

80%

60%

40%

20%

0% 1999 2000 2001 2002 2003 In-patient Out-Patient Emergency Medical care Others

Source: Country Background Papers.

223. This structure of spending is not very different from the structure of health spending in OECD countries (Figure IV.8). On average, OECD countries allocated most of their health funds for inpatient care (38 percent), followed by outpatient services, including ancillary services and home care (31 percent), medical goods, including pharmaceuticals and medical appliances (21 percent), and administration and prevention programs (10 percent). Pharmaceutical expenditures in Canada, Spain, and the UK (countries with tax-financed systems) as well as in Austria and Germany (social health insurance system), are less than the OECD average, whereas they reflect the average in France, the Netherlands, and Switzerland. Inpatient expenditures are above the OECD average of 38 percent in the UK, Switzerland, France, and Austria, all of which have different institutional arrangements and total health spending levels. In the UK and France, a larger part of inpatient care expenditures is allocated to curative care and rehabilitation than in Switzerland. Switzerland, with the highest health expenditures, also reports the highest proportions of spending on long- term care in nursing homes and homecare organizations (Rossel and Gerber, 2004). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 107

Figure IV.8. Structure of Health Expenditures, OECD average (2001).

Source: Orosz and Morgan (2004). 3. Health Sector Deficits and Debts 224. A defining feature of the health sector in almost all the EU8 countries is the widespread and growing indebtedness, which has assumed alarming proportions in recent years.91 Buoyed by generous benefit packages (offering ineffective and non-essential services), and extensive infrastructures handed down as legacies from the pre-independence era, and pressured in recent years by increasing pharmaceutical costs and expectations of higher salaries, expenditures on health in country after country have consistently surpassed the available resources, leaving huge unpaid bills for services already delivered. This situation is particularly severe in the V4 countries, though some of the Baltic countries are also beginning to feel the impact of rapidly rising expenditures in the health sector.

Hungary 225. Public funding of health expenditures in Hungary is channeled through the Health Insurance Fund, which is responsible not only for benefits in kind in healthcare, but also for cash benefits for sick-pay and some types of pensions (like disability). In 2004, the Health Insurance Fund spent HUF 946 billion on healthcare benefits in kind, and paid HUF 422 million on benefits in cash. Revenues of the Health Insurance Fund (HIF) have consistently been below expenditures, but the gap has grown considerably in the last two years. In 2003, expenditures exceeded revenues by over HUF 300 billion, equivalent to 1.6 percent of GDP (Figure IV.9). There was a marginal improvement in 2004, with expenditures exceeding revenues by around HUF 277 billion, or 1.3 percent of GDP.

Figure IV.9. Fiscal Position of the Health System, Hungary.

8 0 -0.4 -0.5 -0.2 6 -0.5 -0.5 DP DP f f o o -0.7 -0.7 G G 4 -1

2 -1.3 -1.5 e e prcent prcent -1.6 0 -2 1997 1998 1999 2000 2001 2002 2003 2004e HIF Revenues (left axis) HIF Expenditures (left axis) Deficit (right axis)

Source: Country Background Papers.

Slovakia 226. Health expenditures in Slovakia have exceeded revenues year after year since 1997. Total expenditures on healthcare in 2003 were SKK 72.9 billion, representing 6.9 percent of the GDP, while health revenues were less than SKK 69 billion, equivalent to 6.5 percent of GDP taking the stock of debts to SKK 31.2 billion, equivalent to

91 Debts refer to payments outstanding and past due, not liabilities which are natural consequences of the financial management of health facilities. 108 over 2 percent of the GDP (Figure IV.10).92 The gap between revenues and expenditures is expected to be closed by 2006, however, as the government’s ambitious health sector reform program begins to yield results.

Figure IV.10. Fiscal Position of the Slovak Health System (percent of GDP).

0 0 0 -0.1 7.6 7.6 7.3 7.3 7.7 -0.2 8 7.2 6.9 6.9 6.5 -0.2 n 6.2 6.6 d 6 7.2 6.4 6.5 -0.4 it 7 6.9 ic

d 6.8 6.1 6.4 6.4 6.4 6.5 f

u -0.4 ntures e

nsa 4 -0.6 -0.5 -0.6 D

pi -0.7 2 -0.8 Ex e Reve e -0.9 -0.9 -0.9 0 -1 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Revenues (left axis) Expenditures (left axis) Deficit (right axis)

Notes: (1) Figures for 2004 are estimates; for 2005 are forecast; (2) Revenues and expenditures in 2004 and 2005 reach relatively lower shares of GDP because of the high growth rate of the Slovak economy; (3) Revenues in years 2003 and 2004 do not include the bailing out of hospitals and HIC (approx. SKK 15 billion) via the state-owned company Creditor. Source: Ministry of Health of the Slovak Republic.

Poland 227. The health sector in Poland has been generating deficits for over a decade now, and despite a major bail-out by the state just before the implementation of health reforms in 1999, the level of matured debts stood at PLN 5,930 million as of March 31, 2005. This figure includes debts created by autonomous units of both local and central governments. A peculiar feature of the health sector debts in Poland is that about 80 percent of the matured debts originate from only 15 percent of establishments (and about one-half from 6 percent of the total number of establishments).93 The most indebted facilities are located in the Wroclaw region (Dolnoslaskie Voivodship, a region that also has the largest healthcare infrastructure in Poland), which generates almost 20 percent of all debts. Other regions with debts above the average are Lodz, Warsaw, and Gdansk. The largest share of hospital-based debts (31%) is towards the public sector, mostly for local taxes, real estate tax, and social insurance on behalf of the employees, the second largest debts are towards the suppliers of drugs and medical consumables (20%), and the third largest debts are towards the employees (19%) resulting mostly from the Act of 22 December 2000 (also known as the “203 Act”) which required all healthcare establishments to pay employees wage increases in 2001 of not less than PLN 203 per month and in 2002 at a level not lower than the average wage growth in the national economy (Figure IV.11).94

Figure IV.11. Structure of Hospital Debts in Poland.

25% 31%

1% 4%

19% 20%

Debt tow ard the public sector (taxes, social insurance, etc.) Debt to suppliers of drugs and medical utilities Debt tow ard employees Debt to suppliers of medical equipment Debt due to investment and renovation Debt to others Source: Poland Ministry of Health (2004).

92 Debts are generally less than accumulated deficits owing to past bail-outs of the health sector. 93 These figures refer to end-June, 2003. 94 Recently, there has also been an intensification of debt collection from defaulting providers by way of seizure by court order, directly from assets of public providers, as well as through the National Health Fund (NFZ). The total value of debt collection through seizures. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 109

Czech Republic 228. The current level of debt in the health sector in the Czech Republic is not high, but rising rapidly. The stock of debt as of December 31, 2004 was approximately CZK 9 billion (less than 0.1% of GDP), all on behalf of one health insurance company, VZP). Projections suggest that debt could reach CZK 24 billion in 2006 (this data is only for VZP, since no other insurer has had any overdue payables, though some are expecting to get into debt within a year or so).95

Lithuania 229. Expenditure of healthcare institutions for services and reimbursable medicines are covered from the budget of the Central Health Insurance Fund (CHIF). However, healthcare services and medicines drive up the actual expenditures to levels higher than planned. Actual expenditures exceeded planned expenditures by Litas 33 billion in 2003, leaving the CHIF in debt to healthcare institutions to the tune of Litas 263 billion in 2003 (0.47 percent of GDP). The largest debt of the CHIF is to personal healthcare providers (about 50 percent) and pharmacies (about 40 percent). The indebtedness of the CHIF to personal healthcare providers, pharmacies and other healthcare institutions grew during the period of 1997-2004, particularly during 2000- 01. As a result of the CHIF’s indebtedness to healthcare institutions for provided services and reimbursable medicines, healthcare institutions are indebted to their staff and to the State Social Insurance Fund.

Table IV.4. Lithuania – Debt of Health Insurance Fund to Providers (billion Litas).

2000 2001 2002 2003 Indebtedness at the beginning of the period 226.603 271.915 363.025 317.445 Targeted expenditure 1,847.170 1,833.456 1,811.585 1,868.722 Actual expenditure 1,836.929 1,936.643 1,847.835 1,901.965 Reimbursed expenditure 1,789.256 1,845.978 1,893.474 1,956.963 Indebtedness at the end of the period 271.915 363.025 317.445 262.875 Source: State Patient Fund, Lithuania.

4. Key Expenditure Areas in the Health Sector Pharmaceutical Costs 230. Drugs are the single largest cost driver in almost all healthcare systems in the EU8, and have been the most dynamically growing element in overall costs of healthcare services in recent years. This is not a situation unique to EU8 countries – according to new data released by the OECD, spending on pharmaceuticals across OECD countries has increased by an average of 32 percent in real terms since 1998, reaching more than US $450 billion in 2003. Growth in drug spending has outpaced total health expenditure in the EU8 countries in the last five years, as well as in most OECD countries. In some OECD countries, like the United States and Australia, spending on drugs grew more than twice as fast as total health expenditures between 1998 and 2003, while in others, like Japan, Italy and Switzerland, the growth was more moderate.

231. In Slovakia, expenditure on pharmaceuticals increased dramatically during 1995-2002, accounting for 32 percent of total health expenditures in 2002. The increase was fueled by both increasing consumption of pharmaceuticals – each year an estimated 52 million prescriptions are made out for a population of 5 million, and which until recently were fully or partly reimbursed by health insurance companies – and by increasing prices, driven up by non-transparent procurement and price-setting policies, as well as by increases in international prices of pharmaceuticals.

232. In Poland, expenditures on drugs constitute over 30 percent of total healthcare expenditures. The Polish market is dominated by imported drugs, which account for 63 percent of the market share, most of which are still in the patent period. The total value of imported drugs has grown at a rate of over 15 percent per annum during 2001-2003, while the value of domestically produced drugs grew at a rate of 10 percent per annum. Both retail prices and the quantity of drugs consumed (especially expensive drugs) have grown during this period.

95 Officially, none of the providers has any debts, but this may well be an accounting machination tacitly supported by MOF and MOH. 110

Figure IV.13. Total Pharmaceutical Expenditure in Latvia (2003 prices).

400 15%

300 10% oth VL g

200 5% w o f i l o l r i

mnL 100 0% rate 0 -5% 2000 2001 2002 2003 Total Spending on Health Total Expenditures on Pharmaceuticals Rate of Growth of Total Health Expenditures Rate of Growth of Expenditures on Pharmaceuticals Source: Country Background Papers.

233. In Latvia, pharmaceuticals accounted for about 29 percent of total expenditures on health during 1999- 2003. Expenditure on pharmaceuticals increased by about 38% during this period (2003 prices), in line with the increase in total spending on health (Figure IV.13). The consumption of reimbursable medicines in Latvia has been higher than allocated resources, and has increased in recent years. The share of reimbursable medicines in total expenditure on pharmaceuticals is small, but has increased from 11.3 percent in 1999 to 16.7 percent in 2003.

Box IV.1. Expenditure on Drugs (OECD countries; 1998-2003) Spending on drugs represented, on average, 18 percent of total health spending in OECD countries in 2003. The share ranged from highs of around 30 percent in Slovakia, Korea, and Hungary, to lows of around 10 percent in Denmark and Norway. In 2003, total drug expenditure per person was highest in the United States (more than US $700 per person), followed by France (just over US $600), Canada and Italy (about US $500); the lowest spending of just over US $100 was in Mexico and Turkey. Variations in drug spending across countries reflect differences in prices and consumption, as well as the pace of the introduction of new and often more expensive drugs. Differences in income levels across countries is also a significant factor affecting spending on pharmaceuticals.

Figure IV.13. Annual Growth in Drug Expenditure and in Total Health Expenditure, 1998 to 2003.

Drug expenditure Total health expenditure %

14 . . 2 2 18 17 .

12 1 14 . 0 12 .

10 96 . 91 . 9 83 . 7

8 0 . . 5 7 69 . . 1 0 6 9 . 64 . . 6 . 6 6 . 4 5 3 58 . . . . 6 5 54 54 5 . 5 . . 50 . . . 49 48 . 1 46 46 46 0 . . . . 43 42 41 4 5 . 4 4 . . . . 37 . . 35 4 3 3 34 8 . 32 32 . . 30 . 28 2 2 26 . 2 . . 18 2 18

0 ) y e n s c l i 1) n c c e n d n d ( ( ( ( a a

u

) 1) ) 1)

a t b n e a a n a p I l o as og d n n E u a n a ee ed nd r m n a1 r Si e Krea p a nark r OCD ay a1) F d Ic l el w G Fi e

e e any rlia2 Ca ray1 ap ( e e( hl d eb r u n Dm o G J

w ett Ir l x ur Astria( h N u S Hg Switzerl d c Net r Aust a Lm eR Unit S Cz Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 111

Around 60 percent of drug spending is publicly financed on average across OECD countries, with the remainder paid by private sources, mainly out-of-pocket payments and private health insurance. By comparison, almost three-fourth of total health spending is funded through public sources on average in OECD countries. Most countries have seen the public share of pharmaceutical spending increase in recent years, including in the United States, although at around 20 percent in 2003, it is still the second lowest public share among OECD countries. Notes: Countries are ranked from left to right by annual growth of per capita pharmaceutical expenditure; (1) 1998-2002; (2) 1997-2001. Source: OECD Health Data 2005 (June).

234. Hungary spends $280 per person per year on drugs, more than the UK and the Czech Republic (both around $240). In fact, drug policy is one of the weakest links of the Hungarian healthcare system, and the sub-budget for drugs is the only sub-budget of the Health Insurance Fund without even an indicative macro limit. The other sub-budgets for primary care, secondary care, and hospitals have limits and are obliged to create reserves throughout the accounting period. When the reserves are not sufficient, and this happens every year, the ministry responds by decreasing the value of the point system used to reimburse outpatient care and decreasing the base rate for the Diagnosis Related Group (DRG) used to reimburse inpatient care, so that the target of the sub-budget is met. Pharmaceutical expenditures regularly cause deficits that the government then pays from the state budget. The second reason for a permanent deficit of the drugs “treasury” is the introduction of new products and drugs to the market. Every drug released on the internal market is registered with the National Institute of Pharmacy, with minimal checks and balances as regards efficacy and use. New drugs are typically priced high, and where refunds are percentage-based, the financial burden for health insurance increases. Hospital infrastructure 235. The current oversupply of hospital infrastructure in the EU8 countries is a legacy of the Soviet influence. The focus of healthcare in the Warsaw Pact countries was the treatment of diseases, and the establishment of a huge infrastructure of hospitals was a natural process in the fulfillment of the deemed priorities of the health sector. At the same time, geopolitical and strategic preparations for potential wartime eventualities also dictated the development of a vast hospital network in the front line cities of the extended Soviet influence. Expectedly, therefore, the EU8 countries inherited a disproportionately large number of hospitals and hospital beds, which soon started to become a drain on public resources in the health sector.

236. Almost all countries have taken many bold steps to reduce the number of hospitals and acute- care hospital beds, but very few have done nearby enough. On average, the number of beds per 100,000 population in EU8 countries declined from 987 beds in 1993 to 765 in 2002, but is still 30 percent more than the average of 611 for EU15 countries (Table IV.5). The Czech Republic has 80 percent more hospital beds per 100,000 inhabitants compared to EU15, followed by Lithuania (50 percent) and Hungary and Latvia (30 percent). The large number of hospital beds, combined with longer lengths of hospital stay and low occupancy rates (for example, the occupancy rate in Slovakia is only 62 percent), generates enormous fixed costs for the system and locks-up scarce financial resources. Of all the EU8 countries, only Estonia and Slovenia have a hospital infrastructure comparable with the EU15 standards in terms of capacity.

237. Regions with an excessive concentration of public hospitals and hospital beds expectedly also generate the largest debts in the health sector. Hospital contracts with health insurance companies do not cover the full costs of maintaining the huge hospital infrastructure and associated fixed costs, and the resources allocated by the central and local governments (who own most public hospitals in most EU8 countries), are not sufficient to pay for all the expenses. In addition, public hospitals tend to provide services irrespective of the value of contracts concluded with the sickness funds, and bear costs that are refunded neither from health insurance nor from the state budget. In short, maintaining the large number of hospitals and hospital beds causes a huge drain on available resources and is one of the largest cost-drivers in the health system. A 30 percent reduction in hospital beds in the region will surely free up significant resources, particularly if they are accompanied by the closure of hospitals. 112

Table IV.5. Beds per 100,000 Inhabitants (EU8; 1993-2002).

Compared 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 to EU15 EU (15 countries) 711.6 705.7 692.0 686.2 664.1 648.9 625.0 622.9 612.8 611.3 1.0 EU 8 987.4 955.0 910.3 877.8 846.6 832.4 820.7 804.5 783.0 765.0 1.3 Czech Republic 1218.5 1209.7 1134.5 1107.4 1125.2 1113.6 1104.1 1092.6 1095.8 1107.1 1.8 Lithuania 1173.9 1108.1 1083.0 1055.6 983.0 961.5 938.0 923.2 869.4 892.8 1.5 Hungary 1004.1 990.6 909.1 910.9 826.4 831.1 836.8 839.1 806.3 806.3 1.3 Latvia 1203.0 1184.3 1099.3 1025.0 961.3 922.2 885.2 855.1 809.5 773.4 1.3 Slovakia 832.7 815.2 804.6 795.7 784.1 766.9 756.9 1.2 Poland 791.7 784.0 768.7 766.9 757.8 744.0 735.1 718.7 717.5 709.9 1.2 Estonia 941.8 830.9 804.1 757.6 738.2 722.8 716.5 682.9 681.8 605.9 1.0 Slovenia 578.7 577.7 573.6 566.6 565.3 559.1 554.0 540.6 516.9 508.9 0.8

Source: Eurostat, 2003. Salaries of Medical Personnel 238. Salaries in the health sector in the EU8 countries are rising faster than the average salaries in the economy, and as salaries account for more than 60% of health expenditures in these countries, this trend is increasing the pressure on overall health spending. Encouraged in part by the greater availability of outside opportunities following EU accession, and in part by the push to bring the ratios of salaries in the health sector to average salaries in the economy in line with EU15 proportions, physicians and nurses have been able to successfully negotiate demands for higher salaries. This has been and continues to be a big issue in most countries, as the examples below from Latvia, Estonia, Slovenia and Poland indicate.

239. In Latvia, the financing of regional sickness funds for healthcare services increased by 18 percent in 2002 in nominal terms, driven largely by an increase in the minimum salary. Most of the additional funding came from the Reserve Fund and was paid out as additional money (i.e., in addition to the regular salaries), on the basis of workload of medical personnel in medical institutions. The additional salary component was settled between the salary levels included in the price of contracted services and salary demanded by medical personal during strikes, and translated to LVL 53 per month for physicians, LVL 9 for paramedical staff, and LVL 7 for junior paramedical staff. In 2003, the financing by sickness funds for healthcare services increased by 17 percent compared to 2002, and like in the previous year, was also connected with an increase in minimum wages. Overall, average monthly salaries in the health sector in Latvia have increased from LVL 141 in 2002 to LVL 190 in 2003, and of doctors from LVL 214 in 2002 to LVL 291 in 2003.

240. Even though the average remuneration of medical personnel employed in the healthcare system has increased considerably in recent years, the low levels of remuneration are considered to be one of the main obstacles in the development of human resources in the healthcare system in Latvia. In order to increase the number of medical personnel in the age group 25-40 years by at least 5 percent, and to decrease the number of patients per physician, the Ministry of Health recently suggested increasing physician remuneration to levels twice the average salary in the economy. Five variants are suggested, which essentially vary only with the speed at which the ratio of 2:1 is achieved. The average salaries in the economy are expected to rise to LVL 265 by 2010, implying that doctors’ remuneration will be set at LVL 430. Whichever variant is finally adopted, it will increase total spending on health by between 5 percent and 7 percent, equivalent to around 0.2 percent of GDP.

241. In Estonia, the recent increase in salaries announced by the previous government has resulted in a huge financing gap of EEK 400 million in the Health Insurance Fund, equivalent to almost 10 percent of total expenditure on health services by the Fund in 2004. The new government is now actively discussing ways to close the financing gap, and while there seems to be still no agreement on the exact course of action, alternatives that are being discussed include increasing the out-of-pocket payments of patients, reducing the benefits paid to insured individuals during sick days or starting the payments a day or two later. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 113

242. In Slovenia, the fiscal balance in the health system received a jolt in 1996 when doctors went on strike demanding increased compensation. The strike was resolved with the government committing to link physician salaries to the salaries of judges, a move that has since been completed. The doctors again went on strike in 2002, but while there was broad agreement that doctors were generally overworked, the government did not agree to yet another pay increase, fearing that it would trigger demands and pressures from other streams of professionals as well, and threaten the government’s efforts to maintain balance in public finances. This strike was resolved with the government agreeing to correct the imbalance between the different professions and prepare a law on public sector pay to this effect.

243. In Poland, average remuneration in the health sector has been lower compared to other sectors. This gap has only been increasing in recent years, and in 2002 average remuneration in the health sector was about 23 percent below average wages in the public sector. This differential persists despite the government decision in 2000 to increase health sector wages in 2001 by no less than PLN 203 per month, and maintain the average growth rate in health sector wages at levels comparable to the national average. However, only about 55 percent of all healthcare establishments raised wages in 2001, with less than half covering the full amount of PLN 203 per month. The situation did not improve much in 2002, and healthcare establishments – citing limited financial resources – continued to be unable to implement the statutorily imposed obligations. Most healthcare establishments implementing the government decision ran up huge debts, estimated to be over PLN 1.9 billion, while more than PLN 2.2 billion (principal only) remains unpaid to employees in establishments that have not implemented the government order.96

244. It is illustrative to compare average physician remuneration in the health sector with average salaries in the country at equivalent levels of educational qualification and experience. This data is not readily available and there are many problems in determining average salaries in the health sector, particularly because of the many different combinations of qualifications, training, specialization and experience. Average salaries at equivalent levels in the Czech Republic, Slovakia, Slovenia and Hungary (Table IV.6). The ratio of physician salaries to the national average is more than 2 in the Czech Republic, but only 1.2 in Slovenia. In a very rough way, this also suggests that salaries are likely to become significant cost drivers in many EU8 countries if the ratio of physician salaries compared to national averages in these countries converges to prevailing levels of about 2 in EU15 countries.

Table IV.6. Physician Salaries (selected EU8 countries; average monthly in local currency).

Year 1999 2000 2001 2002 2003 Czech AMS of physician (CZK) 23 101 25 094 28 139 31 654 33 961 Republic AMS in Czech Republic (CZK) 12 797 13 614 14 793 15 866 16 920 Ratio of Physician Salaries to National 1.81 1.84 1.90 2.00 2.01 Average Slovakia AMS of physician (SKK) 20 050 23 837 27 879 28 973 29 862 AMS in Slovakia (SKK) 11 430 12 365 13 511 14 365 15 825 Ratio of Physician Salaries to National 1.75 1.93 2.06 2.02 1.89 Average Slovenia AMS of health workers (SIT) 173 245 191 669 214 561 235 436 253 200 AMS in Slovenia (SIT) 203 098 224 575 253 131 295 319 308 013 Ratio of Physician Salaries to National Average 1.17 1.17 1.18 1.25 1.22 Hungary AMS of a specialist (HUF) 112 248 127 074 148 348 162 754 AMS in Hungary (HUF) 70 211 79 820 103 254 122 482 Ratio of Physician Salaries to National Average 1.60 1.59 1.44 1.33 Source: Ministry of Health, the Czech Republic; Ministry of Health, Hungary; Statistical Office of the Slovak Republic and Institute of Health Information and Statistics of the Slovak Republic; International Labor Organization, 2005 (Slovenia).

96 GUS communication 19 August 2004, ref DUI-06-3089/04 in Government of Poland (2004). 114 5. Spending on Ageing Populations and Medical Technology Aging Populations 245. Financing and delivering healthcare for the growing population of the elderly is likely going to be the largest new cost-driver in the health systems of the EU8 in the near future. All EU8 countries face the consequences of population ageing caused by reduced fertility and mortality rates on the one hand, and increasing life expectancies on the other. Total Fertility Rates (TFR) in the EU8 fell from 1.8 in 1990-1995 to 1.2 in 2000-2005, and is projected to increase only marginally to 1.5 by 2020-2025 (Figure IV.14).97 Hungary will experience the lowest decline in TFR (0.3), while Lithuania will see the largest decline (0.9).

Figure IV.14. EU8 Actual and Projected Total Fertility Rate (TFR).

2,5

2 ts R e 1,5 t i l ia 1 F rt y a t o

Tle 0,5

0 1980-1985 1985-1990 1990-1995 1995-2000 2000-2005 2005-2010 2010-2015 2015-2020 2020-2025 Cz ec h Es tonia Hungary Latvia Lithuania Poland Slovak Slovenia EU8

Source: Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, World Population Prospects (2004 Revision) and World Urbanization Prospects (2003 Revision).

246. Life expectancy at birth in the EU8 countries has been steadily rising and is expected to continue rising. Overall for the EU8 countries, life expectancy at birth for females increased from 75.8 years in 1990- 95 to 78.1 years in 2000-2005, and is projected to rise to 81.1 years in 2020-2025. Likewise, life expectancy at birth for males increased from 67.7 years in 1990-95 to 70.9 years in 2000-2005, and is projected to rise to 74.5 in 2020-2025 (Figure IV.15).

Figure IV.15. EU8 Actual and Projected Life Expectancy at Birth.

82 t

r 78 i B

t a

y c

n 74 a t c eh x E 70 e f Li p

66 1980-1985 1985-1990 1990-1995 1995-2000 2000-2005 2005-2010 2010-2015 2015-2020 2020-2025 Cz ec h Es tonia Hungary Latvia Lithuania Poland Slovakia Slovenia

Source: Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, World Population Prospects (2004 Revision) and World Urbanization Prospects (2003 Revision).

97 TFR is the average number of children a woman is expected to have by the end of her reproductive period. Since it is measured using information on births of women aged 15-49 in a certain period, it is the average number of children a woman is expected to have between the ages 15-49. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 115

247. The net result of decreasing TFR and increasing life expectancies is that the share of people aged 60 years and older in the total population in all EU8 countries, which was 16 percent in 1995, is projected to increase to 27 percent in 2025 (Figure IV.16). As a result of these movements, there will be 60 million inhabitants in the EU8 by 2050, almost 13 million less than today. Latvia and Lithuania face the largest declines, losing over a quarter of their population by 2050. At the same time, the old age dependency ratio and total dependency ratio are also expected to rise sharply.

248. Older persons are more likely to have greater health needs and are more likely to consume more expensive healthcare services, particularly during the last years of life. Without a doubt, the extent to which ageing affects overall health spending will depend on the utilization of services, forms of elderly care offered, changes in technology and the way elderly care is paid for, but a rapidly ageing population will bring significantly higher needs, simply because the elderly (above 65) have a higher average demand for medical care in their later years in terms of ambulatory, inpatient, and chronic care (Oxley and MacFarlan, 1995). In a study based on Swiss health data, Felder (2005) projects an increase in health expenditures of 24 percent by 2025 in the absence of time-to-death adjustments, and 19 percent if time-to-death adjustments are made (Felder, 2005). The impact on health expenditures of ageing populations is lower, however, if the health status of the elderly improves over time; in other words, if those aged 65 years in 2025 are as healthy on average as those aged 60 years today, then ageing simply shifts rightwards the flatter part of the age-expenditure curve instead of increasing its slope. It is, therefore, difficult to quantify the impact that an ageing population will have on health expenditures, though there is little doubt that an increasing proportion of people over 65 will exert some upward pressures on healthcare costs.

Figure IV.16. EU8 Countries: Population over 60, 1995-2025.

30 20000 ) 25 tl 15000 o ( t '000 a 20 f l

t 15 10000 u p

o 10 u pation

5000 p

5 oation percen o pl 0 0 1995 2000 2005 2010 2015 2020 2025

Source: Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, World Population Prospects (2004 Revision) and World Urbanization Prospects (2003 Revision).

249. A critical issue is long-term care for the very old, which can become a huge financial burden as informal family-based care begins to decline. However, total healthcare costs may decrease if the elderly are placed in less-expensive nursing home beds, which reduce the use of higher-cost acute care beds; or if the use of technical advances, such as modern anesthetics, reduces surgery-related risks for older patients (Cutler, 1994). But increasing long-term bed capacity does not necessarily translate into inpatient cost-savings through less acute care beds. In Switzerland, for instance, the declining number of acute care beds has been accompanied by a steep increase in long-term care beds. Likewise, to respond to the needs of an increasing share of the elderly, Austria anticipates an increase of about 13,000 beds in nursing and residential homes by 2010 and a growing demand for health and nursing staff with advanced training. As a result, health expenditures for long-term care are expected to rise (HIT Austria, 2001). Medical Technology 250. Widespread and extensive use of new and expensive medical technology has already made it a major cost-driver in the health systems in many EU15 countries, and the rapidly growing EU8 countries are poised to witness this proliferation in the near future. This is not to imply in any form or manner that the use of medical technology is necessarily or always wasteful (indeed, medical technology is often the only effective solution and has led to significant improvements in health outcomes); it is simply that the trend of consumption of expensive medical technology is a rising one, and will rapidly become a major cost driver in the health sector in EU8 countries. Felder (2005), in a study using Swiss health expenditures data, 116 assumes that health technology will increase expenditures by as little as 1 percent per year, and projects an increase of over 100 percent in total healthcare expenditures by 2025.

251. The last 15 years have seen particularly dynamic growth in the adoption of new technology in the health sector in many OECD countries. The number of Magnetic Resonance Imaging units, for instance, increased three-fold from 1.7 per million population in 1990 to 6.5 per million population in 2000 (with Switzerland and Austria seeing particularly high increases during this period), and the number of CT scanners increased from 10.1 per million population in 1990 to 17.7 per million in 2000 (OECD, 2003), led again by Austria and Switzerland. From 1989 to 1996, the number of CT scanners increased more than 400 percent and MRIs by 1,000 percent in Austria; by 1996 there were 112 CT scanners in hospitals and 81 CT scanners in private practice, and 32 MRIs in hospitals and 28 MRIs in private practice (Wild, 2004) (Figure IV.17).98

Figure IV.17. Diffusion of MRI Units and CT Scanners, Selected Countries.

16 30 e l

25 p o

12 e o p 20 n lin mlli o t i aon l e

u 8 15 p ill p i o e tr pi i

10 p n m

us 4

5 CTs r 0 0 United Switzerland Spain Germany France Canada Austria Kingdom MRI, 1990 MRI, 2000 CT Sc anners , 1990 CT Scanners, 2000

Source: OECD Health Data 2004.

252. There is no doubt that many technical innovations contribute to cost-saving (e.g., when drugs reduce the need for surgery), but concerns about the impact of technology on increased healthcare costs have led to regulations in the use of high-end technologies in most EU15 countries.99 In the 1960s and 1970s, the Dutch government attributed the main cause of rising costs to the construction of new hospitals and the steady expansion of health technology. The Netherlands has highly-developed policies to regulate technology in its health system with multiple payers and private providers. Dutch hospitals can be denied the purchase and use of new technology through a formal decision made by the health minister, after advice from the HealthCare Insurance Board and the Health Council. This measure has prevented oversupply and stimulates effective use of technologies. Hospitals that ignore the regulations are subject to sanctions and are not reimbursed by the sickness funds (HIT Netherlands, 2004).

253. Countries with tax-based systems such as the UK (National Screening Committee) and Spain have national agencies for high technology and technology control measures in their national health policies (Banta, et. al., 2001). In Germany, the multiple social health insurance funds and hospitals are free to develop services and purchase technologies; however, cost concerns have led to regulation proposals. Switzerland does not have a coherent policy towards technology; although coverage of services is restricted under social health insurance. Market entry for high-cost technology is regulated in Austria, where the purchase of high-technology medical equipment by public hospitals has to be justified on the basis of “medical needs” (HIT Austria, 2001). In France, the distribution of high-cost technologies is based on norms described in the national medical map (HIT France, 2004).

98 The tax-funded health systems of Spain and the UK have the lowest density of MRI units and CT Scanners, and also report the lowest hospital capacities and inpatient expenditure shares, as well as the lowest total health expenditures. On the other hand, Switzerland, with the highest hospital numbers, inpatient cost shares and total expenditures, is also among the countries with a large number of MRI and CT units. Interestingly, high technology diffusion does not correlate with increasing health expenditures in Austria. 99 Also see Weisbrod (1991) and Katz et al, 1996. Weisbrod found that the use of new technology was the main reason behind the Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 117 6. Directions for Further Reform 254. The systemic debts in the health sector in the EU8 countries are indicative of a broad range of underlying issues that threaten the sustainability and effectiveness of the health sectors in these countries, and which will have to be addressed if costs in the health sector are to be managed. These issues are well-understood and have been extensively discussed in most EU8 countries. Likewise, the solutions to the proximate problems facing the health systems in the EU8 countries are also known, and the problem is much more of taking cognizance of the urgency of the situation, and mustering the necessary political will to stand up to those who are likely to be adversely affected in the changing environment. Three examples will illustrate these points.

255. Consider first the generous benefit package that most EU8 countries provide. The constitutional guarantee of free healthcare in most EU8 countries not only creates a very strong sense of entitlement, it also provides the political basis for the strong resistance frequently observed in many countries to co-payments or to health insurance and provision schemes that differentiate or explicitly ration access to health services. It is not surprising, therefore, that most EU8 countries chose to maintain universal coverage and free access to health services, even as large-scale structural reforms were being initiated in other facets of healthcare. While the presence of informal payments in many countries has compromised the theoretical ‘free access’ to health services, few countries have taken the politically difficult step of formalizing the existing flow of informal payments. Facing little or no direct costs at the point of utilization of health services, patients have few incentives to rationally determine their health consumption basket, an unsurprising result of which is the high level of consumption of health services in many EU8 countries.

256. Second, despite a decade of reforms, the underlying supply-side incentives in the health system continue to be weak and generally ineffective, almost matching the near-absence of demand-side incentives which promote cost-consciousness. While most countries have introduced risk-sharing at the primary care level by paying the providers on the basis of capitation, similar incentives are by and large absent at the secondary and tertiary levels. Most hospitals in the EU8 countries are owned by the local or central governments and most hospital managers have no claim to any residual balances that may result from good management of the facilities. Prices for health services are typically set at the level of the central government, usually at levels very close to the costs of production and delivery of these services, and with few rewards and penalties for staying within/breaching the budget, health providers pay scant attention to the cost side of the equation.

257. Third, expected gains from shifting to a social health insurance system have not fully materialized, with insurance companies in many countries behaving simply as collection and forwarding agencies, instead of as rational purchasers of health services on behalf of the insuree. In most countries, insurance companies face very few cost-saving incentives and contract with all hospitals at pre-determined prices for a centrally-defined basket of services.

258. The solutions to these three problems are obvious and include some combination of: (i) rationalization of the benefit package; (ii) introduction of co-payments or other mechanisms to make people take greater responsibility of their health and healthcare-seeking behavior; (iii) creating competition among healthcare providers; (iv) consolidating and closing public hospitals and reducing hospital beds; (v) risk-pooling and rational purchasing of health services; and (vi) introducing provider payment mechanisms that promote incentives for cost management and quality enhancement.

259. The fact that the health systems in three of the eight countries of the region seem to have found some working solutions and do not have any debts is remarkable, and it is worth exploring why the same set of issues that have been the bane of fiscal balance in some countries have not been a problem for others. Unsurprisingly, Estonia, Slovenia and Latvia faced - and continue to face - the same financial pressures from increasing pharmaceutical costs, extensive hospital infrastructure, and increasing demand for higher salaries, but have found simple and yet effective ways to address these problems. Slovenia found the answer in good governance, close scrutiny, and informed oversight. Estonia found the solution in strict adherence to financial rules and in maintaining financial reserves to meet unforeseen expenditures. Latvia contained its expenditures to the resource envelope, and simply did not spend what it did not have. To be sure, all three countries also carried out a host of reforms to control pharmaceutical spending, reduce 118 hospital infrastructure, improve hospital management and so on, but the underlying difference between these three countries and the others is that of sheer fiscal discipline.

260. The most important message from the experience of Estonia, Slovenia and Latvia is that there is no substitute for fiscal discipline, and that the best way of not running up arrears is to stay within the available resource envelope. The resource envelope available for the health sector is usually sufficiently elastic – definitely so in the middle-income EU8 countries – to accommodate the needed expenditures in a way that fiscal balance is maintained and quality health services are produced to meet the healthcare demand. However, such accommodation is possible and sustainable only if backed by a fundamental adherence to good budgetary practices and fiscal prudence.

261. In addition to the continuing financial pressures from rising pharmaceutical expenditures, rising salaries of health personnel, expenditures associated with maintaining the existing hospital infrastructure, aging populations and proliferation of expensive medical technology are bound to place a greater burden on the already fiscally-stretched health systems in the EU8 countries. In many countries, the challenge of containing costs will require systemic changes in the way healthcare is produced, financed and delivered, as well as institutional changes in the way that the health sector is organized and managed. This was the case, for example, in Slovakia, where a set of comprehensive reforms was put in place to meet the fiscal crisis head-on and find a sustainable solution (Boxes IV.2 and IV.3).

Box IV.2. Health Reforms in Slovakia The Stabilization Phase of the health reforms in Slovakia consisted of three measures: the introduction of user fees; changing the methods of purchasing pharmaceuticals; and hospital restructuring. As a result of these measures, the annual recurring health sector deficit has been eliminated, though the stock of debt remains high.

Measure Description Result Introduction of Modest fees (SK20 per visit, SK50 -10% reduction in outpatient visits copayments per hospital bed day, SK20 per -Minimal adverse impact: only 1.5% of those prescription); providers allowed to surveyed stopped physician visits retain co-payments. -Physicians official earnings increased -Drop in corruption (% of respondents associating health care with corruption dropped from 32% in Nov2002 to 10% in Jan2004) Changes in Increased transparency in -Reduction in drug prices and prescription pharmaceutical procurement; price negotiations volumes (e.g., price of Risperidone, an procurement over internet; change in the antipsychotic agent, fell by 76% between procedures process of setting max prices. Nov2003 and Oct2004) -Substantial slowdown in growth of expenditures, from 16.5% annually in 2001 to 8.9% in 2003 and -11.7% in 2004 Pilot projects of Decentralization; consolidation of -Reduction in hospital beds hospital hospitals in Bratislava and Banska- -Reduction in hospital employment restructuring Bystrica. -Sale of hospital buildings Systemic reforms were started under Phase 2, and include the passage of supporting laws. Six health- related Acts were passed by the Parliament in Nov 2004: (i) the Act on Health Insurance, to regulate social and private health insurance; (ii) the Act on Health Insurance Companies and Healthcare Surveillance Authority, to introduce budget constraints, transparent financial relationships, accounting, and auditing; (iii) the Act on Providers, to introduce increased decision-making independence and autonomy of individual healthcare providers, and their responsibility for consequences of their decisions; (iv) the Act on Ambulance Services, to organize and integrate emergency service; and (v): the Act on Healthcare, to unambiguously define healthcare and the forms of its provision, rights and duties, handling of health documentation, and regulation of provision of health-related services; and (vi) the Treatment Act, to define and develop the scope of services covered under social health insurance. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 119

262. Looking forward, the social health insurance system in the EU8 countries will come under even greater pressure unless urgent action is taken to address at least the proximate causes of rising health expenditures. At the very least, the health reform package would have to consist of a combination of stricter supply-side measures, such as management of pharmaceutical expenditures and hospital restructuring, and demand- side measures, such as greater patient responsibility for their own health and greater patient contributions, including cost-sharing for pharmaceuticals. The ultimate objective of these solution-measures in the short- run has to be the stabilization of the fiscal situation so that no new debts are created in the health system. Likewise, the emphasis in the medium term would necessarily have to be on improving efficiency and effectiveness, ensuring access to healthcare, enhancing quality of care, sustaining cost-containment, and ensuring that adequate legislative support is available to sustain the short-term measures and medium term reforms. Managing Pharmaceutical Expenditures 263. The single biggest reason for rising pharmaceutical costs is the increase in volume of drug use, and substantial savings in pharmaceutical expenditures are possible by controlling the prescriptions and use of drugs. The changes in drug expenditures relate to changes in the volume of drug use and prices for new drugs entering the market. The experience of OECD countries has shown that increases in the volume of drugs and the introduction of new products, rather than increases in prices, contributed to the growth of pharmaceutical expenditures (Figure IV.18). The UK, Spain and the Netherlands experienced high percentage increases in both volume of prescribed drugs and in pharmaceutical expenditures. In Canada, the Patented Medicine Quantity Index, a measure of patented drug volume, increased annually by 12.4 percent between 1988 and 2001, considerably more than the average annual drug price increase of 0.8 percent.

264. Drug price regulation is important and necessary, but the experience of many countries shows that it is not enough by itself to control drug expenditures. Many OECD countries have indeed successfully implemented reform measures aimed at price control. In the Netherlands, the pharmaceutical price index declined at an annual average of -1.3 percent from 1996 to 2000, mainly due to price regulation that led to a fall in prices for “old” drugs, and has remained more or less unchanged since. Prices of old drugs fell in Switzerland as well, though the decline was almost fully compensated by higher prices for new drugs at market entry. In the UK, pharmaceutical prices are highly regulated and prices for existing drugs cannot be increased (however, the UK also has the highest pharmaceutical prices in Europe).

Figure IV.18. Price and Volume of Pharmaceuticals, Selected Countries, 2002.

14

12

10

8

6

4

2

0

-2 UK Spain Netherlands Germany France

% Change in price of existing drug % Change in new products entering the market % Increase in volume of prescribed drugs Total Grow th in Drug Expenditures (%)

Source: Nefarma (2002). 120

265. However, most of these measures have not been successful in controlling the high growth of pharmaceutical expenditures. In Austria, for instance, spending on pharmaceuticals grew 1 percent faster since 1990 than the average annual growth rate of total health spending, mainly driven by the increased use of innovative and more expensive drugs. To create competition between pharmacies, Austria relaxed the conditions for establishment of pharmacies based on needs assessments in 1998, which was expected to create an increasing number of pharmacies leading to lower prices (HIT Austria, 2001). However, pharmaceutical expenditure as a percent of total expenditure has continued to increase.

266. In addition to price controls, the regulation of the consumption of pharmaceuticals is critical in order to contain expenditures on drugs. Most countries have adopted demand-side measures for controlling consumption, and cost-sharing has proven to be the most effective measure. In the Netherlands, for example, the introduction of co-payments on prescribed pharmaceuticals (a fixed amount per prescription), led to a substantial decrease in the total number of prescriptions. When this was compensated by an increase in the prescription size (pharmaceuticals per prescription), and in quantities prescribed, the government responded by limiting the number of drugs per prescription and allowing virtually unlimited cheaper prescriptions (HIT, Netherlands, 2004). Reference price systems have been introduced in Germany, the Netherlands, Spain, and France to increase cost-sharing for individuals using branded or higher cost products while assuring access to less costly generic drugs (OECD, 2003). In Germany and France, cost-sharing to control pharmaceutical expenditure has led to an increase in the number of drugs excluded from insurance coverage (e.g., “comfort” drugs or those without proven therapeutic value). In Germany, drug cost-containment measures take the form of cost-sharing, prescription limitations, reference prices, and the pharmaceutical spending cap that makes physicians’ associations liable for any overspending with no upper limit. These measures led to substantive decreases in pharmaceutical expenditures for social health insurance, mainly attributable to price reductions, changes in physicians’ prescribing behavior (resulting in a reduced number of prescriptions by 11.2 percent), and increased prescriptions for generics (HIT Germany, 2005). The French government imposes a fine on pharmaceutical companies if pharmaceutical expenditures surpass budget ceilings, either due to price or quantity increases (HIT France). Spain introduced negative lists, generic drugs list and global reference prices, and reduced wholesaler profit margins from 11 percent to 9.6 percent (HIT Spain, 2000). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 121

Box IV.3. Managing Expenditure on Drugs – the Experience of Slovakia In order to control pharmaceutical expenditures, the government of Slovakia launched an ambitious drug policy in 2003, employing a series of measures aimed at controlling both the price and the quantity of drugs. These included: (i) the introduction of a flat prescription fee (SKK 20); (ii) the introduction of a fixed ratio after the categorization has taken place; i.e., if the pharmaceutical company decreases the price of a drug after the positive list is published, then the ratio between the reimbursement, paid by the health insurance company, and the co-payment, paid by the patient, remains the same; (iii) the introduction of a mechanism where insurance companies reimburse patients on the basis of the lowest price in every therapeutic category. The lowest price, in turn, is determined on the basis of daily dose requirement and published in a handbook that is widely circulated among pharmacies. The pharmacies are required under the law to explain to the patient the substitutability and availability of drugs, and the different co-payments associated with them. Patients choosing the higher-priced drug make the higher co-payment as determined in the handbook; (iv) open competition among pharmaceutical providers, which is conducted on-line so that all bidders have complete information about the bids of their competitors as well; (v) the introduction of the “fast track” process, whereby – if a pharmaceutical company decreases the price of a product by 10 percent or more, compared to the cheapest drug in the cluster – one step in the registration procedure (evaluation by the Categorizing Committee) is dropped.

Figure IV.19. Expenditures on Drugs in Slovakia.

500 40

400 30 in 20 ge

l 300 i

m 10 o

200 centa r

0 e p euro l 100 -10 0 -20 1996 1997 1998 1999 2000 2001 2002 2003 2004

Expenditures on Drugs Annual Growth in Expenditures

Data supplied by health insurance companies show a substantial slowdown in the growth of expenditures allocated to drugs. While in previous years that growth was regularly in the double digits, in 2002 and 2003 it dropped to single digits. Figures for the first half of 2004 were also encouraging, with drug expenditures falling by 11 percent year-on-year (Figure IV.19). However, preliminary expenditure numbers reported by one insurance company (VsZP) for the first four months of 2005 seem to suggest that pharmaceutical expenditures are on the rise again, and could even surpass the 2003 levels.

267. Lowering drug costs requires a multi-faceted approach, necessarily involving the policy-makers working together with payers, doctors, pharmacists, and patients, and involving a package of measures aimed at pricing, prescription, and safety of drug use. As the experience of many countries shows, price controls are necessary but by no means sufficient to contain pharmaceutical costs. Controlling consumption is critical, either by demand-side measures such as cost-sharing, or through supply-side interventions, such as managing physician prescription practices. The latter is clearly more difficult, but is becoming more important as the cost-sharing limits are being reached. Rationalizing the benefit package 268. The move from general tax financing to health insurance financing in the EU8 countries did not significantly reduce the scope of services provided free of charge to the patients. It remains important to clearly and unambiguously define which services would be covered by the health insurance, and which services would not. In order to ascertain the appropriate use of available and mobilized resources for the health sector – including non-public sources such as private out-of-pocket payments and private insurance premiums – and in order to structure the system to effectively use these resources, it is necessary for the health system to provide more specificity with respect to the scope of publicly covered services. Without such specificity, it is also difficult to ascertain the role that might be played by private supplemental insurance and 122 other non-public resources. The absence of clarity of services covered also contributes to the pervasiveness of informal payments. If people are not aware of the precise benefits to which they are entitled, they will be more susceptible to requests for additional payments. Clarification of what is covered will also help strengthen the budgeting process, as it will enable the government and healthcare providers to make accurate estimates of future expenditure needs.

269. Creating a universally acceptable benefit package to be funded from public sources is an attractive, but in some sense impossible mission. Therefore, health insurance funds (or governments) tend to exclude some treatments or entitlements, which do not solve the problem. For the fiscal sustainability of health systems, a flexible mechanism needs to be in place that allows for a benefit package that responds to a population’s needs and can be serviced within the given budget constraints.100 Introducing private insurance101 270. Private health insurance can play several roles within the health care systems of the EU8 countries.102 Under one common option, private health insurance supplements coverage provided by the government or social insurance fund(s). Examples of countries with supplemental private health insurance include Australia, Canada, France, Spain and the United Kingdom.103 This supplementation can take several forms. If a benefit package does not include certain commonly desired coverage, the insurance coverage can complement the government benefits. For example, the insurance contract may guarantee reimbursement for the cost of specific health services omitted by the public benefit package. Alternatively, or together with this type of supplementation, the insurance can provide reimbursement for any costs associated with receiving services covered by the public system.104

271. Another option for private health insurance is to cover a range of benefits, including comprehensive coverage. The scope of benefits is determined by the coverage contract, as well as by the applicable laws and regulations. There may be circumstances in which this coverage serves as an individual or family’s only protection against the cost of expected and unexpected health services. This can occur if there is no social insurance system, individuals are not eligible for coverage under such a system, or they are permitted to opt out of the social insurance system if they purchase private coverage. The latter type of coverage, referred to as “substitutive coverage,” is an option in the Netherlands, Switzerland, Germany, and Chile. These comprehensive packages may have significant or minimal cost-sharing. The consideration of a “substitutive” option raises several issues and must be approached carefully. This type of system can greatly impact the distribution of risk between the public and private system, and once created, it may be difficult to alter at a later date.105

100 In Slovakia, the Scope of Benefits is derived from the principle that an insured person has the right to equal treatment in case of an equal need. Due to the infinite nature of needs, it is however necessary to define a certain maximum extent of care, based on the list of priorities that are in line with the fiscal capacity of the Slovak economy. The priority list is a positive list of diagnoses where there is zero co-payment of insured patients. The list of priority diseases contains approximately 6,700 diagnoses, which is almost two thirds of the total list of diagnoses (11,000) listed in ICD-10. The list of priority diagnoses is adopted by the Parliament on government proposal. All diseases are subject to the process of cataloguing, where they would be assigned a list of interventions fully reimbursed from public healthcare insurance. Standard diagnostic and therapeutic procedures are thus created. For diseases not listed on the priority list, the extent of patient co-payment is determined for all interventions. 101 This section draws heavily from a background note on the subject prepared for the author by Tapay, Nicole (2001). Also see Colombo and Tapay (2005). 102 Experts have developed somewhat different categorizations of private insurance products but these often seem to encompass a similar range of functions as the categories described herein. 103 In Australia, private insurance plays a dual role. It can supplement the coverage under their social insurance fund, by covering services not covered by the public system, but it can also provide coverage that is “parallel” to the publicly available coverage, replacing the public funding. As an example of its supplemental role, this coverage can pay up to 25% of the fee to cover the gap in social insurance coverage for medical services provided in hospitals. 104 Note that private supplemental coverage of the same benefits otherwise covered by the public or social insurance system, particularly private coverage of cost-sharing expenses associated with the receipt of certain publicly covered services, may impact utilization. In the U.S., for example, researchers have found that private Medicare supplemental coverage of cost-sharing expenses related to the receipt of benefits otherwise covered under Medicare, may have resulted in increased utilization by those holding these private policies, except in the case of coverage offered through managed care HMO plans (see Christensen and Shinogle, 1997). 105 For example, Chile provided its employed population with the option to obtain coverage through the private funds (ISAPREs) in lieu of coverage through the public social insurance fund (FONASA). The regulations focused on contract compliance rather than content. This system has resulted in a higher proportion of the younger and wealthier population receiving coverage under the private funds. This fragmentation occurs despite the fact that there are significant additional revenue streams (in addition to the payroll tax) for both the public and private aspects of the health care system. FONASA receives 60% of its revenues from general taxes, and the ISAPREs are able to charge additional premia with few limits. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 123

272. Another coverage option is parallel coverage. Such coverage may or may not offer a comprehensive benefit package. Rather than focusing on filling particular gaps in a social insurance system – as is the case with supplemental coverage – this coverage may offer access to higher quality or more comfortable accommodations in connection with the same health care services also covered by the social insurance system. It also may enable individuals to bypass any waiting lists that develop. Examples of such “alternative” private insurance can be found in Ireland, Great Britain and Spain. Since this type of coverage also complements social insurance systems, it is sometimes also referred to as a type of supplemental coverage.106

273. A report of the European Parliament’s Committee on Employment and Social Affairs argued for some minimum rules for the supplementary (or complementary) health insurance marketplace in order to reduce certain inequities in access to this type of private coverage. The report noted that that “[i]t is both necessary and sensible to oppose a trend which, supposedly on the grounds of economy, good management and making people more responsible for their own welfare, has the effect of reducing high-quality health cover and transferring certain risks…to the supplementary schemes…The proportion of the cost which has to be met by the consumer….is becoming too high for people on low incomes…if supplementary health insurance organisations are to play an increasingly important role in providing health care, goods and services for the public, the sector must be subject to rules to ensure nobody is excluded and everyone has access to high-quality necessary treatment.”107 Also, the ability of certain segments of society to receive services through private coverage may impact the political will to improve elements of the social insurance system.

274. A private health insurance market has the potential to offer certain advantages over solely public sources of health care financing. If a government chooses to target public coverage to certain population groups, rather than provide universal coverage, the availability of a private insurance option becomes a critical component of a nation’s health financing scheme. Where the government assures broad coverage, private insurers may relieve some portion of the burden from the public system of financing (and depending upon the delivery system, it also may relieve the public delivery system). The extent to which the private system is able to relieve a burden from the government depends on many structural aspects of the system, such as whether the privately insured are still covered through some governmental funds (whether from general revenue or payroll taxes) as well as whether private insurers are paid an additional amount that is separate from the funding stream for public coverage. The potential for relief of the governmental burden also depends upon the incentives and regulation that can impact the division of the population between the public and private systems. If the private system is permitted to cover a disproportionately younger, wealthier or healthier population, with little regulation, it may raise significant issues for the public system. 108

275. Another potential advantage of private health insurance is that it may help providers rebuild infrastructure and amortize needed investments. The combination of many factors may impact whether private financing can positively and broadly impact providers’ funding and access to high quality care. Relevant issues include the risk composition of the insurers’ covered populations, the presence or absence of a risk adjustment or cross-subsidization mechanism among insurers and between the public and private system, and the extent to which there are existing inefficiencies in the system that private insurance could address (and that could not otherwise be addressed through public financing).

276. Finally, private insurers may promote innovation in financing and delivery, help offer consumers a broader choice of providers and create incentives for more efficiency in aspects of the system. Innovative methods of provider payment may emanate from the private system. If the public system does not assure

106 See e.g. the discussion of two types of voluntary health insurance (VHI) in the European Union (EU) in Rocard, M., “Report on supplementary health insurance,” European Parliament, 4 October 2000, Committee on Employment and Social Affairs. This report divides voluntary health insurance in the EU into two categories, that which is supplementary to public entitlement, and that which substitutes for, and is mutually exclusive from, the statutory health insurance scheme (as in Germany and the Netherlands). 107 See Rocard (2000), at 16-17. 108 For example, in Ireland, a survey of the Organization for Economic Co-operation and Development (OECD) found that “private health insurance has operated in a way that tries to ensure that a significant number of people stay in the private system, relieving the cost of hospital care to the public finances.” Department of Health and Children, Ireland, “White Paper: Private Health Insurance” (Dublin, 1999) at 14, citing OECD, “Economic Survey of Ireland 1997.” 124 consumers true access to the providers of their choice, there also may be a demand for a system of financing that does a better job of assuring consumers access to high quality providers of their choice. This might occur if some providers selectively contract with private insurers.109

277. Yet, despite the possible advantages of private insurance, there are several problems that can arise from its introduction. Adverse selection – the migration of high risk cases to certain insurers or insurance products – represents a significant risk in the sale of private insurance, particularly when the purchase of health insurance is voluntary. Even if predictably high cost cases are spread somewhat evenly among products and insurers, if the risk is segmented among a significant number of insurers and products proportional to the population, the occurrence of a few unanticipated high cost cases can cause an unstable situation for a particular product pool or insurance company. There also are significant incentives for private health insurers to engage in “cream-skimming,” the practice of covering low-risk persons and not serving a broad range of risk profiles. This can be done by refusing to offer or renew coverage to high-risk individuals or charging significantly higher rates to such individuals (thereby reducing the likelihood that they will purchase coverage and making them pay a higher premium if they do). Another common method of reducing the risk they cover is to exclude coverage of certain “preexisting conditions.” This practice can serve the important purpose of protecting insurers against adverse selection. If misused, however, it can enable insurers to reduce the extent to which they cover medium and higher risk individuals’ health needs.

278. Further, rising medical costs can make it challenging for any private financing sector to maintain affordable rates, just as it poses challenges to a publicly financed system. If rates are permitted to vary based upon risk, affordability can become an acute issue for those with health concerns. Depending on the level of market fragmentation and the size and diversity of risk pools, there is also a risk of “rate spirals.” These spirals occur when premiums increase substantially on an annual basis due to an increasingly high risk case load within an insurer’s or a product’s pool of covered individuals. This risk may be especially high in relatively small marketplaces or in the case of smaller insurers.

279. Another concern with private insurance relates to the extent to which private insurers use the funds they receive for administrative purposes. Private insurers’ expenditures on administrative and non-benefits- related costs, such as marketing, sales, agent commissions and profits (in the case of for-profit entities) may be in stark contrast to the percent of premiums paid on administrative costs by social insurance funds.110

280. Informing consumers can also be a challenging process, particularly within a complex private health care insurance marketplace.111 The ability of insurers to offer many different products may increase choice and benefit competition. However, in the absence of standardization of policies, and even in its presence in some cases, consumers may purchase duplicative coverage.112 Meaningful information disclosure and dissemination can be a challenging but important piece of a successful private health coverage marketplace.

281. Many of the risks and disadvantage of private insurance can be addressed, at least to some extent, through the imposition of appropriate, enforceable and enforced regulatory requirements. While a detailed and exhaustive discussion of regulatory structures and options are beyond the scope of this paper, it is useful to list some of the essential elements of such a regulatory system.

109 Chollet and Lewis (1997). 110 For example, in the U.S., administrative costs for the Medicare program (a program for the elderly and certain disabled individuals), as a proportion of overall expenditures, was 2.66% in 1997. In contrast, the administrative and non-benefits related costs associated with private policies in the U.S. are reported to be much higher (Braden et al, 1998). 111 The Rocard report noted that medical insurance products are complicated and, notwithstanding the regulations covering contract terms, individual consumers are unlikely to find health insurance policies easy to grasp. It also noted that coverage under policies are not identical and therefore make it difficult for those outside the industry to compare policies in terms of value for money. 112 This was the case in the U.S. even after standard packages for certain Medicare supplemental coverage was developed. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 125

282. Policymakers should encourage the largest possible pooling arrangements, bringing together individuals of diverse risks.113 Requiring insurers to accept all applicants can be an important way to help address risk fragmentation. On the other hand, in order to help prevent adverse selection against insurers, particularly in a voluntary market, policymakers may want to consider limiting the timeframe during which insurers must accept applicants, such as through the use of an “open enrollment” period. Furthermore, limits on exclusions based on preexisting conditions, particularly if coupled with incentives for people to maintain continuous coverage, can help promote the appropriate use of exclusions to prevent adverse selection, while at the same time preventing their abuse. It also can help promote portability among insurance products. Requirements relating to the renewability of coverage can also help assure that consumers are afforded the ability to continue coverage after they become sick.

283. Affordability of coverage is one of the more challenging issues for both public and private financing mechanisms. Rating restrictions that prohibit or restrict the use of risk factors in the calculation of premium charges can help equalize the costs across the system. However, these mechanisms may result in premiums that discourage purchases by the young and healthy, in which case the governments may need to examine creative methods of risk equalization or other ways to encourage risk distribution. If policymakers wish to address some of the inequities that arise between high and low-income individuals, they may want to consider some method of cross-subsidization based on income. This could include direct subsidies to help lower income individuals and families pay for coverage. Alternatively, the use of a percentage payroll tax may also help cross-subsidization, as may the use of general revenues.114

284. As noted above, the extent to which consumers can compare plans is an important piece of a competitive market. If they are unable to compare the value of packages by readily comparing benefits and costs, it may undermine market forces and also result in the purchase of inappropriate coverage. Policymakers may want to ensure that certain aspects of the plan and its operation are disclosed in the coverage contract. They also may want to help ensure that the information is presented in a readily understood, and if possible, comparable fashion.

285. Requirements relating to the expertise required of those who make decisions about access to care on behalf of a private plan may help address some concerns that arise from plan access restrictions. Additional protections may include requirements that certain decisions relating to access to appropriate care (“utilization review” decisions) are made within a certain time period and that this time period is escalated in the case of emergencies.

286. The establishment of mechanisms for consumers to complain about and appeal insurer decisions – both regarding claims payment and access to care – can help promote confidence in the private system. The establishment of such procedures internal to the plan can be very useful and regulatory standards for such procedures may help assure that issues are resolved within a reasonable timeframe.

287. Ideally, the government should devote resources and technical expertise to assuring that insurance companies are in a financial position to deliver on their promises. In the absence of these requirements, there is significant risk of financial trouble, resulting in insurers’ or providers failing to receive the reimbursement they were guaranteed in their coverage contract. A government also may want to develop a safety net mechanism to protect insurers in such a case, yet this might require the allocation of significant resources to a backup or “guaranty” fund. Governments should examine the financial standards applicable to other types of insurance and ascertain when these might appropriately be applied to health insurance products. It likely would make sense for the relevant government agency to coordinate or utilize existing regulatory expertise in the area of financial standards. Regular reporting requirements, and systems to analyze company financial data and indicate when companies are in trouble, can help avert the hardship of insolvencies. However, these types of reporting requirements can require the development and maintenance of significant expertise at the regulatory level and should not be underestimated.

113 World Health Report (2000): Health Systems: Improving Performance (Geneva, Switzerland: 2000). 114 Chile uses a payroll tax to help promote cross-subsidization among incomes and also includes significant resources from general taxation in funding its social insurance fund. However, membership in the private funds is still skewed towards those with higher incomes. 126

While an extensive treatment of private health insurance is beyond the scope of this paper, it is important to emphasize that private insurance can have a significant impact on an overall national system of health financing and delivery. The potential for significant impact – both good and bad – underscores the need for coordinated rules and structures governing both public and private financing.115 It is also important to keep in mind the potential impact of private financing on the overall costs of a country’s health care system. If the private insurers’ payment structures and arrangements with providers (or absence thereof) result in payment for medical charges with little control over the amount of spending, it can result in runaway spending and impact nationwide spending patterns.116 In addition, since private insurance is not inherently accountable to government and national objectives, “external incentives and regulation are needed to make sure that benefit packages and insurance practices are coherent with national priorities and policies regarding health, financial fairness and responsiveness.”117 Greater Individual Responsibility in Managing Own Health 288. To a great extent, demand for healthcare is controlled by the individual consumer. The key to containing excessive and unnecessary demand for health services lies in making individuals and families sensitive to the costs of additional health spending, while still limiting each family’s maximum outlays to affordable levels. This was the logic behind the introduction of patient co-payments in the health sector in Slovenia and Slovakia, a move that in both countries resulted in significant savings. Despite the logical appeal of such insurance, it has been politically difficult in countries like the Czech Republic and Poland to formally introduce co-payments as a means of sensitizing individuals and families to the costs of additional health spending.118

289. One real concern with co-payments is that it may impose an additional financial burden on the poor, and dissuade them from seeking healthcare when ill. In Slovakia, for instance, out-of-pocket payments have increased by SKK 2.7 billion following the introduction of co-payments, and some of this undoubtedly has come from the pockets of the poor (though survey results show that only 1.5 percent of the population did not seek care because of the co-payments). The adverse equity effects of co-payments can be neutralized to a large extent by exempting the poor and vulnerable from making official co-payments, though it is not easy to set up a system by which the poor and the vulnerable are appropriately identified. The Slovak system managed these equity concerns by giving a monthly grant of SKK 50 to individuals (listed by the government to be poor or vulnerable), in order to defray any additional health expenditures because of the co-payments.

290. In any case, it is important to note that in many countries, patients are already making huge out-of- pocket payments, most of which are of an informal nature (Box IV.4). These informal payments have had an adverse impact on equity and access, since the poor are unlikely to be able to match the informal payments by the non-poor, and thus would always find themselves at the farthest end of the queue of patients for healthcare. While estimates of informal payments vary tremendously across and within countries, and some countries have better estimates than others, there is a general consensus that widespread informal payments in the health sector in many EU8 countries adversely affect utilization of services, and introduce an element of uncertainty in the transaction between the patient and the provider. The pervasiveness of informal payments in health in many EU8 countries has actually become a serious impediment to healthcare reform. Besides contributing to the general environment of corrupt practices and the growth of a parallel healthcare financing system, informal payments introduce perverse incentives in the health system, and compromise government efforts to improve efficiency, accountability, and equity in the delivery of health services. The non-transparent and discretionary nature of informal payments adversely affect access to healthcare, particularly for the more vulnerable segments

115 The World Health Report (2000) noted the contradictory incentives of private insurance and social insurance. “The social security and risk-related private insurance approaches are contradictory, and their coexistence creates different incentives for consumers. All consumers whose risk category is such that private insurance would charge them less than the amount they would have to pay under social insurance have the incentive to avoid contributing to social insurance and use private insurance if they are allowed to. High-risk people, however, have the incentive to contribute to social security, loading it with high-risk members and increasing the per capita cost of services for members of the pool.” World Health Report 2000 (Geneva, Switzerland: World Health Organization) at 109. 116 Provider-driven cost escalation occurred in the U.S. prior to government regulation in the 1970’s and the regulation of the doctor- patient relationship by market forces in the 1980’s (Rosenblatt 1999). 117 World Health Report (2000), at 111. 118 Those opposing this move point toward the adverse effects of patient co-payments on the poor, for whom out-of-pocket payments would constitute a greater burden in terms of share of income, and go on to argue that even those who could otherwise afford to make the co-payments would tend to defer treatment until the problem becomes more acute, and perhaps more expensive to treat. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 127 of the population who have to pay disproportionately large amounts for health services that are supposed to be available free of charge. The act of asking for and receiving informal payments cannot be entirely pleasant for all providers as well, many of whom are also concerned by the unethical nature of this practice.

291. In a novel approach to facilitate the inclusion of people into the decision-making process and make them assume greater responsibility for their own health, a group of reformers in the Czech Republic has proposed the integration of social health insurance with Personal Health Accounts, through which people are enabled to become respected partners to health insurers and healthcare providers in a system of social (or publicly funded) health insurance. The Czech proposal views such Personal Health Accounts as a means to include people in the decision process concerning the healthcare system while preserving high solidarity, general accessibility to healthcare, and respecting the limited information patients have concerning their disease and possible available treatments. According to the Czech proposal, the Personal Health Accounts represent a way in which personal preferences can be revealed and people can exercise their freedom of choice.119

Box IV.4. Polish Health Service Goes Private The Polish healthcare system today is reminiscent of communist times, when commercial stores began to emerge: it appears to be public, but in fact it is private to a large extent,” writes the author of an article in Wprost weekly. “The constitution of the Third Republic of Poland copies the communist provision concerning free healthcare. However, nearly all Poles are aware that free medical care in Poland is a fiction. Poles spend more than ZL 51bn on healthcare, over 6 percent of GDP. At least ZL 22bn of the total (43 percent) is spent in private medical centers, or in the form of bribes in public hospitals. Half of the remaining difference (held by the National Health Fund - NFZ) is also privatized in a number of ways. This means that in fact some 70 percent of Polish healthcare services have been privatized. If so, what prevents the Polish healthcare system from being fully privatized? In a poll conducted by the Stefan Batory Foundation, doctors were named the second most corrupt professional group, just after politicians. Thirty-six percent of respondents admitted having offered bribes to medical personnel. Bribes are a form of privatization of healthcare services in Poland. Fortunately, there are also other, more legal ways. Dentistry is nearly fully privatized. At the beginning of the 1990s, 12 percent of Poles had their teeth treated at private dentists; now it is 80 percent. More and more often it is hospitals that are being privatized, surprisingly mainly those situated in small towns. A hospital in Kwidzyn in Poland’s north-western Pomerania region used to be constantly up to its ears in debt. At present, it breaks even, as Zdrowie (health) company. The local authorities are satisfied with the privatization as they receive the hospital’s financial statement every month and have control over the money spent. The local authorities managed to pay off a ZL 6m debt and the hospital opened four new wards. A number of other hospitals are planning to be privatized soon. The local authorities of the southern town of Kielce offered hospital personnel a lease on the hospital. Starting in January, they will manage the hospital, while the town will pay its debt. To date, 50 out of 700 Polish hospitals have been transformed into commercial companies. Hospital privatization is underway, though the Act on public support for healthcare services does not assume financial support for privatized medical centers. What is more, the provision granting them the possibility to become commercial companies was removed from the bill. Therefore, hospitals are privatized on the basis of the Healthcare Service Centers Act of 1991, which allows them to function as public or non-public institutions and as commercial companies. Opponents of healthcare privatization say that the process will lead to the closure of expensive, though essential, wards or will augment hospital fees. So far this grim scenario has not materialized though. Privatized medical centers offer a wider range of services to attract patients and increase revenue. The fees are regulated by the market. The hospital in Kwidzyn has opened an expensive emergency unit. In Ozimek in south-western Poland, where the hospital was purchased by EuroMediCare Company (EMC), 90 percent of revenue comes from the contract with NFZ, with 10 percent from diagnostic examinations that patients pay for. Healthcare service privatization has a large group of opponents because it reveals the way money is wasted in this sector. One of the reasons is excess personnel in hospital administration. A greater number of hospitals in Poland could be privatized and thus become more patient-friendly, were it not for a strong lobby of professors and heads of hospital wards, who earn in the shadow zone and block privatization. Hospitals more and more frequently tend to use outsourcing in order to reduce costs… In total, firms reduced costs by 20 percent thanks to outsourcing. The question is why Poland sticks to public healthcare, if private services seem beneficial to all?”

Source: Wprost, August 7, 2005.

119 Personal Health Accounts, or Medical Savings Accounts as they are also known, are thus individual savings accounts that are restricted to spending on health or medical care. They have generally been introduced: (i) to encourage savings for the expected high costs of medical care in the future, because average income and capability to save for an average person are usually high through working years compared to retirement; (ii) to enlist consumers in controlling costs; and/or (iii) to mobilize additional funds for health systems (Hanvoravongchai 2002). 128 Reducing hospital infrastructure 292. Almost all the EU8 countries have an excessive supply of health facilities, including hospitals and hospital beds, and almost all have made attempts to consolidate the many hospitals and reduce the number of hospital beds. The approach in most countries, however, has been half-hearted, and the problem of excessive infrastructure remains unsolved. At one level, rationalizing hospital beds is a straightforward exercise, made complicated only by the real and imagined political obstacles that may stand in the way of its execution. But the fact that some countries have achieved remarkable success holds promise for others who have probably not taken this problem seriously enough.

293. The successful downsizing of the hospital infrastructure in Estonia is noteworthy, not only for the successful result of the exercise, but also for the scientific elegance and practical simplicity of the approach. In 2000, Estonia developed a Hospital Master Plan to reduce the excess capacity in healthcare supply by setting targets for reducing the number of hospital from 78 to 13 by 2015 (which, following political pressure, was subsequently increased to 21). Capacity requirements were established based on facility optimization and the type of facility (secondary and tertiary) was determined based on the catchment’s population. Demographic and epidemiological models were used to determine the demand for healthcare, with suitable assumptions about the development of medical technology. Simple criteria were used to determine the location of hospitals by defining a constraint that time taken to reach the hospital should be no more than one hour. And finally, separate estimates were made for elderly and nursing care needs. Based on these criteria, three regional hospitals (tertiary care catchment’s populations of 600,000–800,000), four central hospitals (catchment’s population of 100,000–150,000), 11 county hospitals (catchment’s population of 30,000–50,000) and three local hospitals with smaller catchment areas were identified. By 2001, 41 hospitals and outpatient polyclinics had been merged into six networks, which then restructured service delivery. By 2003, 7 facilities in the network had been closed following managerial decisions. The total number of acute care hospitals was reduced to 50 by 2003, and over 1,500 beds had been closed down. At the same time, the average length of stay in hospitals was reduced from 15.4 days in 1993 to 8.5 in 2002, and is targeted to fall to 4.0 by 2015 for acute care.

294. In Slovakia, rationalizing the number of hospitals and hospital beds by consolidating hospitals and reconfiguring the mix of hospital beds is a priority item in the ongoing reform program of the Ministry of Health. A detailed assessment of all hospitals has been carried out and a master plan of hospitals is close to finalization. A network of essential hospitals is being determined and will be strengthened during the rationalization process. Hospitals and hospital beds identified as ‘excess capacity’ in the three big cities of Bratislava, Banska-Bystrica and Kosice are included in the first round of the consolidation process. The Ministry of Health has also established a Hospital Restructuring Fund to support specific capital investment in the hospital sector in the near future.120 Since the hospital restructuring program is focusing on the three big cities of Bratislava, Banska-Bystrica and Kosice where over-capacity is most severe, the rationalization of hospital facilities is not likely to have a negative impact on access to healthcare services. However, the Ministry of Health recognizes that the rationalization of hospital services in rural areas will need to be conducted with caution. This is particularly true for mountainous areas, where travel time to the nearest facility might be severely extended in case of a hospital closure. These facilities are proposed to be included in the network of essential hospitals and protected from closure.

120 In addition, steps are being taken to corporatize the public hospitals and make them fully autonomous in their functions. Corporate hospitals will not have an implicit government guarantee, which will send an important signal to suppliers that they cannot simply continue to lend to hospitals. This is expected to create hard budget constraints for hospitals and compel them to accord high priority to efficiency. Corporate hospitals will be allowed to retain savings and keep revenues, which they may then re-invest in the hospital operation, thus generating additional incentives for hospitals to change their input mix and align inputs more closely with outputs. For hospitals that continue to build arrears, there will be the prospect of bankruptcy or the possibility of selected bailouts by the government or commercial banks in exchange for a rationalization plan. Since the hospital restructuring program is focusing on the three big cities of Bratislava, Banska-Bystrica and Kosice where over-capacity is most severe, the rationalization of hospital facilities is not likely to have a negative impact on access to healthcare services. However, the Ministry of Health recognizes that the rationalization of hospital services in rural areas will need to be conducted with caution. This is particularly true for mountainous areas, where travel time to the nearest facility might be severely extended in case of a hospital closure. These facilities are proposed to be included in the network of essential hospitals and protected from closure. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 129 7. Conclusions 295. The continuing indebtedness in the health systems of the EU8 countries is symptomatic of the deep- rooted structural problems involving all aspects of health production, delivery and finance, and indicative of the underlying faults in the ways that the incentive systems are organized in the health sector. Piecemeal solutions may provide temporary relief but are unlikely to resolve the more fundamental issues in the health sector, and will almost never be sustainable beyond the immediate short-term. The challenge that the EU8 countries face, therefore, is that of carrying out comprehensive reforms, but in a way in which the proximate problems are prioritized to yield quick and immediate results.

296. Managing pharmaceutical expenditures will be the single-biggest challenge that the health reform measures would need to address. As discussed above, while it is no doubt important to closely monitor pharmaceutical prices, the key really is to manage consumption of pharmaceuticals. This would require a multi-faceted approach, one that involves the payers, doctors, pharmacists and patients, and harmonizes both demand-side measures – in the form of greater cost-sharing with patients – and supply- side measures – in the form of managing physician prescription behavior. Several countries have had good experience with demand-side measures, and there is increase evidence to suggest that measures that increase patient responsibility for own health would have to be at the forefront of any health system reform package whose objective is to ensure financial sustainability. Such measures would typically take the form of patient copayments at the point of service, which would have to be designed in way that make individuals and families sensitive to the costs of health spending but limit each family’s maximum to affordable levels.121

297. Another key area of health sector reform necessary in almost all the EU8 countries relates to the management of the excessive supply of health facilities. While many countries have started the process of consolidation of hospitals and reducing hospital beds, a lot remains to be done to address the problem of excessive infrastructure and free up the scarce fiscal space for quality-enhancing initiatives. There is little doubt that the introduction of patient copayments as well as rationalization of hospitals and hospital beds can be politically contentious issues; but there is even little doubt that the costs of inaction on these fronts are very heavy and perhaps unsustainable in any health system.

298. It is worth reiterating that at least three of the eight countries in the region found effective and sustainable solutions by successfully applying simple rules of governance and management to the complex set of problems in the health sector. These countries – Estonia, Slovenia and Latvia – are exposed to the same set of financial pressures as the other countries in the region, namely the pressures from rising pharmaceutical costs, extensive hospital infrastructure, and perpetual demand for higher salaries in the health sector, but have responded strongly by maintaining equilibrium through good governance, strict adherence to the rules of financial discipline, and by simply not spending what they do not have. The single-biggest lesson from the experiences of these countries is that there is no substitute for fiscal discipline, and that irrespective of the nature and extent of the systemic reform measures, the best and certain way of not running up debts is to stay within the resource envelope.

299. The EU8 countries may gain some comfort from the fact that other countries were in a similar state when joining the European Union, but found ways of managing their budgets without bringing about huge destabilizing changes. Consider the example of Austria, which acceded to the European Union in 1995. Constrained by the SGP, Austria placed the reduction of public spending at the top of its political agenda and began focusing on reducing government consumption, controlling healthcare costs, reforming the public pension system, and controlling subsidies. Austria also embarked on a major social sector reform, which led to the freezing of public expenditures on personnel costs, and stabilization of operating costs, social spending and subsidies. In the health sector, spending decreased from 5.9 percent

121 One result of the introduction of co-payments in Slovakia, for instance, is that in the two years since co-payments were introduced, Slovaks have paid SKK 2.7 billion to physicians, pharmacies and hospitals, but survey results show that only about 1.5 percent of the population was affected adversely enough by the co-payments to not seek care when ill. Moreover, equity concerns following the introduction of co-payments in Slovakia have been managed by giving a monthly grant of SKK 50 to individuals (listed by the government to be poor or vulnerable), in order to defray any additional health expenditures because of the co-payments. 130 of GDP in 1994 to 5.4 percent in 2003. While total health expenditures also declined slightly during the same period, the share of public spending declined from 74.4 percent in 1993 to 69 percent in 2003.122

300. Fifteen years ago, the EU8 countries went through the first phase of their transition from centrally- planned economies to market-based ones, and that process was undoubtedly painful for all countries, but rewarding overall. As far as the health sector was concerned, the reforms brought about fundamental changes in financing, delivery and organization, but maintained the legacy of state paternalism insofar as the comprehensiveness of the basic benefits package – which included, and continues to include, many non- essential and ineffective interventions – was concerned. A decade-and-a-half later, the EU8 countries find themselves yet again at the crossroads of difficult decision-making, but this time around they have to give back some of the gains in the interest of fiscal stability. As before, the process will be painful, but as before, the long-term rewards promise to be plentiful.

122 To be sure, the Austrian reforms did have an impact on equity, particularly since the reforms were accompanied by an increasing share of private expenditures in financing total health spending. In a study of household survey data from Canada, Spain, the UK, Germany, Austria and the Netherlands, Van Doorslaer and Masseria (2004) found that the mean number for total physician visits (GPs and specialists) was distributed according to the need for such visits in all selected countries with the exception of Austria, where there was a significant pro-rich inequity in total physician visits. However, the findings reveal that the rich have a significantly higher number of specialist visits per year than the poor in all countries, with the exception of the UK and Netherlands, and to this extent these results of Austria are not so striking. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 131 References: Australian Productivity Commission (1996), “Private Health Insurance, Discussion Draft”, December. BAG, Bundesamt fuer Gesundheit (2005), „Statistik der obligatorischen Krankenversicherung“, Bern, Switzerland. http://www.bag.admin.ch/kv/statistik/d/2005/KV_2003_DE_v030205.pdf Banta, D. et. al. (2001), “Health Policy, Health Technology Assessment, and Screening in Europe”, International Journal of Technology Assessment in Health Care, 17:3, pages 409-417.

Boccuti C, Moon M. (2003), “Comparing Medicare and private insurers: growth rates in spending over three decades”, Health affairs, 22, 2, pages 230-237.

Braden, B.R., Cowan, C.A., Lazenby, H.C., Martin, A.B. et al (Fall 1998), “National Health Expenditures, 1997” in “Health Care Financing Review”, Vol. 20, Issue 1, Washington, DC.

Cabe, Inga and Mukesh Chawla (2005), “Latvia: Fiscal Reform Options in the Health Sector”, Background Paper prepared for The World Bank (2005), “Selected Issues in Fiscal Reform in Central Europe and the Baltic Countries 2005”, Washington DC.

Chakraborty, Sarbani and Jan Bultman (2005), “Slovenia: Fiscal Reform Options in the Health Sector”, Background Paper prepared for The World Bank (2005), “Selected Issues in Fiscal Reform in Central Europe and the Baltic Countries 2005”, Washington DC.

Chalkley, M, Malcolmson J. (2000), “Government purchasing of health services”, in A Culyer and J. Newhouse, “Handbook of Health Economics”, Elsevier.

Chawla, Mukesh and Marzena Kulis (2005): “Poland: Fiscal Reform Options in the Health Sector,” Background Paper prepared for The World Bank (2005), “Selected Issues in Fiscal Reform in Central Europe and the Baltic Countries 2005”, Washington DC. Chawla, Mukesh and Peter Pazitny (2005), “Czech Republic: Fiscal Reform Options in the Health Sector”, Background Paper prepared for The World Bank (2005), “Selected Issues in Fiscal Reform in Central Europe and the Baltic Countries 2005”, Washington DC. Chawla, Mukesh and Peter Pazitny (2005), “Hungary: Fiscal Reform Options in the Health Sector”, Background Paper prepared for The World Bank (2005), “Selected Issues in Fiscal Reform in Central Europe and the Baltic Countries 2005”, Washington DC.

Chollet, Deborah J. and Lewis, Maureen (1997), “Private Insurance: Principles and Practice” ,at 77, 79 in “Innovations in Health Care Financing: Proceedings of a World Bank Conference, March 10-11, 1997”, World Bank Discussion Paper No. 365. Christensen, Sandra and Judy Shinogle (Fall 1997): “Effects of Supplemental Coverage on Use of Services by Medicare Enrollees,” Health Care Financing Review volume 19, issue 1, Washington, DC.

Colombo, F, Tapay, N. (2005), “Private Health Insurance in Countries: The Benefits and Costs for Individuals and Health Systems”, OECD Health Working Papers No. 15. http://www.oecd.org/dataoecd/34/56/33698043.pdf

Couffinhal, Agnes and Joseph Kutzin (2005), “Health Financing in Estonia: Challenges and Recommendations”, WHO Regional Office for Europe, Copenhagen.

Cutler D. (1994), “A Guide to Health Care Reform”, Journal of Economic Perspectives. Vol.8, No. 3, pages 13-29. Docteur E, Oxley H. (2003), “Health-Care Systems: Lessons from the Reform Experience”, OECD Health Working Papers No. 9.

Dor, A., & Farley, D.E. (1996), “Payment source and the cost of hospital care: Evidence from a multiproduct cost function with multiple payers”, J Health Economics, 15, pages 1-21. 132 Ellis, R. (1998), “Creaming, skimping and dumping: provider competition on the intensive and extensive margins”, Journal of Health Economics 17: 537–555

European Observatory on Health Systems and Policies (2000), “Health care systems in transition: Spain”, European Observatory on Health Systems and Policies, WHO European Centre for Health Policy, Belgium.

European Observatory on Health Systems and Policies (2001), “Health care systems in transition: Austria”, European Observatory on Health Systems and Policies, WHO European Centre for Health Policy, Belgium

European Observatory on Health Systems and Policies (2004). “Health care systems in transition: Germany”, European Observatory on Health Systems and Policies, WHO European Centre for Health Policy, Belgium

European Observatory on Health Systems and Policies (2004), “Health care systems in transition: France”, European Observatory on Health Systems and Policies, WHO European Centre for Health Policy, Belgium

European Observatory on Health Systems and Policies (2004) “Health care systems in transition: Estonia”, European Observatory on Health Systems and Policies, WHO European Centre for Health Policy, Belgium

European Observatory on Health Systems and Policies (2004), “Health care systems in transition: Netherlands”, European Observatory on Health Systems and Policies, WHO European Centre for Health Policy, Belgium

Felder, Stefan (2005), “Increased Life Expectancy and Health Care Expenditures: A New Calculation Using the Costs of Dying,” Paper presented at the 5th Congress of International Health Economics Association, Barcelona, July 13. Fortin, G (2004) “SHA-Based Health Accounts in Thirteen OECD Countries: Canada National Health Accounts 1999.” OECD Health Technical Papers No. 2. http://www.oecd.org/dataoecd/33/19/33733723.pdf

Government of Poland (2004), „Finansowanie ochrony zdrowia w Polsce - Zielona Ksiêga,” Ministry of Health, Warsaw

Graham, Dr. Martin (2000): “Health Insurance Regulation in the UK,” part of “Issues in Health Sector Regulation,” Oxford Policy Institute (OPI) Seminar Series, Trinity Term 2000, Week 3.

Hanvoravongchai, Piya (2002), “Medical Saving Accounts – Lessons Learned from International Experience,” World Health Organization Discussion Paper No. 52.

Hofmarcher, M, et al. (2005): Die Reform des Gesundheitswesens in der Slowakei: Pflichtprogramm oder Kür? in Health System Watch. Hauptverband der österreichischen Sozialversicherungsträger, http://www.ihs.ac.at/departments/fin/HealthEcon/watch/hsw05_1d.pdf

Ireland, Department of Health and Children (1999): “White Paper: Private Health Insurance” (Dublin, 1999).

Jemiai, N; Thomson, S; Mossialos, E. (2004): “An overview of cost sharing for health services in the European Union”, Euro Observer, Vol. 6, No. 3.

Katz S.J., McMahon L.F., Manning W.G. (1996): Comparing the Use of Diagnostic Tests in Canadian and US Hospitals, Medical Care, Vol. 34, No. 2, pp. 117-125. Kumpiene, Janine (2005): “Lithuania: Fiscal Reform Options in the Health Sector,” Background Paper prepared for The World Bank (2005): “Selected Issues in Fiscal Reform in Central Europe and the Baltic Countries 2005”, Washington DC.

Leu R, Schellhorn M. (2004): “The evolution of income-related inequalities in health care utilization in Switzerland over time.” Diskussionsschriften. Universitaet Bern. www.vwi.unibe.ch

Manning, W.G., J.P. Newhouse, N. Duan, E.B. Keeler, A. Leibowitz, M.S. Marquis (1987), “Health insurance and the demand for medical care: evidence from a randomized experiment”, The American Review, Volume 77, Issue 3, June, pp.251-277. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 133

Mossialos, E., Dixon A., Figueras J., Kutzin J., Eds. (2002) “Funding health care: options for Europe.” European Observatory on Health Care Systems Series. Buckingham, the UK: Open University Press. http://www.euro.who. int/document/obs/Fhcpolbrief4.pdf

Mossialos, E., Thomson, S. (2004): “Voluntary health insurance in the European Union. Copenhagen”. The European Observatory on Health Systems and Policies.

Mossialos, Elias; Le Grand, J (eds.). Health Care and Cost Containment in the European Union. Ashgate, 1999

Newhouse, J.P. (1977), “Medical care expenditure: a cross-national survey”, Journal of Human Resources, 12:115-125.

Newhouse, J.P. (1992), “Medical care Cost Increases: How Much Welfare Loss?” Journal of Economic Perspectives, Summer, 6, 3-21.

Newhouse, J.P., (1996). “Reimbursing health plans and health providers: efficiency in production versus selection.” J. Econ. Lit. 34, 1236–1263.

OBSAN (2004): „Entwicklung der Gesundheitskosten in der Grundversicherung von 1999-2003. obsan fact sheet“. Swiss Health Observatory. Nr. 1. www.obsan.ch

OECD (1997) “Economic Survey of Ireland 1997”. .

OECD (1999), “Challenges for Fiscal and Monetary Policy in the Euro Area.” Economic Outlook No. 64.

OECD (1999), “White Paper: Private Health Insurance”. Department of Health and Children, Ireland, Dublin.

OECD (2000), “Competition and Regulation Issues in the Pharmaceutical Industry.” DAFFE/CLP 29.

OECD (2000): “A System of Health Accounts” (SHA): www.oecd.org/EN/document/0,,EN-document-684-5-no-27- 24508-0,00.html

OECD (2004), “In Figures and Country HIT Reports.” OECD Health Data 2004. http://www.oecd.org/document/16/0,2340,en_2825_495642_2085200_1_1_1_1,00.html

OECD. (2003), “Health at a Glance.” OECD Indicators 2003. Paris, France. http://www.oecd.org/searchResult/ 0,2665,en_2649_201185_1_1_1_1_1,00.html

Orosz E, Morgan D. (2004) “SHA-Based National Health Accounts in Thirteen OECD Countries: A Comparative Analysis”, OECD Health Working Papers No. 16.

Oxley H, MacFarlan M. (1995): “How can OECD countries achieve their health-care goals while keeping rising expenditure under control?” OECD Economics Studies No. 24.

Panopoulou, Panagiota and Mukesh Chawla (2005): “Slovak Republic: Fiscal Reform Options in the Health Sector,” Background Paper prepared for The World Bank (2005): “Selected Issues in Fiscal Reform in Central Europe and the Baltic Countries 2005”, Washington DC.

Rocard, M. (2000): “Report on supplementary health insurance,” European Parliament, Committee on Employment and Social Affairs, 4 October 2000, (A5-0266/2000).

Rossel R, Gerber Y. (2004): “SHA-Based Health Accounts In Thirteen OECD Countries Country Studies: Switzerland National Health Accounts 2001.” OECD Health Technical Papers No. 12.

Rosenblatt, Law and Rosenbaum (1999), “Law and The American Health Care System,” Ch.2(A)(2), “Professional Autonomy, Equality, and Market Competition With Respect to Health Care Financing.” 134 Saltman R, Dubois H. (2005) “Individual incentive schemes in social health insurance systems”. Eurohealth Vol 10 No 2. pp. 21-25. Saltman R, Figueras J. (1998): Analyzing the evidence on European health care reforms. Health Affairs, 17(2):85-108.

Saltman, R et al (eds) (2002) “Regulating entrepreneurial behaviour in European health care systems”, http://www. euro.who.int/observatory

Saltman, R et al (eds)(2004) “Social health insurance systems in western Europe”, http://www.euro.who.int/ observatory

Statistik Austria. (2005). Jahrbuch der Gesundheitsstatistik. Bundesanstalt Statistik Oesterreich. http://www. statistik.at/pub/neuerscheinungen/2005/gesundheit2003.pdf

Tapay, Nicole (2001): „Issues Surrounding the Introduction of Private Health Insurance in Poland,” Background paper prepared for The World Bank (2001): „Evaluating the Impact of Health Care Reforms in Poland: The Challenges of Systemic Changes in a Transforing Environment,” Washington DC.

The World Bank (2000): “Chile: Health Insurance Issues, Old Age and Catastrophic Health Costs,” Human and Social Development Group, Argentina, Chile and Uruguay Country Management Unit, Latin America and the Caribbean Region, The World Bank. Thomson S, Mossialos E. (2004): “Private health insurance and access to health care in the European Union.” Euro Observer, Vol. 6, (1).

UK National Statistics (2003). Total Health Expenditures.

United Nations Secretariat (2003 and 2004), “World Population Prospects (2004 Revision) and World Urbanization Prospects (2003 Revision).” Population Division of the Department of Economic and Social Affairs.

Van Doorslaer E, et al. (1999): “The redistributive effect of health care finance in twelve OECD countries.” Journal of Health Economics; (18):291-313.

Van Doorslaer E, Jones, A. M. (2004): “Income-related inequality in health and health care in the European Union.” Health Econ. 13: 605–608.

Van Doorslaer E, Masseria C. (2004): “Income-Related Inequality in the Use of Medical Care in 21 OECD Countries”. OECD Health Working Papers No. 14.

Van Doorslaer, E., Koolman, X. and Jones, A.M. (2003), “Explaining Income-related Inequalities in HealthCare Utilisation in Europe: a decomposition approach”, Ecuity II Project Working Paper No. 5, October.

Wagstaff A, et al. (1999): “Equity in the finance of health care: some further international comparisons.” Journal of Health Economics; (18):263-290.

Wagstaff A, van Doorslaer, E: “Equity in the delivery of health care: methods and findings”. In: Van Doorslaer, Eddy, Adam Wagstaff, and Frans Rutten, eds. 1993. Equity in the Financing and Delivery of Health Care: An International Perspective. Oxford: Oxford University Press.

Weisbrod, B.A. (1991): „The Health Care Quadrilemma: An Essay on Technological Change, Insurance, Quality of Care, and Cost Containment,” Journal of Economic Literature 29, no. 2 (1991): 523–552.

Wild, C. (2004), “Health Technology Assessment in Austria.” International Journal of Technology Assessment in Health Care, 16(2): 303-3 World Health Report (2000): “Health Systems: Improving Performance,” World Health Organization, Geneva, Switzerland: 2000. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 135

V. MANAGING FISCAL RISKS IN PUBLIC–PRIVATE PARTNERSHIPS123 1. Introduction 301. There has been a renewed interest in the use of public–private partnerships (PPPs) to improve economic efficiency and to satisfy large investment requirements in the new EU member states (NMS) from Central Europe and the Baltic region.124 This interest has perhaps been stimulated by attempts to address the difficult challenge of reducing the existing disparities between old and new member states in the quality and availability of public services while maintaining fiscal stability or, in some countries, conducting fiscal adjustment. At the start of their transition to market economies, EU8 countries, like other countries from the region, inherited extensive infrastructure networks that were in serious disrepair and ill suited to their transforming economies (EBRD, 2004). Since the early 1990s, the European Investment Bank, the World Bank, and the European Bank for Reconstruction and Development have lent some €21 billion for infrastructure in the EU8 countries, plus Romania and Bulgaria. The total investment in infrastructure (public and private) in these countries during the same period has been estimated at €100 billion.125 The process of modernizing infrastructure was the fastest in telecommunications, where the Czech Republic, Estonia, and Poland have already nearly reached the EU level. By contrast, road infrastructure remains much inferior in the NMS, and efforts to modernize the infrastructure of the new EU members are yet to narrow the gap significantly to the incumbent members. However, narrowing the infrastructure gap is essential for improving economic competitiveness and the living conditions of the population in the NMS.

302. There is still a large gap between the public infrastructure (be it economic or social) investments needed to improve the quality and availability of public services and the limited public financial resources of the EU8 countries. Public investment in 2003 was close to or above 3 percent of GDP in all EU8 countries, except in Latvia (1.5 percent of GDP). Recognizing increasing investment needs, public investment as a share of GDP is projected to increase to 3.5 percent on average for the EU8 in 2006. While the average public investment level in the EU8 is slightly above the level in the EU-15 (around 2 percent of GDP), investment required to bring the EU8 infrastructure standards to their average for the union entails much larger costs. For example, studies indicate total investment requirements for the EU8 plus Bulgaria and Romania of around €500 billion during the next 15 years (Deutsche Bank, 2004). The long-term needs for transport infrastructure alone in these countries are estimated at nearly €100 billion (EC, 2001). Furthermore, the environmental investment needs of the EU8 are estimated at €47–69 billion. Poland has the largest environmental investment needs (€22–45 billion or 10-21 percent of 2003 GDP), followed by Hungary (€10 billion or 12 percent of 2003 GDP) and the Czech Republic (€9.4 billion or 10 percent of 2003 GDP) (CASE, 2005).

303. Further increases in public investment may be constrained by fiscal concerns, including the Maastricht fiscal criteria for Euro adoption. In 2003-04, four of the EU8 countries (the Czech Republic, Hungary, Poland, and Slovakia) had budget deficits higher than 3 percent of GDP, and deficits are projected to remain above this level through 2006. Meanwhile, the three Baltic countries have strong fiscal positions reflected by small deficits or overall surpluses. Public debt was lower than 60 percent of GDP in all EU8 countries, with the Visegrad countries around 40% of GDP and the Baltic countries in the low double digits. Nevertheless, long- term sustainability may be at risk in all EU8 countries (except perhaps Estonia) because of population ageing and inadequate monitoring and control of contingent liabilities (EC, 2005).

123 Prepared by Hana Polackova Brixi, Nina Budina, and Timothy Irwin. Valuable contributions were received from Malgorzata Markiewicz, Marian Moszoro, Jan Pesek, Joanna Siwinska, Krisztian Szentessy, Marketa Jonasova, Zhicheng Li Swift, and Sergei Shatalov. The paper benefited from valuable comments from Drahomira Vaskova, Allen Schick, and Sergei Shatalov (peer reviewers), Dirk Sommer, Thomas Laursen, Roger Grawe, Edgar Saravia, Christoph Rosenberg, Vikram Nehru, Brian Pinto, and Santiago Herrera. An earlier version of this chapter was discussed with EU8 government officials and international experts in a workshop held in Warsaw in June 2005. 124 The European Council supports PPPs and, in its European Growth Initiative of 2003, sets as one of its objectives promoting the use of PPPs to develop infrastructure that is judged necessary for reinforcing the potential growth of EU (EIB, 2003). The appetite for PPPs is also increased by the required co-financing for EU structural funds, as government co-financing capacity is limited and the use of private financing to co-finance infrastructure development is permitted (although there are only a few examples of such co- financing, notably the Irish road program). 125 These calculations were quoted in Deutsche Bank Research, Current issues (2004) and the original source is in Christian von Hirschhausen (2002). 136

304. Given the vast infrastructure needs in all of the EU8, the issue of how to create “fiscal space” for additional public infrastructure spending has gained prominence, as it has in Latin America. Heller (2005) defines fiscal space as “the availability of budgetary room that allows a government to provide resources for a desired purpose without any prejudice to the sustainability of the government’s financial position.” Creating fiscal space is critical when additional fiscal spending would likely boost medium-term growth and also pay for itself in terms of future fiscal revenue. There are different ways to create fiscal space: through revenue mobilization; optimization of the composition of budgetary expenditure (e.g. by cutting low priority public expenditure); measures to limit contingent liabilities and quasi-fiscal operations of other public entities; additional borrowing or money creation; and, finally, structural reforms to promote growth.126

305. It has long been debated whether to increase the flexibility of the Stability and Growth Pact (SGP) by excluding public infrastructure investment from the fiscal deficit, taking into account their asset-creating nature and different structural characteristics of the EU member states (Blanchard and Giavazzi, 2004). While public infrastructure spending will result in higher primary deficit and debt in the short run, it will create productive assets that will generate higher output – and hence taxes – while contributing user charges over the long run.127 The full impact of public infrastructure spending on long-run solvency can only be assessed by looking at the public sector’s net worth and the government’s inter-temporal budget constraint. Despite good arguments in favor of such an approach, there are important issues concerning the quality of public investment, possible short-term financing constraints, or conflicts with the need to reign in excess demand. Further, it may also distract attention from difficult fiscal reforms related to reprioritization of expenditure, including cuts in non-priority current expenditure, dealing with fiscal risks and contingent liabilities and eliminating off-budget expenditure, including on infrastructure. Freeing public investment from any fiscal constraint may also invite creative accounting, or classifying current spending as investment, and thereby excluding it from fiscal targets.

306. Fiscal space can also be created by rebalancing the existing portfolio of public investment projects, ensuring that the state would retain its involvement only in projects where this was productive. The government can stimulate private infrastructure investment, leading to improvements in the quantity and quality of public services, while lowering their cost. For example, increased competition in infrastructure (e.g. electricity), tariff increases toward cost recovery (e.g. water and transport), hard budget constraints on public utilities, and so on, can stimulate private investment and generate large fiscal savings.

307. PPPs operate at the boundary of the public and private sectors, being neither fully public nor fully private. PPPs are defined here as privately financed infrastructure projects in which a private firm either: (i) sells its services to the government; or (ii) sells its services to third parties with significant fiscal support in the form of guarantees. Despite these common elements of PPPs across sectors, there are differences in the type of arrangements that are typical in each sector (Box V.1).

308. In what follows, we focus on whether and when using PPPs can create fiscal space for additional infrastructure investments in the EU8. In doing so, the paper will examine the fiscal risks of PPPs and the role of fiscal institutions in this regard, including how these affect the use and design of PPPs and thus the potential for creating fiscal space while promoting investment in infrastructure. Section 2 distinguishes the illusory from the real fiscal effects of PPPs. Section 3 relates the extent to which PPPs reduce fiscal costs to the nature of fiscal institutions. Section 4 explains how fiscal institutions can be improved to encourage fiscal prudence in the use and design of PPPs. Section 5 concludes.

126 Monetary financing is not an option in the EMU. 127 For example, if public infrastructure investment leads to alleviating infrastructure bottlenecks, and it is a complement to private investment, it may well be that creating fiscal space for this investment will have a strong growth effect. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 137

Box V.1. How Does the Use of PPPs Differ by Sector? In roads, many types of PPP are observed. On roads with real tolls, the high costs of investment and uncertainty of traffic often lead governments to provide minimum-revenue or other guarantees. On untolled roads, governments sometimes link their payments to the volume of traffic (shadow tolls) and at other times agree to make availability payments conditional only upon the state of the road. In electricity generation and in wholesale water supply and wastewater treatment, municipal or state- owned utilities often enter into contracts that have some form of availability (or “capacity”) payments. Though not usually described as PPPs, the contracts are structurally similar to PPPs; the main difference is that the purchaser is a government subsidiary, not the government itself. In electricity and water distribution, PPPs, as defined, are less common. Privately financed providers abound, but governments are unlikely to give them guarantees - perhaps because these industries don’t require such large and lumpy investments as do roads and power generation. Governments do sometimes subsidize privately provided water services, for example, by making publicly owned assets available to a private water company at no charge, but though these arrangements are sometimes described as PPPs, they don’t raise the same fiscal issues as contracts with guarantees or long-term commitments to make availability payments. PPPs are also commonly used to procure “building facilities,” including schools, hospitals and prisons. In such cases, the private party usually finances, builds and maintains the building, but does not provide services, such as education and healthcare, to end users. The government typically makes availability payments to the private party, supplemented sometimes by usage-related fees. The government can usually control the latter (for example, by deciding how many prisoners are sent to the private prison), but not the former, which depend only on the buildings being available for use in proper condition.

2. The Fiscal Effects of PPPs 309. PPPs are generally used to generate fiscal savings while promoting investment (to create fiscal space). PPPs may generate fiscal savings, but, as we have defined them, they entail fiscal obligations that are often not captured in the fiscal accounts. Thus, their fiscal effects are often obscure. In what follows, we discuss the fiscal obligations created by PPPs and then consider whether, despite these obligations, PPPs might have a positive real effect on a government’s underlying fiscal position. Fiscal obligations from PPPs 310. A government that uses public financial resources to build infrastructure needs to make large upfront payments for construction. Unless the government increases taxes, its reported deficit and debt immediately increase. A government that uses a PPP, on the other hand, usually need not pay anything upfront. Construction bills are paid by the private partner, and the government’s reported deficit and debt do not immediately increase. Therefore, PPPs can be attractive to governments under pressure to reduce their deficits and debt.

311. However, fiscal comparisons between public financing and PPPs can be deceptive. First, the true, long-run fiscal effect of publicly financed public services is not always reflected in measures of a government’s fiscal position and performance. As advocates of greater fiscal space for public infrastructure investment have noted, public financing requires large upfront expenditure, but usually generates future revenues (Blanchard and Giavazzi op. cit.; Easterly and Serven, 2003; IMF 2004c and 2005a; Irwin and Serven, 2005). If tolls, tariffs, or other user fees cover the full costs of providing the service, a publicly financed investment pays for itself over the long run. Its effect on the government’s net worth – the present value of future revenues less expenditures – is neutral. Even when user fees cover only some of the costs, the net final fiscal effect of publicly financed investment is not as large as the immediate increase in the deficit and debt. The same holds true if the investments raise growth and general government revenues.

312. Second, the comparison is deceptive because even though PPPs do not immediately increase the government’s debt, they do create direct or contingent fiscal obligations.128 In one type of PPP, popularized in Britain under the name of the ‘private finance initiative’ and emulated by Australia, South Africa and

128 Privately financed public services do not necessarily impose such obligations; but PPPs, as defined in this paper, do. A toll road without government guarantees is an example of private financing that is not a PPP as defined here. In sectors such as roads where the service is provided to end users, governments tend to prefer PPPs to purely private financing when fully cost-covering user fees are either economically undesirable or politically infeasible. 138 many European countries, the government commits itself to purchase the output of the private partner. For example, the government may agree to make regular payments to a private provider of road, school, hospital, or prison facilities, conditional only upon the availability of the facilities in satisfactory condition. Such “availability” payments do not depend on the government’s subsequent demand for the services: so long as the private partner has properly constructed and maintained the facilities, the government must pay. While the government’s obligations to make these payments may not entail an accounting liability, they do entail an economic liability. When the government is the sole purchaser of the output of the private partner, the government usually needs to commit itself to payments that equal, in present values, the cost of the investment. Subject to some qualifications set out in the next section, the economic liability may be the same as the debt the government would have incurred had it used public finance. Lower public debt associated with these PPPs is in large part an artifact of governments’ standards for financial reporting.

313. In another type of PPP, final users purchase the services supplied by the private partner, but the government provides financial support in the form of guarantees. PPPs of this type were particularly popular in the nineteenth century, when governments promoted railways by guaranteeing investors’ returns (Ville, 1990). The guarantees helped get railways built, but their design sometimes encouraged cost overruns. Many railways turned out to be unprofitable in the early years, requiring governments to spend large sums for support. The guarantees often complicated budgeting and sometimes precipitated fiscal crises. Today, toll roads are the type of infrastructure most likely to benefit from guarantees. Chile, Colombia, Korea, and Spain, for example, have given both revenue and exchange-rate guarantees to toll roads. The former compensate investors when toll revenue falls below a certain threshold, the latter when local-currency depreciation increases the cost of servicing foreign-currency debt. These guarantees have helped attract private investment in valuable infrastructure, but have sometimes been expensive. Spanish exchange-rate guarantees, designed to help investors finance projects from foreign rather than domestic sources, ended up costing the Spanish government more than $342 billion pesetas between 1969 and 1990 (Gomez-Ibanez and Meyer, 1993). The guarantee on a single road in Korea, running from Seoul to a new airport at Incheon, may cost the government as much as $1.5 billion (about one-quarter of a percent of GDP) (Irwin, 2004).

314. Even when governments make no explicit financial commitments, they can feel obliged to bail out private infrastructure firms that become financially distressed. The Mexican government, for example, secured around $10 billion in private investment in toll roads in the early 1990s, but many of the concessions ran into trouble, in part because the government’s policy encouraged short concessions with high tolls, which discouraged traffic, and in part because of the effects of a financial crisis in 1994. To resolve the problems faced by the concessionaires’ creditors, the government took on $7.7 billion (or close to 2 percent of GDP in 1994) of private debt (Ruster, 1997; Gomez-Ibanez 1997). More recently, the British government took on several billion pounds of liabilities when Railtrack, the private company that owned the country’s rail infrastructure, became financially distressed (Ehrhart and Irwin, 2004). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 139

Box V.2. PPPs in the EU8

PPPs are not new to the EU8. Indeed, the very first railway in Europe with a government-guaranteed return ran from Warsaw to Vienna (Westwood, 1964). However, as in other countries, PPPs have returned to the limelight in the region in the last few years. As in the rest of the world, some involve the government signing long-term purchase contracts, others guarantees.

Hungary, for example, has entered into long-term purchase contracts to secure a prison and several university buildings, at an estimated total cost in present value terms of 483 million (CASE, 2005). It has also signed a long-term purchase contract under which it will make availability payments to the M5 motorway, which runs connects Budapest with Bratislava and Bucharest. The estimated present value of the payments is €914 million. The two amounts add up to about 2 percent of GDP.

Both Poland and Hungary have given guarantees to toll roads. The first privately financed toll road in Poland (at least in the modern era) is a segment of the A2, a motorway under construction running from the German border to Warsaw and beyond. The A2’s costs were estimated at around €745 million, of which €235 million came from equity and loans from the shareholders of the concession company and €235 million from senior lenders. The remaining €275 million came from a subordinated zero-coupon bond from the European Investment Bank, with an implied interest rate of just 6.5 percent, on the condition that the government guaranteed full repayment (Esty, 2004). Hungary originally provided a guarantee to the M5 motorway (which now benefits instead from availability payments). The government agreed that if toll revenue fell below a specified threshold for any reason, it would extend to the concessionaire a loan that would not have to be repaid before the concessionaire’s senior lenders were repaid in full. When the road opened, revenue was below the threshold and the government had to make payments, which it was later able to reduce by discouraging traffic on alternative public roads (EC, 2004).

In other cases, EU8 governments have sought private investments in infrastructure that have not required guarantees or long-term purchase commitments. At least one such project nonetheless had significant fiscal implications. The first privately financed toll road in Central and Eastern Europe was the M1/M15 motorway in Hungary, which was financed purely on the back of forecast toll revenues. Yet when the motorway was opened, traffic was less than forecast and grew more slowly. The tolls, among the highest in Europe, were unpopular, and a court ruled that they had to be reduced. As a result, the concessionaire could not finance its debt, and eventually the government took over the road and assumed part of the debt (52 billion Hungarian forint, or about a quarter of a percent of GDP).

Lastly, public enterprises in the region have entered into long-term purchase agreements that are like PPPs even if they are not usually described as such. These agreements create contingent obligations for the government. In many countries, governments guarantee that the private partner will receive the payments it is due from the public enterprise. Even when governments do not give formal guarantees, they may feel obliged to honor the commitments of public enterprises or important private enterprises, for example power-purchase agreements signed by a state-owned electricity utility. Long-term power-purchase contracts complicate the creation of competitive power-markets. Hungary, for example, cancelled its power-purchase agreements in 2002. It may be possible to ensure that customers pay for the costs of compensating the investors; if not, governments must pay themselves. 140

315. One useful way of thinking about fiscal liabilities associated with PPPs is to put them in the cells of a table that distinguishes between liabilities according to whether they are direct or contingent and whether they are explicit or implicit (Table V.1).

Table V.1. Selected Liabilities from PPPs and Other Privately Financed Infrastructure Projects.

Direct Contingent

Obligation in any event Obligation if a particular event occurs

Explicit Obligations to make availability payments State guarantees of revenues under PPPs for services Poland (A1 motorway) The Czech Republic (motorway D47, Government Hungary (M5 motorway) “Internet for schools”) liability created by State guarantees of debt of PPPs a law or contract Hungary (prison, educational facilities, sport facilities, and the M6 motorway) Poland (A2 motorway, energy project in Krakow, water and sewage treatment in Warsaw and Torun, Slovakia (planned PPPs for the D1 road and rail infrastructure in Lodz) motorway and for the National Theater Statutory guarantees on liabilities and other obligations in Bratislava) of various entities, including financial institutions, Tax expenditures, like exemptions possibly involved in PPPs (infrastructure development Poland (tax exemptions for state- funds, etc) owned companies) Czech Republic (Railway Transport Infrastructure Administration) Hungary (State Development Bank) Poland (Bank for Environment Protection providing guaranteed credit lines for small and medium enterprises and programs co-financed by EU) Estonia, Latvia, Lithuania, and Slovakia (investor protection guarantees) State guarantees on service purchase contracts Poland and Hungary (possible obligations arising from the past power-purchase agreements) Litigation Slovakia (legal claims by CSOB and the Slovak Gas Company)

Implicit Not typical in PPPs; would arise if Possible assumption of the obligations of state-owned utilities that have entered into unguaranteed power- governments were politically, but not A “political” purchase agreements with independent power legally, obliged to continue to provide producers obligation of subsides to a PPP. Hungary (past PPA contracts in the energy sector, government that PPAs in waste water and solid waste) reflects public and Poland (PPA contracts in the energy sector) interest-group Possible assumption of the debts of privately financed toll pressures roads, beyond the extent of formal guarantees Hungary (M1/M15 motorway construction concessions – partly implemented through the Road Construction Corporation of the State Development Bank) Poland (claims arising from expressway construction concessions) Possible demands by local governments for assistance in covering their own debt, guarantees, arrears, letters of comfort and other obligations possibly related to PPPs Poland (local government debt and guarantees related to regional development) Czech Republic (bail-outs related to hospital arrears) Claims by financial institutions involved in PPPs Slovenia (regional guarantee schemes)

Note: The table shows the obligation of the general central government (rather than the consolidated public sector). It is by no means comprehensive. Source: The authors, using the framework set out in Polackova (1998). Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 141 Possible overall fiscal savings from PPPs 316. Despite creating fiscal obligations, a PPP might improve the government’s net worth. In particular, the present value of the obligations might be less than the reduction in government expenditure or increase in revenues associated with the PPP (Table V.2)

Table V.2. Possible Real Fiscal Effects of PPPs on Net Worth.

Avoiding constraints imposed by fiscal If fiscal rules prevent public investment rules with more or less arbitrary but allow guarantees and long-term accounting conventions purchase contracts, and PPP investment leads to sufficiently large increases in user fees or tax revenue, PPPs can increase net worth.

Reducing whole-of-life costs If the private partner is better than the government at coordinating construction, operations and maintenance, PPPs can lower the whole-of-life costs of public services, increasing government net worth. Facilitating sustainable increases in user If it is easier to achieve or sustain cost- fees covering user fees when the service is privately financed, PPPs may reduce the need for taxpayer-funded subsidies, thereby increasing government net worth.

317. First, if the government is constrained by fiscal rules such as the Maastricht criteria, a PPP can allow the government to undertake more investment than it otherwise could. More or less arbitrary differences in accounting treatments may allow a PPP to proceed when an economically similar publicly financed investment would breach the constraint. In this case, the PPP’s true effect on government net worth depends on the circumstances. If the project generates no user fees for the government, it could increase the government’s net worth only if it stimulated economic growth so much that the present value of increased tax revenues surpassed the present value of the government’s payments. This seems unlikely, but not impossible (Brixi and Irwin, 2004). If the government used the asset(s) procured with the PPP (such as a road on which the government made availability payments to the private partner) to generate user fees, the PPP is more likely to have a positive effect on government net worth. Conversely, if the PPP allows the government to undertake investments that do not pay for themselves in future taxes and fees, it reduces government net worth.

318. Second, PPPs in which the government is the purchaser of the output under a long-term contract can increase the government’s net worth, if they reduce the total (“whole-of-life”) cost of providing the service. Private companies are usually better than governments at building infrastructure assets on time and on budget and at operating and maintaining them in a cost-effective and efficient manner. That creates an argument for allocating the risks of operations, maintenance and construction to a private firm, as is typically the case in a PPP. When a private firm bears these risks, project costs should be lower, and this should reduce the amount the government has to pay. Yet the advantages of contracting out construction, operations, and maintenance do not immediately create an argument for private financing of a project: governments can also contract out these functions in publicly financed investments (Figure V.1). Whether a PPP leads to lower costs generally depends on whether the government or the firm is better at coordinating construction with operations and maintenance. This matters because the quality of construction can affect the costs of operations and maintenance: high-cost construction may allow lower-cost operation and vice versa. Some argue that firms are likely to be better at coordinating the two functions and that delegating coordination to a private firm should therefore lead to cost savings; others are skeptical, arguing that fiscal sleight of hand is the real motivation for private financing in these cases (Quiggen, 2004). 142

Figure V.1. Comparing Public Finance to a PPP with a Long-Term Purchase Contract.

Public finance PPP with long-term purchase contract

Gov’tGov’t

OpCo Finance OpCo Operating Service contract contract Operating contract contract

Construction Finance Construction Firm contract providers contract

ConCo ConCo Finance providers

319. Third, PPPs may increase the government’s net worth if they make it easier for the government to introduce or raise user fees. User fees usually do not cover the full cost of water, power and transport infrastructure and, while governments often plan to raise the fees, they find it hard to carry out the plans in the face of political pressure (World Bank, 2004). Raising user fees is still difficult when the service is provided under a PPP, but when the private partner gets its revenue from the user fees, the private investors act as a counterweight to the political pressure for lower user fees. In particular, they will not invest unless the government can credibly commit itself to cost-covering tariffs or a combination of cost-covering tariffs and subsidies. Although governments can commit to cost-covering subsidies, not tariffs, they often choose cost-covering tariffs. Over time, the commitment may lead to higher user fees, lower subsidies and hence, higher government net worth.

320. PPPs are sometimes thought to have a higher cost of capital than publicly financed projects, which – with all other things being equal - would make them more expensive. The public party can usually borrow at a lower rate than the private partner, and it is sometimes thought that these extra interest payments should be weighed against any possible fiscal benefits. Borrowing costs do not give a complete picture of the cost of capital, however. The cost of capital invested in a project, whether the capital is public or private, is best thought of as its opportunity cost – the returns that could be earned by investing the capital in the best alternative project. Because projects differ in their risks, sophisticated assessments of the cost of capital adjust for differences in risk (using theories such as the capital-asset pricing model). The implicit approach taken here is to assume that the opportunity costs are the same for the public and private sectors, and that the risk-adjusted cost of capital is unaffected by the choice between a PPP and publicly financed investment (Brealey, Cooper, and Habib, 1997; Klein, 1997).

321. Net worth, the key measure of government solvency, is probably the most fundamental fiscal consideration. But indebtedness increases a government’s financial vulnerability, heightening the risk of adverse changes in net worth, and matters especially to governments that are already highly indebted. Even when PPPs do not increase a government’s net worth, they may reduce its indebtedness (counting the PPP obligations as debt). Consider a case in which the private partner sells its services to final users, and suppose that costs and tariffs are the same under public and private financing, so that the government’s net worth is unaffected by whether the project is undertaken publicly or privately. Even if the government gives some guarantees, its total liabilities, along with its assets, will be lower under a PPP than under public financing. This will not matter much to governments with low debt, but could be important to others. How to assess the overall fiscal impact of a PPP project? 322. An adequate assessment of whether a particular PPP project would lead to fiscal savings, compared to pure public financing, requires an adequate basis for comparison. First, it is necessary to estimate the true fiscal consequence of the publicly financed investment project, taking into account future revenues, as well as required future recurrent costs, besides the value of the initial investment. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 143

323. Second, such an assessment would require an estimation of the cost in present value terms of the government’s (or public sector’s) commitment in the PPP contract. This is relatively straightforward for long- term purchase contracts in which the government’s payments are fixed in advance. The simplest option is to assume that the private partner will achieve the performance targets it is set. All that needs to be done, then, is to determine the payments required by the contract and calculate their present value at the risk-free interest rate appropriate to the term and currency of the obligation. For example, if the contract requires availability payments of €10 million at the end of each of 20 years and the risk-free interest rate is 5 percent, the fiscal cost is approximately €125 million. If it were reasonable to suppose that the private partner would not fully meet the performance requirements, the government could estimate the expected penalties that could be deducted from the availability payments.

324. Estimating the cost in present value terms of guarantees and other forms of contingent support is more difficult, but often possible.129 Option-pricing formulas can be used to estimate the cost of simple guarantees.130 Monte Carlo simulation and other numerical techniques can be used to accommodate the features of more complex guarantees. Countries with extensive experience in the valuation of government guarantees include Canada, Colombia, Chile, the Netherlands, Sweden, Turkey, and the United States.131

3. How Fiscal Institutions Affect the Fiscal Cost of PPPs 325. Fiscal institutions affect the use, design, and, ultimately, fiscal cost of PPPs. Fiscal institutions are conventionally defined as institutional arrangements and management practices that relate to public resource allocation, resource use and financial management. The nature of these arrangements and practices directly affects government incentives, information and capacity. Among fiscal institutions, attention tends to focus on arrangements and practices of budget management, and more recently, government management of financial assets and liabilities. With respect to infrastructure, fiscal institutions affect the extent and type of fiscal support. They affect the extent to which governments use PPPs as opposed to traditional public financing, and accept risk exposures as opposed to providing cash subsidies under PPPs. Incentives 326. As EU8 countries seek to generate fiscal savings while promoting investment, their fiscal institutions affect their incentives toward the use and design of PPPs. Conventional fiscal institutions tend to promote incentives to: a. Favor PPPs even when public investment would deliver equal results at a lower cost in the long term; b. Accept risks (for example, offer explicit and implicit guarantees), rather than providing cash subsidies under PPPs;132 and c. In the design of PPPs, let the public sector accept risks that the private sector is more suited to bear.

129 See IMF (2005a) for a technical summary of the available methods for valuation of guarantees. 130 It is possible to illustrate how a relatively simple guarantee can be valued. Take a loan guarantee similar to that given to the A2 motorway in Poland. Assume that the project’s initial cost and value is €100 million, of which €30 million comes from equity and €70 million from a zero-coupon bond maturing in 20 years. The zero-coupon bond’s repayment has been guaranteed by the government, and therefore attracts an interest rate equal to the government’s borrowing rate of 5 percent (continuously compounded). At that interest rate, the required repayment is €190 million. Now suppose that without the guarantee the bond would carry an interest rate of 10 percent. The value of the guarantee is the difference between the present values of the promised future repayment of €190 million at the two interest rates: €70 million - €26 million = €44 million. Given the assumptions, most of the value of the debt financing comes from the government, even though it offers no cash. The guarantee can also be valued without making any assumptions about the interest rate that would apply in the absence of the guarantee by using the Black– Scholes formula (Black and Scholes 1973; Hull 2003). In this case, however, it is necessary to estimate – or make an assumption about – the volatility of the value of the asset financed by the guaranteed bond (where volatility is the annualized standard deviation of the return to the asset). The two measures are equal in this example at a volatility rate of 45 percent. 131 For Colombia, see Lewis and Mody (1997) and Echeverry and others (2002). 132 As the previous section illustrated, private investors often ask governments to provide financial incentives in the form of cash 144

327. This is largely because conventional fiscal institutions devote weaker scrutiny to non-cash fiscal support and long-term obligations, compared to cash-based support and immediate outlays. The difference mainly relates to the rules for fiscal reporting, accounting, and budgeting, for measurement of fiscal savings, and for government accountability with respect to fiscal performance. Measuring fiscal savings (and fiscal adjustment) in terms of the immediate impact on the deficit and debt that are not adjusted for government risk exposures has been known to encourage governments to provide contingent forms of fiscal support and assume risk and long-term obligations in exchange for short-term reductions of cash spending (Irwin et. al., 1997; Brixi and Schick, 2002).

328. Hence, with respect to infrastructure, measuring fiscal savings in terms of conventional government deficit and debt (that is, without considering the future fiscal cost of contingent liabilities, such as guarantees, and long-term obligations, such as take-or-pay contracts) creates an illusion of fiscal savings when investment and services are delivered without immediately raising the budget deficit and government debt. This illusion makes government risk exposure and long-term obligations under PPPs look “cheap” compared to public financing and cash subsidies. Since the fiscal cost of PPPs typically surfaces in the long term, the illusion holds even when countries develop a medium-term fiscal framework.

329. Although not explicitly captured by ESA95, some guarantees in EU8 countries are included in the fiscal accounts under the Maastricht criteria as countries have expanded their definition of general government.133 Specifically, these countries have brought the cost of quasi-fiscal operations implemented through extra- budgetary funds and off-budget agencies into their general government deficit and debt figures under the Maastricht fiscal framework. The commitment to bring extra-budgetary funds and off-budget agencies of a fiscal nature within the scope of general government has been, however, uneven across the EU8 countries. The Czech Republic and Slovakia may be considered the leaders in this regard.

330. Commendably, the Czech Republic since 1997, after publicly disclosing its large contingent liabilities arising from the so-called transformation institutions,134 off-budget funds and credit guarantees, has brought these into the Maastricht fiscal framework. Transformation institutions and off-budget funds have been either dismantled or scheduled for dismantling. Those that have remained in operation are now accounted for as part of general government, and the cost of their activities thus directly affects the deficit and debt figures.

331. The Czech Republic and Slovakia have also started to impute the future fiscal cost of government credit guarantees into their deficit and debt figures. These two countries have assessed most of their outstanding government guarantees as risky and started to reflect the maximum amount the government could be required to spend as part of their reported government deficit and debt. In 2003, the year in which the accounting policy changed, the increase in reported debt also showed up as an increase in the reported deficit. In the following years, any future increase in guaranteed debt will show up in the deficit in the year in which the guarantees are issued. Although this decision has negatively affected the countries’ reported deficit in the short term, it will technically prevent most past guarantees from raising future deficit figures and thus complicating future fiscal adjustment. This comes in contrast to the situation in other new EU member countries, notably Cyprus, Malta, and Poland, who have a large portfolio of outstanding government guarantees that have not been reflected in government debt and deficit figures.135

133 ESA95 and SNA record contingent liabilities when these liabilities are tradable and have market value. State guarantees are very rarely tradable (EC, 2004). 134 Transformation institutions had been created as off-budget agencies to borrow, issue guarantees and finance government support programs for banks, enterprises, and other entities. Some of the transformation agencies were covered by an explicit government guarantee, others were considered backed by the government implicitly. 135 In Poland, although a state guarantee regulation introduced in 1997 requires decent disclosure and prudential provisions, the regulation tends to be implemented in a less comprehensive and less binding manner. Hence, current government debt figures only partly reflect the future fiscal cost of outstanding guarantees. This is particularly so for guarantees issued by local governments, extra-budgetary funds, and off-budget agencies. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 145

332. Recent improvements in the coverage of general government deficit and debt reported by EU8 countries, however, fall short of fully capturing the future fiscal costs related to government risk-taking under PPPs. Within the EU and beyond, the EU8 countries are not unique in facing the problem of internalizing the fiscal cost of government long-term obligations and risk exposures. Specifically, take-or-pay contracts (including those signed by public utilities) and various types of guarantees, provided by the local as well as central levels of government under PPPs, are not easily captured by ESA95, nor other accounting standards. According to a 2004 ruling by Eurostat, for example, a PPP remains off a government’s balance sheet if the private partner bears construction and availability risks (that is, risks related to construction costs and delays and whether the private partner has constructed, operated, and maintained the asset to ensure that it can provide the required service) (Eurostat, 2004). However, in a PPP that meets the two criteria from the above referenced Eurostat ruling, but where the government is the sole purchaser of the output, the governments would accumulate debt-like liabilities without affecting its fiscal deficit and debt figures (at best, mentioning the liabilities only in a note to financial statements).

333. Similarly, accountability structures in EU8 countries as well as in a number of other EU countries, although improving, fall short of ensuring fiscal prudence in the use and design of PPPs. EU8 countries have been improving the accountability of policy makers with respect to medium-term fiscal performance as they have been establishing medium-term fiscal frameworks and compliance with EU fiscal surveillance. EU8 countries have also been strengthening their audit mechanisms (namely internal audit by the ministry of finance and external audit by the supreme audit institution), so as to promote accountability of policy makers for fiscal performance. Hungary’s National Audit Office, for instance, has taken initiatives to implement performance audits (rather than pure financial audits) and assess the government’s handling of fiscal risk. The existing accountability frameworks in EU8 countries (as well as most other EU countries) are, however, still incomplete with respect to government risk taking and risk management.

334. With respect to PPPs, policy makers do not seem accountable for the long-term fiscal risk arising from take-or-pay contracts and various types of guarantees offered by local and central governments. Similarly, there is no clear accountability for the adequacy of risk analysis that supports government decisions about fiscal support to infrastructure. Governments’ accountability for managing government risk exposures under PPPs is also limited. Information 335. Good information on and understanding of the long-term fiscal cost of PPPs is important for promoting risk awareness (that is, an open discussion and acknowledgement of risks and government risk exposures). EU8 countries, however, have only limited information on the risks involved in PPPs and limited understanding of the long-term fiscal cost of PPPs. Moreover, these countries make very little of such information publicly available. PPP contracts and their content are considered confidential. This makes it difficult for policy analysts to assess the long-term fiscal cost of PPPs – and for the public to exercise appropriate pressure on policy makers for fiscal prudence. Many countries, such as Australia, Argentina, Brazil, Britain, and Peru now publish at least some contracts or licenses issued to private firms for the supply of public services.136 Others such as Chile publish summaries of the contracts, including descriptions of the government’s fiscal obligations in their official gazettes. But around the world much more could be done.

336. Without good information and understanding, the risks and long-term fiscal cost of PPPs tend to be underestimated. This further contributes to making PPPs and government risk-taking under PPPs an attractive choice of fiscal support to infrastructure. Capacity 337. Building government capacity to evaluate and manage risks and long-term obligations that arise from PPPs takes time and effort. Yet building the capacity to mitigate risk at source, create risk-sharing arrangements, and manage any residual risk is essential for PPPs and will be discussed in the next section. Weaknesses in government capacity to evaluate and manage risk may surface in the form of inefficient risk allocation and excessive government risk exposure under PPPs. Promoting PPPs without having such capacity has proven costly in a number of countries, including Indonesia and Turkey.

136 Some contracts and licenses are available on the internet at . 146

338. EU8 countries have been building their fiscal risk management capacity. Most have centralized the authority for issuing government guarantees and for managing fiscal risk in their ministries of finance, have developed a central database of government guarantees and local government borrowing, and have built capacity to gather and analyze information on fiscal risk. For instance in the Czech Republic, the capacity to analyze fiscal risk (as well as disclosure) has been evolving (Bezdek et. al., 2003). The capacity to manage government debt and its risks has also improved and, for instance in Hungary, is comparable to international good practice in terms of fine-tuning the instruments for the issuing of public debt, and the systematic use of benchmarking in order to minimize risk and costs (Currie, Dethier, and Togo, 2003). The capacity to actively manage government risk exposures arising from contingent liabilities and control long-term obligations has been more limited.

339. Experience shows, however, that useful capacity can be built relatively quickly once the ministry of finance is committed to do so. Turkey (Box V.3) and Chile, for instance, have recently been able to employ relatively sophisticated risk valuation models to regularly assess government risk exposures under PPPs. Sweden has been notable for managing government contingent liability risk exposures as part of comprehensive government asset-liability portfolio management at the Swedish National Debt Management Office. In fact, based on its risk assessment, the Swedish National Debt Management Office is providing a menu of possible contract designs, risk exposure, and cost to the government.

Box V.3. Building Risk Management Capacity in Turkey During the 1980s-90s, the Turkish Treasury provided many credit guarantees to SOEs and municipalities, revenue guarantees in the electricity sector, debt guarantees in the water sector and guarantees on take- or pay contracts signed by the state-owned gas importer. Many of these guarantees were called during the economically difficult years of 1998-2002. In that period, the cost to the Treasury was about US$1 billion (0.3-0.8 percent of GDP) per year. As part of its 2001 reform program, the Turkish Government undertook policy and institutional reform in public liability management. While these reforms are yet to be completed, a number of good measures have been introduced: A stronger legal framework: The new laws on Public Finance and Debt Management (2002) and Public Finance Management and Control (2004) established a risk management framework for the entire portfolio of government (contingent as well as direct) obligations, and strengthened the role of the Treasury. Stricter guarantee issuance procedures: The Treasury established a guarantee ceiling and a Risk Fund (a cumulative contingent liability fund). Guarantee issuance is no longer automatic and involves the assessment of government risk exposure and risk-sharing possibilities. A credible collection policy: To reduce moral hazard under guarantees, the Treasury has tightened the use of write-offs and offsets on the amounts due from the delinquent, applied short maturities and interest on restructured amounts and seized revenues from delinquent debtors. Improved oversight and disclosure: Quarterly Debt Reports provide detailed information on government contingent liabilities and on the Turkish liability management system. An effort to build risk management capacity: Within the Treasury, a high-level Debt Management Committee and Deputy General Directorate for Risk Management have been established to implement an integrated public liability management framework. The effort to build risk management capacity includes utilization of innovative risk valuation tools. With the support of a foreign private firm, the Treasury has employed a stochastic simulation model for its US$6- billion portfolio of contingent liabilities. The model simulates both macroeconomic risk and sector- and contract- specific risk. Monte Carlo simulation generates the loss distribution. The model helps identify the optimal portfolio composition that would minimize losses in most scenarios. Furthermore, the model helps to determine the guarantee fee to be paid by the beneficiary (according to government risk exposure), future guarantee ceilings and annual budgetary transfers to the Risk Fund. The Treasury has been constantly updating the model and using it to decide on new guarantee issuance as input to budget formulation and to promote public understanding of fiscal risk. One of the outcomes has been a drastic reduction in guarantee issuance during 2003-2004, and further tightening of risk-sharing and collection rules. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 147 4. Directions for Further Reform 340. Fiscal institutions need to set the basis for transparency, accountability and other measures to support fiscal prudence in the use and design of PPPs. It will mainly be a task for the domestic fiscal institutions to promote fiscal prudence in the use and design of PPPs. While separate PPP legislation may not be required, public finance laws, regulations and practices need to be comprehensive in order to address all possible forms of government fiscal support, including government commitments under PPP contracts. International mechanisms, however, may help if they better address government contingent liabilities and long-term obligations. In EU8 countries, EU fiscal surveillance in particular could become more effective in promoting fiscal prudence beyond the budget.137 Promote awareness of risk and long-term fiscal cost 341. When government fiscal support is warranted, the decision on whether to undertake a PPP and on how to design government support in the context of this tends to be influenced by the extent of government risk awareness. In this regard, an open discussion of government long-term obligations and risk exposures enhances policy choices about fiscal support in infrastructure and improves government dealing with risk under PPPs. Similarly, at the level of local governments, introducing an open discussion and acknowledgement of risks, their sources, types, and possible long-term fiscal cost may deliver significant benefits in the soundness of local government involvement in PPPs. Most EU8 countries have been trying to collect, analyze, and discuss information about government risk exposures emerging from state credit guarantees. The Czech Republic, Estonia, and Latvia have been expanding the discussion to the whole portfolio of main government contingent liabilities and fiscal risk. This helps to build awareness about the risks and obligations under PPPs.

342. Credible valuation of risks and assessment of long-term obligations contribute to risk awareness. Scenario analysis, for instance, can be useful to make policy makers aware of the potential fiscal impact of the worst possible outcomes under PPPs. Given the tendency of the proponents of PPPs to underestimate their long-term fiscal cost, impartial valuation based on credible assumptions and methodologies, conducted possibly by the ministry of finance, provides a useful basis for acknowledging risks and future possible fiscal costs. In addition to risk valuation methodologies, there are practical analytical frameworks that promote risk awareness. The fiscal risk matrix presented in Table V.1, for instance, has been used by a number of countries, including China, the Czech Republic, India, South Africa, and the United States (Government Accountability Office) to promote government risk awareness.

343. Open discussion and acknowledgement of risks by policy makers is the key to promoting effective risk awareness and can be achieved even without relying on sophisticated risk valuation methodologies. In the Czech Republic, Hungary, Thailand, and South Africa, among others, ministries of finance have successfully promoted risk awareness by reaching out to the central bank, sector ministries, government and non- government agencies, academic institutions, and private sector agents to gather relevant information. Impose disclosure 344. Disclosure raises scrutiny, fiscal prudence, and the contestability of resources. Information that is disclosed invites scrutiny by people outside the government and by the government itself. When disclosure rules have broad coverage they enable the government to better monitor lower-level governments and public sector units and expand the share of government activities that is open to public scrutiny. Scrutiny is likely to generate pressure for greater fiscal prudence applied by governments at both the central and local levels.

345. Good-practice financial reporting standards require the disclosure of commitments, contingent liabilities, and some other sources of fiscal risk. Adopting such standards automatically creates a requirement to disclose information about most government risk exposures under PPPs. Also, since the government auditor must express an opinion on the accuracy of the disclosures, it automatically creates an enforcement mechanism.

137 This section draws on Brixi and Irwin (2004) and Brixi (2005). 148

346. However, promoting disclosure should not be held hostage to the improvement of these standards. Statements of government risk exposures, for instance, can complement any government financial statement or report. At the level of central government, Australia, Canada, the Czech Republic, the Netherlands, New Zealand, the United Kingdom, and the United States offer some good practices to consider. In these countries, the government publishes a list of its risk exposures, and discusses their sources, nature, sensitivities, and possible financial and allocative implications.

347. Statements of government risk exposures and obligations can be self-standing documents or part of regular budgetary documents. Government statements of risk exposures and obligations (including long- term purchase and subsidy contracts) can provide an estimate of the associated future fiscal cost. Chile, for example, discloses information about its guarantees beyond what is required by the financial reporting standards it follows (Government of Chile, 2003). The government intends to report according to modern accrual accounting standards in the future, but at present uses cash accounting. Nonetheless, in a report on public finances accompanying the budget, it discloses information on the costs of the revenue and exchange rate guarantees it has granted to toll roads. First, it presents estimates of the total cash flows it expects to pay or receive over the next twenty years as a result of the revenue and exchange rate guarantees (lumping together all the different concessions). Second, it provides estimates of the value of each of the revenue and exchange rate guarantees by concession.

348. With respect to PPPs, governments can disclose their fiscal risks before as well as after deals are completed. By disclosing draft PPP contracts, the government would allow for a wider discussion of the fiscal risks involved and for a possible adjustment in the contract before it is signed. Disclosure of the signed PPP contracts would promote public understanding of government fiscal risk exposures, which in turn would promote government accountability. In addition, governments could be required to disclose a description and analysis of their risk exposures under PPPs. Finally, to promote transparency and accountability, the fiscal cost of past PPPs should be disclosed in a timely manner, possibly as part of a broader discussion on realized fiscal risk in annual budgetary documents.

349. There are several prerequisites for both risk awareness and disclosure. These include having a database of government direct and contingent obligations to form a basis for analysis; adequate institutional capacity, including the capacity to gather and analyze relevant information, evaluate risk exposures, and conduct revenue and expenditure projections under PPPs (as well as for the government as a whole); and an adequate enforcement mechanism, including a supportive political and legal environment (for instance, with respect to the reporting by local governments, public sector units, and public utilities to ensure compliance. 350. Given the long-term nature of the fiscal costs of PPPs, it is appropriate to disclose long-term fiscal scenarios that would capture the possible fiscal effect of PPPs (along with other long-term obligations, such as those related to retirement and healthcare in ageing societies). Australia, New Zealand, the United Kingdom and the United States are countries that publish long-term fiscal projections covering periods of up to 75 years, and offer good practice to consider. Long-term fiscal forecasts could also include “fiscal” balance sheets, showing the present value of cash flows in various categories, including assets and liabilities not recognized according to prevailing accounting standards.

351. Long-term projections and fiscal balance sheets can assume current policies as the base case and assess whether those policies are sustainable. Projected cash flows from long-term purchase contracts in PPPs could easily be included in the projections. If the expected expenditures created by guarantees had been calculated, these amounts could also be included. Clearly, there’s room for reasonable disagreement about the assumptions underlying long-term fiscal models, so governments may want to present good and bad scenarios as well as a base case and to make the model available for public scrutiny. Enhance fiscal planning, accounting and budgeting 352. For countries with large portfolios of contingent liabilities and long-term obligations, fiscal planning, to be meaningful, needs to reflect the possible long-term fiscal implications of these obligations. Long-term fiscal projections discussed above would support government fiscal planning. Specifically, they would allow governments to reflect possible future spending pressures in its fiscal targets and, if needed, to adjust fiscal targets to maintain fiscal stability vis-à-vis long-term obligations and risks. Fiscal targets may be complemented by ceilings on government risk exposures, possibly in the simple form of ceilings on the face value of guarantees outstanding and of new guarantees issued in the coming budget year. Such Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 149 simple ceilings have been introduced, for instance, in Hungary, Latvia, and Poland. In these countries, the ceilings are approved by Parliament as part of the annual Budget Law. In Latvia, the annual Budget Law also provides ceilings for local government guarantees and borrowing. Furthermore, it specifies an allocation to cover the cost of past contingent liabilities that are considered likely to be realized in the given fiscal year. The ceilings, however, may not always be considered binding beyond a given year. In Poland, for instance, the guarantee ceilings fluctuated between 1.0% and 3.7% of GDP during 1999-2003. More developed approaches to support fiscal planning by addressing government risk exposures and long-term obligations include budgeting for risk and comprehensive asset and liability management frameworks (both discussed below).

353. Accounting reforms can address some but not all of the problems described here, in part because they can create requirements for routine, audited disclosure of some of the information described in the previous section. Accounting standards have the potential of enhancing the management of entities (government and public-sector entities as well as private-sector entities) and supporting the government’s aggregate fiscal discipline (which is our focus here). First, the adoption of “accrual” accounting, according to any reasonable standards, requires governments to report the assets they own and a measure of the fiscal deficit that reveals approximate changes in the government’s net worth, thus addressing the first reason why comparisons of PPPs and publicly financed investments are deceptive (see above). The New Zealand government, for example, produces a balance sheet showing assets as well as liabilities, of which state highways and state-owned power generation and transmission companies recently made up about 18 percent. It also reports an accrual measure of the budget deficit, which counts depreciation on existing infrastructure assets, but excludes cash spent on publicly financed investments.138 In terms of accounting and statistics, Slovakia belongs to the most advanced among EU 8 countries, as it now reports accrual-based government finance statistics to the IMF (IMF 2004b).

354. Second, the adoption of some accrual accounting standards requires governments to report as liabilities some of the obligations created by PPPs. In the United Kingdom, accounting standards setters published guidance designed to help reporting entities see beyond the “legal form” of PPPs and to capture their “economic substance.”139 As a result, the government reports assets and liabilities associated with some, though not all, of its PPPs containing long-term purchase contracts. Accounting standards can also require reporting entities to recognize the cost of guarantees in the year in which they are issued. The United States government, for example, has designed an accrual accounting standard that ensures that the net long-term present value of certain loan guarantees is captured in its accounts in the year in which the guarantees are issued.

355. However, even the best existing accrual accounting standards are not sufficient for solving all the problems. Accrual-based accounting standards do not force all costs and liabilities out in the open: they do not necessarily require the costs of guarantees to be included in calculations of budget deficits, and they do not necessarily require the liabilities created by long-term purchase agreements to be recognized alongside ordinary debts on the balance sheet. Fortunately, the leading international standards appear to be improving: International Financial Reporting Standards, International Public Sector Accounting Standards (which modify the International Financial Reporting Standards for use by governments), and Generally Accepted Accounting Principles in the United States, for example, all appear to be converging toward more accurate accounting for such instruments. According to each of these three sets of standards, many guarantees would be recognized at the present value of their future fiscal cost, while the value of most other guarantees would at least be disclosed. It will likely be some time, however, before the standards

138 For more information on the fiscal accounts for New Zealand see the Treasury web page: http://www.treasury.govt.nz/budget2005/ fiscalstrategy/annex1.asp 139 See “Private Finance Initiative and similar contracts,” a note appended to the UK Accounting Standards Board’s Financial Reporting Standard 5, “Reporting the substance of transactions”. 150 become fully satisfactory (Irwin, 2003).140 Moreover, although adopting accrual standards can help address the problems, the problems can also be addressed without adopting such standards.

356. In government budgeting, the objective is to expose any proposed PPP and its related fiscal obligations and risks to the same extent of scrutiny as cash spending. In this respect, contemporary approaches reflect two important principles for budgeting for fiscal risk (Schick, 2002): a. Apply a joint ceiling to the cost of budget and off-budget support for each sector in a fiscal year. Off- budget support is considered a form of subsidy and thus subject to the same scrutiny and limits as any spending program. The size of the hidden subsidy is calculated as the present value of the future expected fiscal cost; b. Have the budget immediately reflect the full likely fiscal cost of contingent support when such a scheme is approved. In countries subject to fiscal ceilings, such as the EU countries, this principle could imply that the net present value of the future fiscal cost associated with government guarantees issued in a given year would need to fit within the deficit ceiling, and the net present value of the future fiscal cost associated with government guarantees outstanding would count against the debt ceiling.141

357. Another, possibly complementary, option is to create a contingent-liability fund. Some governments have created a special fund (to accumulate financial assets from budgetary transfers and/or fees collected from guarantee beneficiaries) that is used to meet future calls on guarantees and other liabilities. When guarantees are issued, the sector ministry can be required to transfer to the fund an amount equal to the estimated value of the guarantee.

358. In Canada and the Netherlands, which follow the two principles above and also have a special fund, the finance ministry computes the expected annual payout on contingent liabilities undertaken on behalf of the programs of each line ministry. The finance ministry then deducts these expected payouts from the annual budgetary allocation for the ministry concerned. Similar arrangements, including mechanisms to provide reimbursement to the line ministry for such provisions if a payout on the contingent liability does not occur ex post, have also been tried in Colombia. In Sweden, based on its assessment of government risk exposure, the Swedish Debt Office requires the payment of a guarantee fee based on the guarantee’s likely future fiscal cost as a condition for the issuance of any guarantee. In case the guarantee beneficiary is not willing or able to pay the guarantee fee in full, the sector ministry is required to use its budget to cover the difference between the full guarantee fee and the amount paid by the beneficiary. This way, the likely future fiscal cost of a guarantee that is not covered upfront by the beneficiary is brought under the budgetary ceiling of the given sector.

359. These principles have several important implications for government fiscal performance. Budgeting for risk may or may not affect cash-based estimates of the government’s fiscal deficit. It depends on whether the effect on the deficit is recorded when money is transferred from the budget to a contingency fund (then no effect is recorded when a guarantee is called and paid for from the contingency fund) or only when actual cash payments are disbursed from the program account. But budgeting for risk makes policy makers more cash neutral – that is, neutral between alternative forms of providing government support in terms of deficit measurement, budget ceilings, or medium-term fiscal outlook. And, perhaps most importantly, budgeting for risk promotes risk awareness among policy makers.

360. In any case, improvements in the analysis, disclosure and management of the fiscal costs and risks of PPPs need not wait upon the adoption of better accounting and budgeting standards. Experience suggests that the benefits of greater scrutiny, cash neutrality, and risk awareness can be achieved gradually with or without a comprehensive transition of the accounting and budgeting systems to accrual basis. Countries that have successfully combined reporting of contingent liabilities (and wider disclosure of risk) with cash accounting include the Czech Republic and South Africa, and those budgeting for risk within a cash-based budgeting system include Canada, Colombia, the Netherlands, and the United States. Similarly, fiscal risk can

140 In addition, some international statistics rules can be helpful with regard to government risk exposure. For instance, according to the IMF (2004a), non-financial public enterprises that are not commercially run should be included in fiscal statistics. 141 Alternatively, using conventional accounting approaches, one could think in terms of reducing the deficit ceiling by an amount equal to the net present value of the future fiscal cost associated with government guarantees issued in a given fiscal year. This principle would also imply that the debt ceiling would be reduced by the amount of the net present value of the future fiscal cost associated with government contingent liabilities outstanding. This would hold not only for guarantees but also for other forms of contingent fiscal support. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 151 be brought into the government’s medium-term fiscal framework. Setting the government budget and risk exposures in the context of a publicly announced medium-term budgetary framework makes any subsequent departures from the original risk analysis apparent. This has already strengthened the accountability of policy makers and the quality of fiscal policy in countries such as Australia (including New South Wales at the local level), Canada (including British Columbia and Ontario at the local level), Hungary, and South Africa. Advance fiscal risk management to reduce government risk exposure 361. The experience of governments trying to actively manage their risk exposures shows that fiscal risk management is very demanding. Governments find that to manage their risk exposures they need the capacity to gather relevant information, a comprehensive database of all major risk exposures, the opportunity for open discussion; the ability to understand fiscal risk (which may be assisted by useful analytical frameworks), and incentives to act correctly – incentives that are supported by adequate disclosure, accounting and budgeting rules, as discussed above.

362. Proper incentives in dealing with local government risk are supported by appropriate accountability structures. Policy makers need to be accountable for the adequacy of their risk analysis, assumptions, and decisions that involve fiscal risks and for managing the overall risk exposure of the government. Therefore, the role of supreme audit institutions (and the local audit bureaus) is to audit all aspects of government risk analysis and risk management. The General Accounting Office in the United States sets an excellent example, and its published reports provide for some very instructive reading.

363. Practice has shown the importance of three additional features of risk management: a clear risk management strategy (to specify to what extent the government is prepared to take on fiscal risk), centralized risk-taking authority (possibly in the budget office of the ministry of finance), and risk analysis and monitoring that are separate from risk taking (authorizing a separate entity as being responsible for analyzing and monitoring risk).142 The division of responsibilities and functions in risk management and the underlying reporting arrangements need to be very clear to provide a basis for adequate accountability structures.

364. In a decentralized environment, it is appropriate to control the risk-taking by local governments. Just as local government143 borrowing is overseen by the central government, the involvement of local governments in PPPs should also be subject to central government control. Moreover, given the technical demands related to PPP contracts, it may be appropriate to have the central government (for instance, a central PPP unit if established) involved in the design and implementation of PPP-related activities of the local governments.

365. For fiscal risk analysis and monitoring to be effective it needs to be comprehensive. Specifically, it needs to cover the whole range of channels through which governments at the central as well as local level generate fiscal risks, including letters of comfort, credit and guarantee funds, development corporations, local government-controlled enterprises, and so on.

366. Among government agencies and departments, the debt management office is often most able to analyze, monitor, and manage government risk exposures. Specifically, the debt management office is often best equipped to gather and analyze information about government contingent liabilities, evaluate government risk exposure and future possible implications of contingent liabilities on government debt, reflect on the analysis of contingent liabilities in the borrowing and debt management strategy, and advise the government on the future possible fiscal cost of newly proposed programs and on how to structure these programs to reduce government risk exposure. Debt management agencies are likely also to be in a good position to understand off–balance-sheet debt in the form of long-term purchase agreements. Entrusting the debt management offices with the analysis and monitoring of risk would be beneficial also, as it would separate these functions from risk taking, which is important in order to reduce moral hazard in risk analysis.

367. Debt management offices have been placed in charge of risk analysis and management, often in the context of the entire government balance sheet, in several countries, most notably in Sweden, Ireland, and Belgium. Among EU8 countries, Hungary’s debt management office has developed good risk management

142 This role of independent risk analysis and risk monitoring can be undertaken by the debt management office (e.g. Sweden) or by the supreme audit institution (e.g. US Government Accountability Office). 143 See Chapter II for a discussion of debt controls on local governments. 152 capacity that could be extended to the management of government long-term obligations and risk exposures. Depending on the specific institutional arrangements and capacity in a given country, another dedicated office within the ministry of finance or separate agency may be entrusted the responsibility for analyzing, monitoring, and managing government long-term obligations and risk exposures under PPPs. For instance, where appropriate, these functions may be conducted by a dedicated PPP office. Some oversight by the debt management office, however, would be beneficial for better and more comprehensive management of government assets and liabilities.

368. Ideally, for risk management purposes, government balance sheets would be extended to include future revenues, contingent liabilities and long-term obligations, as well as assets and liabilities. Such an extended assets and liabilities management framework (see Brixi and Mody, 2002 for further details) would provide a useful context for the government’s debt strategy (for instance, the selection of debt instruments and debt portfolio decisions to off-set government risk exposures arising from contingent liabilities) as well as fiscal planning (discussed above).

369. Related to the use and design of PPPs, reducing government risk exposure entails three complementary tasks: involving the private sector, transferring the risk to parties better able to bear the risk in the design of PPPs, and managing any residual risk that cannot be mitigated or transferred. Involving the private sector mainly implies mitigating the risk at the source and developing the financial markets. Ultimately, risk mitigation with private sector involvement is the most desirable long-run strategy. It not only reduces the government’s exposure to fiscal risks but also reduces the vulnerability of the economy to shocks.

370. In infrastructure, policy makers may need to ask how to reduce the dependence of private providers and investors on government guarantees and other kinds of support. Countrywide legal, regulatory, and administrative changes and proper debt management strategies can facilitate the establishment of an efficient domestic bond market, which in turn will smooth the progress of private infrastructure, as well as improve the government’s capacity to absorb risk. Private investors and providers in infrastructure may also be more willing to forego government guarantees when the investment climate in the country improves. Regulatory changes can encourage large international insurers to access the local market and pool risks, such as weather risk, that are uninsurable in a small economy. New financial instruments, such as asset-backed securities or catastrophe bonds, may help domestic financial institutions manage risk better, thus reducing their demand for government guarantees. Strategies to promote risk mitigation and financial market development, however, often hinge on fundamental sector reforms, such as reforms in energy pricing, production, and distribution systems.

371. Risk transfer mainly implies creating risk-sharing arrangements in PPP contracts. Creating a good risk- sharing mechanism in the PPP design requires clear policy objectives and understanding of all the underlying risks in a project. For both central and local governments, so far, the primary method of transferring risk has been through risk-sharing provisions in guarantee and insurance contracts. In private infrastructure, recent practice has suggested that carving out commercial risk from the coverage of government guarantees reduces moral hazard under the project, and limits government risk exposure.

372. Residual risk can sometimes be hedged. The private sector and, for some risks, international financial institutions offer useful risk mitigation tools. Governments and public sector entities, for instance, sometimes use currency swaps and commodity futures to hedge their foreign exchange and commodity price risks. They have also purchased reinsurance for disaster risk and weather risk from large international re-insurers. Increasing integration and liberalization in the market for insurance has made it easier to pool risk across countries and, increasingly, to insure risks that were until recently considered uninsurable. Governments might use some of these tools to hedge their exposure to risks in infrastructure projects. For the largest projects exposed to catastrophic risk, governments might also be able to issue catastrophe bonds, which offer lower yields when a catastrophe occurs. Given the still nascent stages of the international catastrophe bond market and weaknesses in the derivatives market, however, it is likely that governments will be able to reduce their risk exposure more effectively by first focusing on policies to mitigate the risk at source and develop the domestic financial markets. Similarly, a government’s own limitations in risk management capacity make hedging approaches less plausible.

373. Risks that cannot be avoided or hedged must be absorbed, requiring the government to manage its financial assets so that it has cash when it needs it. If the government cannot avoid bearing a risk and cannot Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 153 hedge this, it has no choice but to absorb the risk: that is, to bear any losses. It must therefore have sufficient cash on hand to enable it to make payments when they fall due. There are three basic ways of doing this: a) Put cash in a contingent liability fund (as discussed above) and hope the funds are sufficient to meet future payments; b) Use the cash to reduce debt and hope it can use tax revenues or additional borrowing if and when it needs to make payments; and c) Enter into a standby credit agreement with a bank that will allow it to borrow if it needs to make payments.

374. Each option has advantages and disadvantages. Having cash in a fund may give the government stronger assurance that cash will be available when needed, but it may also have a cost, because the cash could otherwise be used to repay higher yielding debt or invest in public services. The existence of cash in a fund may also tempt the government to use the money for other purposes. One option is to contract out management to a reputable foreign entity. The contract could specify permissible reasons for withdrawing cash from the fund without penalty and make other claims subject to a penalty and to prior public disclosure. Using the cash to repay debt may be cheaper, but leaves open the question of whether the government will be able to borrow or raise taxes when liabilities fall due – possibly at a time of crisis. A standby credit agreement, if available, solves the last problem, but at a cost that may be high.

375. The options are not mutually exclusive. A government can, and may indeed have to, use more than one option. The contingent liability fund, for example, cannot cover all contingencies. Even if the fund has the limited purpose of meeting calls on guarantees, it will be large enough to meet the worst possible losses only if the contributions are set according to the face value of the guarantees, rather than their expected costs. If contributions are smaller, the fund may need to be combined with reliance on taxing/borrowing or on a standby credit agreement. International mechanisms to promote fiscal prudence vis-á-vis PPPs 376. While countries’ own efforts to promote fiscal prudence in the use and design of PPPs is fundamental, international institutions could potentially play a larger role than currently is the case. Ideally, general EU and other multilateral fiscal surveillance would continue to broaden its coverage of fiscal risks in member countries, including those related to PPPs at both the central and local government levels.

377. More could also be done to reward countries for disclosure in international fiscal transparency assessments (that is, “upgrade” for transparency rather than “downgrade” for risks revealed). Experience suggests that countries may be punished, rather than rewarded, when they reveal contingent liabilities. For instance, in 1997, when the Minister of Finance of the Czech Republic volunteered detailed information about then-unknown contingent liabilities and launched an effort to bring contingent liabilities under control, international institutions and sovereign credit rating agencies reacted negatively. Rather than commending the formidable steps toward fiscal transparency and discipline, analysts rang warning bells. Being unusual in voluntarily revealing government contingent liabilities, it was portrayed as if it was unusual in having contingent liabilities. This implies that transparency may appear costly in the short term.

378. International organizations can also continue to enhance international standards for disclosure, accounting and budgeting. In the European Union, Eurostat could continue improving the guidelines on how to account for PPPs. It could require government explicit contingent liabilities, such as guarantees, to be disclosed and accounted for based on their net present value. The European Commission could require countries to prepare and disclose long-term fiscal projections reflecting the costs of guarantees and long- term purchase obligations.

379. Further enhancing auditing standards and capacity of supreme audit institutions with respect to PPPs would also help. In this regard, INTOSAI, benefiting from the experience of the US General Accounting Office, has been successful in advancing auditing for government risk exposures and risk management.

380. Finally, advances in international mechanisms could be complemented by continued efforts to build government capacity to design PPPs and deal with the related fiscal risks. Some of the support for capacity building could come from international institutions, such as the European Commission or the World Bank; some of it could be purchased from the private sector. 154

Table V.3. Improving Fiscal Institutions for PPPs at a Glance.

Goal Options Time horizon Risk Collect and centralize information on PPP contracts S awareness Discuss risks and long-term fiscal cost of PPPs as part of government decision- S making M Promote measurement of fiscal risks and the valuation of fiscal obligations Disclosure Disclose past fiscal costs of PPPs S Disclose outstanding PPP contracts S Disclose government analysis of risks and long-term obligations under PPP S-M contracts S-M Disclose draft PPP contracts and government analysis of related risks and M obligations M Disclose long-term fiscal scenarios reflecting the future possible fiscal effect of PPPs Enhance the whole system of financial reporting standards to require disclosure of commitments, contingent liabilities and other sources of fiscal risk in the public sector ACCOUNTING, Reflect the future possible fiscal effect of PPPs in fiscal planning S BUDGETING, Set overall limits on government risk exposure—either as simple ceilings on the S AND FISCAL face value of government guarantees, or as part of joint ceilings on cash PLANNING outlays and the discounted cost of guarantees Reflect the net present value of expected fiscal cost of PPPs in the S-M measurement of government debt when the government obligation originates (and changes in the net present value in the measure of the S-M government deficit) S-M Consider charging guarantee fees in the amount of the net present value of expected fiscal cost of PPPs M Consider reducing deficit/debt ceilings by risk-adjusted values of contingent liabilities issued/outstanding and/or establishing a contingent liability fund ENHANCE THE WHOLE SYSTEM OF ACCOUNTING AND BUDGETING STANDARDS TO ADDRESS GOVERNMENT COMMITMENTS, CONTINGENT LIABILITIES, AND OTHER SOURCES OF FISCAL RISK ON A SYSTEMATIC BASIS Risk Monitor government risk exposures and obligations S management Centralize government risk-taking authority (except where the accountability of S decentralized organizations is strong enough) Audit government risk analysis and risk management S Build capacity to evaluate and manage risk S-M Develop extended assets and liabilities management framework M

Note: S = short term, M = medium term. 5. Conclusions 381. It is important for EU8 countries to raise the likelihood that PPPs will be used and designed well, and will deliver fiscal savings as well as promote investment in infrastructure. In this respect, strengthening of fiscal institutions to enhance transparency and accountability in the use and design of PPPs is paramount. Specifically, countries in the region would benefit from promoting awareness of risks and long-term fiscal costs, imposing disclosure, enhancing fiscal planning, accounting and budgeting, and advancing fiscal risk management to reduce government risk exposure.

382. Institution building takes time, but international experience suggests that major advances in dealing with government risks and obligations of PPPs can be achieved when the government is committed to fiscal prudence. Such commitment, in turn, grows with better fiscal institutions.

383. International institutions could also do more to stimulate the commitment to change and support efforts to enhance fiscal institutions in the EU8 countries. In particular, EU fiscal surveillance could be expanded to cover contingent liabilities, bring government contingent liabilities within the criteria by which a country’s fiscal performance is assessed and develop mechanisms to reward transparency and punish opacity and excessive risk taking. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 155 References Allan, W. and T. Parry, “Fiscal Transparency in EU Accession Countries: Progress and Future Challenges”, IMF Working Paper WP/03/63. Bezdek, Vladimir, Kamil Dybczak and Ales Krejdl (2003), “Czech Fiscal Policy: Introductory Analysis.”, Czech National Bank WP, November. Black, Fischer, and Myron Scholes (1973), “The Pricing of Options and Corporate Liabilities.” Journal of Political Economy, volume 81, pages 637–654. Blanchard, Olivier and Francesco Giavazzi (2004), “Improving the SGP through a proper accounting of public investment”, CEPR Discussion Paper Series No. 4220. Brealey, Richard A., I. A. Cooper, and M. A. Habib (1997), “Investment Appraisal in the Public Sector.” Oxford Review of Economic Policy, volume 13, issue 4, pages 12–28. Brixi, Hana Polackova (2005), “Contingent Liabilities in New Member States.” in “Fiscal Surveillance in EMU: New Issues and Challenges”, ed. European Commission. Edward Elgar Publishing (forthcoming). Brixi, Hana Polackova and Timothy Irwin (2004), “Fiscal Support for Infrastructure: Toward a More Effective and Transparent Approach.” World Bank (processed). Brixi, Hana Polackova and Ashoka Mody (2002), “Dealing with Government Fiscal Risk: An Overview.” in “Government at Risk: Contingent Liabilities and Fiscal Risk”, ed. by Brixi and Schick. Washington DC: World Bank and Oxford University Press. Brixi, Hana Polackova and Allen Schick (editors) (2002), “Government at Risk: Contingent Liabilities and Fiscal risk.” World Bank and Oxford University Press. Burns and Yoo (2002), “Public Expenditure Management in Poland,” OECD, Economics Department WP No. 346.

CASE (2005), “Background Paper on Fiscal Risks from PPPs,” Mimeo, Warsaw, Poland. Currie, Elizabeth, Jean-Jacques Dethier, and Eriko Togo (2003) “Institutional Arrangements for Public Debt Management.” WPS3021. World Bank. Deutsche Bank (2004), “Infrastructure as a Basis for Sustainable Regional Development”, Deutsche Bank Research, June.

Easterly, William, and Luis Servén (2003), “The Limits of Stabilization: Infrastructure, Public Deficits, and Growth in Latin America.” Stanford University Press and the World Bank. EBRD (2004), “Transition Report” , London. Echeverry, Juan Carlos, Verónica Navas, Juan Camilo Gutierrez, and Jorge Enrique Cardona (2002), “Dealing with Contingent Liabilities in Colombia.”, Chapter 12 in Brixi and Schick (2002). Ehrhardt, David, and Timothy Irwin (2004), “Avoiding Customer and Taxpayer Bailouts in Private Infrastructure Projects: Policy toward Leverage, Risk allocation, and Bankruptcy.” WPS 3274, April. EIB (2003), “The Growth Initiative: Further Report to ECOFIN”, July. EC (2004a), “The EU Economy: 2004 Review,” DG Economic and Financial Affairs, October. EC (2005), “Assessments of the December 2004 Updates of Convergence Programs of EU8 countries.” EC (2001), “White Paper – European Transport Policy for 2010: Time to Decide.” EC (2004b), “Resource Book on PPP Case Studies.” DG Regional Policy. Eurostat (2004), News Release No. 18, February 11, “Treatment of Public Private Partnerships.” Available via the Internet: http://europa.eu.int/comm/eurostat/Public/datashop/print-product/EN?catalogue=Eurostat&product=2- 11022004-EN-AP-EN&mode=download Gómez-Ibáñez, José A. (1997), “Mexico’s Private Toll Road Program.” Kennedy School of Government, Harvard University. 156 Gómez-Ibáñez, José A. and John R. Meyer (1993), “Going Private: The International Experience with Transport Privatization.” The Brookings Institution. Government of Chile (2003), “Report on Public Finances: Government Budget Bill for 2004.” Heller, Peter S. (2005), “Understanding Fiscal Space.” PDP/05/4. IMF. Hull, John C. (2003), “Options, Futures, and other Derivatives.” IMF (2004a), “Public-Private Partnerships”, Washington DC. IMF (2004b), “Government Finance Statistics Yearbook.” IMF (2004c), “Public Investment and Fiscal Policy.”, March. IMF (2005), “Government Guarantees and Fiscal Risk.”, March. Irwin, Timothy (2003), “Accounting for Public-Private Partnerships: How Should Governments Report Guarantees and Long-Term Purchase Contracts?” Irwin, Timothy (2004), “Measuring and Valuing the Risks Created by Revenue and Exchange-Rate Guarantees in Korea.” Chapter 3 of Developing Best Practice for Korea’s PPI Market: With a Focus on PSC. Private Infrastructure Investment Center of Korea. Irwin, Timothy and Luis Servén (2005), “Fiscal Space”.LAC Regional Study Summary Paper. Irwin, Timothy, Michael Klein, Guillermo Perry, and Mateen Thobani (editors) (1997), “Dealing with Public Risk in Private Infrastructure.” World Bank. Klein, Michael (1997), “The Risk Premium for Evaluating Public Projects.” Oxford Review of Economic Policy, volume 13, issue 4, pages 29–42. Lewis, Christopher M., and Ashoka Mody (1997), “The Management of Contingent Liabilities: A Risk Management Framework for National Governments.” Chapter 6 of Irwin and others 1997. OECD (1999), “White Paper: Private Health Insurance”, Department of Health and Children, Dublin, Ireland. PriceWaterhouseCoopers (2004), “Developing Public Private Partnerships in New Europe.” Quiggin, John (2004), “Risk, PPPs, and the Public Sector Comparator.” Australian Accounting Review. July, volume 14, number 2, pages 51–61. Ruster, Jeff (1997), “A Retrospective on the Mexican Toll Road Program (1989–94).” Public Policy for the Private Sector. Note number 125. Serven, Luis (2005), “Fiscal Rules, Public Investment and Growth,” Mimeo; World Bank, Washington DC. Schick, Allen (2002), “Budgeting for Risk.” In “Government at Risk: Contingent Liabilities and Fiscal Risk” ed. by Brixi and Schick. World Bank and Oxford University Press. Ville, Simon P. (1990), “Transport and the development of the European economy, 1750–1918.” New York: St. Martin’s Pres. Von Hirschhausen, Christian (2002), “Infrastructure Policies in the Central and east European Accession Countries – A Survey in the Wake of EU-Enlagrment”. Westwood, J. N. (1964), “A History of Russian Railways.” London: George Allen and Unwin Ltd. World Bank (2004), “World Development Report: A Better Investment Climate for Everyone.” New York: Oxford University Press. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 157 SUMMARY OF CONFERENCE DISCUSSIONS144 Opening Remarks In his opening remarks, Tomas Prouza noted that the conference was an opportunity for a quick retrospective and review of the main fiscal challenges for the EU8 countries. He pointed to the World Bank’s close involvement in both the past transformation and in the more recent diagnostics and assessment of fiscal reforms needs in the Czech Republic, highlighting in particular the Country Economic Memorandum (“Towards EU Accession”) and Public Expenditure Review (“Enhancing the Prospects for Growth with Fiscal Stability”) as great sources of inspiration for fiscal measures. Mr. Prouza noted that these studies had contributed importantly to the local understanding of the needs and challenges of transforming economies in the region and were respected for their independent opinion in the nationwide debate on public finance reform.

Mr. Prouza noted that all EU8 countries had come a long way and succeeded in their goal of becoming EU members, with half of them having gone even further and already entered the ERMII. In contrast with the past, when in the EU8 public finance measures were focused on short-term savings and fast consolidation, today the focus was more on structural reforms and the quality of public finance – similar to the developed countries. In order to get ready for future challenges such as ageing, globalization and increased competition, all EU countries, including the New Member States (NMS), had to restructure public spending programs, increase efficiency and improve the quality of public finance and thus create fiscal space for maneuvering to be able to cope with these problems.

The renewed Lisbon strategy (as reflected in the “Integrated Guidelines” and endorsed by the 2005 Spring Council) provided for an increased focus on growth and jobs. This made it all the more important to prepare for demographic trends and the fast integration of new countries into the international economic system, which would put strain on public finances and labor supply. The subsequent dialogue between the Commission and member states on the new Lisbon governance architecture had identified in the macroeconomic part public finance sustainability as the most important challenge for 13 member states, including the Czech Republic, Hungary, Poland, and Slovakia. The emphasis should be on structural reforms to ensure economic dynamism and on the contribution that macroeconomic policies and healthy public finances could make in this respect. Low employment, education and lifelong learning were identified as main challenges in all EU8 countries and infrastructure was also critical in several countries. Mr. Prouza noted that fiscal discipline had not been relaxed in the revised Stability and Growth Pact (SGP); but rather that the instruments for EU economic governance and policy coordination had been better interlinked in order to enhance the contribution of fiscal policy to economic growth and support progress towards realising the Lisbon strategy.

In sum, the current and future challenges were to increase and stimulate growth through infrastructure, business friendly environment, enhanced competition and increased productivity; face population aging by keeping pensions and healthcare costs under control and extending the working life; and increasing the effectiveness of the public sector through expenditure programming, output oriented budgeting, and project management and financing in order to provide better services for taxpayers and reduce costs.

Mr. Prouza noted that the Czech Republic had been reforming its public finances continuously since the collapse of the communist regime, with efforts to make public finances sustainable further intensified on the eve of EU accession. In 2003, the Czech Republic had launched a comprehensive public finance reform with the objectives of (i) consolidating public finances; (ii) promoting economic growth; and (iii) preparing for EU entry and euro adoption. The reform rested on a mix of revenue and expenditure measures and institutional changes.

On the expenditure side, the primary objective had been to accomplish budgetary savings, but many measures were also aimed at improving economic incentives, stimulating employment, and strengthening the supply side of the economy. Measures included parametric changes in the pension system (further rise in the statutory retirement age; abolition of actuarially unfair early retirement schemes; contained indexation of pension benefits; and abolition of the restriction on drawing pension benefits simultaneously with earning income), less generous social benefits (cuts in sickness benefits and state social support benefits; and removal of the upper limit on the earnings of parental allowance recipients), cuts in the public wage bill (decrease in the

144 Prepared by Thomas Laursen. This section reflects a combination of written comments and presentations by peer reviewers as well as discussions during the conference. 158 number of employees at the level of central government; and contained average wage growth), and other savings through expenditure ceilings at the level of individual budget chapters.

On the revenue side, tax reforms were aimed at (i) supporting the consolidation of public finances; (ii) promoting growth and employment; and (iii) harmonizing the tax system with EU requirements. A rise in indirect taxes (VAT, excises) was compensated by reduction in taxation of capital (CIT) and labour (PIT). The CIT rate was cut from 31 % to 24 %, depreciation periods were shortened, and deductible allowances in the area of R&D were increased. PIT changes would make low-skilled workers and families with children better off through joint taxation of married couples, transformation of all deductible allowances into tax credits, and reduction of rates in the two lowest tax brackets.

On the institutional side, the budgetary process was now anchored in a medium-term expenditure framework (MTEF). It introduced binding multi-annual expenditure ceilings, which were approved for three years together with the budget draft and updated on a rolling basis. The MTEF was a basic instrument to ensure compliance with the medium term fiscal targets. Further, the activities of the so-called transformation institutions that were sources of contingent liabilities had been terminated. The privatisation agency (National Property Fund) would be closed by the end of 2005 and the agency in charge of managing non-performing loans (Czech Consolidation Agency) would be liquidated by end-2007.

The Czech Republic was currently focusing its reform agenda on long-term sustainability of public finances (preparation of pension and health care reforms). A thorough analysis of both systems had been prepared and now the political decision had to be made. The country also wanted to further improve its budgetary framework, including reinforcing the binding medium-term expenditure ceilings through penalties for non- compliance and further emphasis on changing from input-based budgeting towards output and performance budgeting conducive to efficiency gains.

Summing up, Mr. Prouza noted that as full members of the EU, the EU8 countries could not only learn from the “old” members and gather expertise from one another but they could also inspire the “old” members in some areas. The EU8 was lagging behind in some areas (spending on R&D, infrastructure, higher education attainment, and employment rates, especially for older workers), but in other areas they had been more courageous and experimental than the rest of the EU. This included tax reforms (the flat tax in the Baltic States and Slovakia), pension reforms (Poland, the Baltic States, and Hungary), and health care (Slovakia). It was acknowledged that the Czech Republic could still learn from experience in other parts of the region, and Prouza stressed that studies like the one at hand would help to disseminate knowledge, offered an invaluable and welcomed opportunity to exchange experiences, and tried to establish best practices in the region thus creating additional peer pressure beyond that of the EU.

Danny Leipziger noted that fiscal policy was under intense scrutiny in many regions and countries. Key questions included: had fiscal policy over the past decade been pre-occupied with stabilization and neglected growth? Had stabilization come at the expense of public investment and growth? How did the composition of public expenditure influence economic growth and thus the sustainability of public finances? What was an appropriate balance between public investment (and maintenance) and other spending? Were infrastructure investments particularly vulnerable to cuts during fiscal adjustments? Was donor bias skewing financing and expenditure towards the social sectors at the expense of infrastructure? How could demographic pressures for social spending be managed without undermining growth? These diverse concerns had led to the key issue of how governments could create or sustain “fiscal space” for growth.

Policy instruments for fiscal space included revenue enhancement (if marginal benefits exceeded marginal costs of taxation or user fees), deficits financed by debt issuance (if larger debts could be sustained), aid in the form of external grants (if available), and expenditure efficiency improvements (cutting low value programs, reducing waste, and boosting technical efficiency) to generate fiscal savings. Options such as PPPs and monetary financing of the deficit offered limited scope and posed risks. The availability of large amounts of EU funds offered a unique opportunity, but serious technical and administrative capacity constraints had to be addressed to make full and effective use of these funds. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 159

In the EU8, fiscal space challenges arose from the Lisbon strategy (making public finances more growth- oriented) and aging populations. Some countries in the region were in a better position to deal with these challenges owing to strong initial fiscal positions, low debt, and small governments (the Baltic states and Slovakia) while others would have a harder time owing to weak initial fiscal positions, high debt, and big governments (Hungary, Poland, and the Czech Republic). Slovenia was a special case with a strong fiscal position but big government. In all EU8 countries, improving the quality of spending by boosting high productivity spending and cutting consumption was key to growth. Raising revenue was not an option for big government cases but improving the efficiency of taxation was a necessity. This included changing the tax structure (moving away from labor taxes), reducing exemptions and tax expenditures, broadening tax bases, and moving from taxes on mobile production factors toward non-mobile sources.

Managing spending pressures was a major challenge in most countries. Current policy commitments would need to be revised to avoid crowding out space for future public investments. Further, demographic prospects would create pressures for additional social spending, including in education and health. In education, there was growing concern in the region about a trade-off between quantity and quality as well as inequitable access to higher education. The required policy response included increased private financing of higher education (tuition fees, fund raising, etc.), effective student loan schemes, needs-based grants and scholarships based on both merit and need, and enhanced competition among educational institutions. In health, most EU8 countries had costly health care systems that did not control quality or excessive consumption of health services, and debts in the health care system were increasing in several countries. The required policy response included defining a clear basic benefits package, increasing private financing and incentives for reducing consumption, supplementary health insurance for non-covered services, and hospital rationalization. Fiscal Challenges for the EU8 Countries Jan Svejnar welcomed the study as one dealing with important issues, based on careful analysis, and with plausible results and recommendations. At the same time, he noted the data limitations and that the analysis left important questions still to be answered. Regarding the econometric analysis, he cautioned that the small sample could only yield tentative results and that robustness checks were critical. Nevertheless, it was the right way to go as a start.

The key overall ideas in the study were that Central Europe and Baltic countries needed to create fiscal space for development spending to accelerate income convergence, that there were pressures to improve public service delivery (strenghten the quality of human capital formation and create equal opportunities for citizens), and that fiscal space was constrained by the need to maintain sustainable fiscal positions and by limited absorptive capacity for additional public spending. The fiscal challenges were particularly acute in Hungary, Poland, and the Czech Republic, with weak fiscal positions, sizeable debts, and large governments. Strong output growth, low interest rates, and currency appreciation had helped, but there would be a problem if these conditions were reversed. Further, labor taxes were high and population aging would add fiscal stress.

There was scope for rationalizing existing spending programs (subsidies and inefficiencies in social spending and public administration) and for broadening tax bases and sources, which should be geared to increasing employment and reducing poverty while cofinancing inflows of EU regional aid funds. In health care and education, the role of competition was believed to be underestimated. However, the role of flat tax systems and fiscal decentralization remained uncertain. Mr. Svejnar also saw a greater role for immigration in mitigating demographic pressures, as a complement to pension reforms.

In commenting on what he believed was an excellent background chapter on fiscal challenges in the EU8, Filip Keereman noted that it was useful to look at these challenges in the following way: on the one hand there was a need for fiscal consolidation to ensure macroeconomic stability, euro adoption and long-term sustainability of public finances (threatened by aging), while on the other hand, there were pressures on the expenditure side (linked to catching-up, implementing the EU acquis and realising the Lisbon objectives) and on the revenue side (linked to globalisation and tax competition). The circumstances in which these challenges had to be dealt with were difficult as the gradual loss of monetary policy autonomy implied a greater burden on fiscal policy and transition economies were less stable, characterised by more data 160 revisions weighing on forecast accuracy and creating a greater risk of policy mistakes. A positive element in the surrounding environment was the EU which could help reduce the costs of fiscal consolidation through enhanced credibility derived from adherence to the SGP. Furthermore, the EU helped the NMS with the structural and cohesion funds.

The challenges could be met by pursuing the right policies, but the results in the EU8 were mixed. First, general government deficits had been reduced, but mainly thanks to strong growth and lower interest rates so that deficits remained largely of a structural nature. In particular in Hungary and Poland, there had been little effort to decisively tackle government deficits. In these countries expenditures as a percent of GDP had been increasing since 2000, while lasting fiscal consolidations should rather rely on expenditure cuts. Missing budgetary targets should be avoided in order not to undermine credibility. Unfortunately, in several NMS (but not the Baltic States) government balances had fallen short of the objectives put forward in the Pre-Accession Economic Programmes and their successor Convergence Programmes, although the situation had improved recently (except in Hungary).

Second, expenditure composition had improved overall and as such contributed to the quality of public finances. The relative share of subsidies and collective consumption had declined, although in Poland and Hungary less so. The functional classification of expenditure indicated a relatively large share of efficiency- oriented spending and compared favourably to the old Member States. Nevertheless, the expenditure share of GDP remained high compared to countries with similar income levels and suggested a large interference of the government in the economies.

Third, the tax burden had declined with cuts in personal and corporate income tax rates in several EU8 countries along with a shift towards indirect taxes. However, the tax wedge on low income earners remained high, impeding the mobilization of the less skilled part of the labour force.

Fourth, economic and statistical governance had to be improved by creating stronger institutions. This would help avoid that political elections impinged on the fiscal consolidation agenda and reduce data revisions which had been large in the fiscal domain. Greater forecast accuracy and better policy analysis would be the fruits. Assessing Intergovernmental Fiscal Relations Gabor Peteri welcomed the fact that the chapter went beyond analysis of fiscal instruments to discuss some of the institutional aspects of intergovernmental fiscal relations, but felt that these aspects could benefit from further discussion. The establishment of democratic local governments had been of greater significance than just introduction of new techniques of fiscal decentralisation. It had created new rules of the game in the public sector. Depending on the traditions of public administration and governance in the countries of the region, decentralisation had changed the relationship between governments and citizens, politicians and voters, and service providers and customers.

The report did not emphazise sufficiently that the new local governments had contributed to economic growth or at least modernization of public services in the period of economic decline in the first half of the 1990s. Institutional changes had contributed to the development of efficient and innovative local governments. Organisational and management aspects of decentralization had been critical factors of public sector reform. Reflecting political and historical factors, two types of territorial structures had been established: in the Czech Republic, Hungary, Slovakia, Latvia, and Estonia, the size of municipalities was relatively small, while in Slovenia, Poland and Lithuania the municipal governments were less fragmented. This had an important impact on own source revenue bases. Accordingly, the roles and functions of the intermediary level of government mattered a lot. In some countries with fragmented municipal governments, the first wave of reforms had created parallel structures of local and regional state administration. In the Czech Republic, Slovakia, Latvia, as a first step elected local governments were highly dependent on the fiscal and service decisions of local state organs. In Poland and Hungary, the supervisory, auditing and regulatory functions of the state were separated from the local governments making the intermediary level of local governments more autonomous with independent functions and budgets.

Beyond the territorial-administrative structures, intergovernmental fiscal relations were influenced by the transformation of sectoral rules in public service provision. The former “dual subordination” of soviet- Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 161 type local administrations had been gradually modernized. In the 1990s this complex task of creating new lines of supervision in human services (e.g. education); transformation of state-owned public utilities, elimination of sectoral subsidies, and development of new regulatory functions and legal-administrative supervision of local governments had laid the foundation of modern local government systems. Two areas of these broad ranging reforms would have deserved additional attention according to Peteri: first, the privatisation of state-owned enterprises had made local governments owners of physical and financials assets, and modified the basis of delivering local utility services; second, financial management had been forced to go through deep reforms. Regardless of the assignment of functions, the scale of local own source revenues and schemes of intergovernmental transfer, the rules of budgeting, accuracy of accounting and fiscal information, autonomy in spending, and audit mechanisms had all influenced local government finances.

Taking all these above mentioned institutional and managerial aspects as critical components of transformation at local level, Mr. Peteri suggested that perhaps the evaluation criteria might have been expanded to further dimensions. First, the accountability mechanisms, partly through greater transparency, were critical factors of successful fiscal decentralisation. Predictability was more related to the certainty and stability of local government finances. Further, dues, fines and user charges, as relatively stable local revenues, were not discussed among own source revenues.

Within the narrower field of intergovernmental fiscal relations, there were some issues, which could be evaluated in a slightly different way. Shared revenues in most EU8 countries were regarded more as part of intergovernmental transfers and less as local own revenues. The key issue in the system of intergovernmental fiscal relations was the incentives of the transfer system, i.e. the extent to which the revenue assignment compeled local governments to fulfil their primary functions. The three models of grant allocation were: (i) direct control over local expenditures and revenues; (ii) central control over transfers only; and (iii) grant allocation which took into consideration standardised municipal expenditures and local revenue capacity. Under the first model local budget autonomy was more constrained, but national fiscal control was high. When central budget control was managed through general purpose, global transfers, localities had greater discretion. This created better incentives for innovation in service delivery. The third model provided the best incentives to local governments for economizing on expenditures and for the most efficient use of their local revenue base. The first model characterized the educational grants in Estonia and Latvia as well as the calculation of grants in Slovenia. The second model was the basis of intergovernmental transfers in Hungary, Poland and as earmarked current budget grant in the Czech Republic. The third model was partially used for revenue equalisation purposes in Hungary and Poland.

Gabor Peteri also suggested that the chapter could have looked at capital expenditures in the context of debt control and bankruptcy regulations. On average in the EU8, 44% of general government capital expenditures were at the local level. The local discretion in this field was very much influenced by the use of matching grants, the role of extra-budgetary funds and planning rules. This issue was getting greater importance with EU accession, when structural and cohesion funds were made available for local governments.

In concluding, Mr. Peteri noted that the “success” of fiscal decentralization depended significantly on the political momentum of the transition process: those countries that had initiated reforms in the period of major political changes (Poland and Hungary) had been able to achieve more than those who waited (e.g. Slovakia). Looking ahead, the driving forces of intergovernmental fiscal reforms would be the need for higher efficiency of local public services, better service quality and increased equity. These objectives were all linked to territorial-administrative reorganisation, improved design of intergovernmental transfers, and establishment of new policy formulating and regulatory roles of line ministries.

Paul Bernd Spahn noted that all EU8 countries had been willing to embrace democratic models of local governance – guided in particular by the European Charter for Local Self-Government – but that much of the “spirit” of the old regime remained. The governance structure of subnational governments was still in a state of flux in most transition countries. Local governments and their responsibilities were still relatively small, and the role of intermediate levels of government, where they existed, did not always appear transparent. It would be essential to establish or strengthen institutions geared toward fiscal policy coordination, both from the point of macroeconomic stability and efficiency of the public sector as a whole. 162

The extensive list of local government functions raised the question whether municipalities were in a position to fulfill their legal obligations in a “self-governing” fashion. Apart from financing, there was the issue to which extent they were constrained by laws, statutes, standards, norms, and administrative rules as well as other constraints such as collective bargaining contracts. Mr. Spahn viewed the decentralization of functions in transition countries as a process that needed time to converge toward a “steady state”. Although legislation had to be in place to delineate the scope of local government responsibilities a priori, it would be risky to implement all legal provisions in a “big bang”. Initially administrative capacities were often frail, local officials needed time to acquire managerial skills, and democratic processes could still be weak in controlling their actions. The “big bang” approach would risk jeopardizing effective public service delivery, and this could undermine the confidence in local administrations.

In education, it was not clear if the assignment of responsibilities reflected true local self-government. Apart from the setting of teaching standards, which most countries confined to a national institution, key questions were whether local governments owned and managed their assets (school buildings etc.), whether they could determine the number of positions and their qualifications at schools, whether they controlled the hiring and firing of personnel, whether they could set appropriate incentives, for instance through performance contracts, etc. At the same time, a certain degree of central control of primary and secondary education was necessary to ensure high-quality education and that children in poor areas received first-rate education. This would not conflict with the principle of local self-government to the extent that municipalities were allowed to manage their assets, recruit personnel within standard limits, control the quality of the services provided, and provide performance-oriented incentives.

Mr. Spahn thought that health as a local function was overemphasized. The financing of these systems was typically based on the payroll and all had run into severe difficulties due to increasing costs of health services combined with a narrowing financial base. This would drive them into exactly the same problems as faced in countries such as Germany or France. Payroll financed systems also needed to be accompanied by supplementary health schemes for those that fell through the safety net, but while primary health services could indeed be a local responsibility, its delivery through a public institution did not seem to be the most efficient. Mr. Spahn argued for a fully privatized health sector with support for the socially weak, which would require municipalities to care for those that could not afford the service. However, this meant local “provision” of health services, not necessarily “production” of the service.

The most important function of local governments in the present and particularly in the future was considered to be social assistance, including housing for vulnerable groups such as the homeless, orphans, or the old. Social assistance as a local public function was likely to become even more important in the future as some of the national schemes would come under financial strain. Local social assistance had one great advantage over standardized national schemes: it could be targeted to the needy, and it could be run more effectively. Infrastructure was another key area that would become increasingly important.

General revenue was seen as the backbone of local self-government since it constituted unconditional revenue to be spent entirely in accordance with local preferences. This was a precondition for an efficient use of public resources. EU8 countries had not assigned any major taxes to local governments as in Scandinavian countries, for instance. Moreover, typical local taxes (such as the property tax) produced little revenue and allowed little policy discretion, if at all. The bulk of local general revenue came from tax sharing, in particular of the PIT. In this regard, it was crucial to adopt allocation formulae that provided appropriate incentives, or were at least incentive-neutral.

On debt, Mr. Spahn noted that most countries restricted long-term borrowing of local governments to capital formation only, and wondered why Hungary and Poland had chosen a different approach. The use of borrowing for public consumption or transfers entailed macroeconomic risks, not only for subnational public entities but also for the national government. In the case of Poland, he speculated that the answer could be that the authorities considered the quantitative restrictions imposed ex ante sufficient, but stressed that this might be treacherous. At the same time, a corporatist approach to restricting local borrowing reduced the potential for efficiency gains through decentralizing government. This suggested the need for an appropriate coordination machinery that responded more flexibly to needs while maintaining macroeconomic discipline. More flexibility in local borrowing arrangements in turn required appropriate monitoring and reporting instruments. Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 163

Mr. Spahn suggested putting a certain percentage of subnational debt into an interest-free national escrow account. Such “reserves” would serve several purposes: they would act as an automatic reporting and monitoring device; they would increase the cost of subnational borrowing by including a “risk premium” in the form of interest foregone, they would safeguard against insolvency to the extent that debt amortization was partially served from the escrow account; and finally they would allow interest accruing to the escrow account to be used for hedging against risks. Given the risks of moral hazard associated with the latter, the insurance scheme would have to be designed with care. It should include penalties as imposed by several EU8 countries already such as the loss of budget autonomy, forced sale of assets, or the dismissal of local officials.

Looking ahead, Mr. Spahn recommended assignment of broader based taxes to the local level, and even some discretion in setting tax rates. He was sympathetic to revenue schemes similar to those adopted in Scandinavian countries that operated through surcharges on the national income tax with a locally determined rate. More importantly, however, the EU8 countries needed to strengthen institutions and procedures aimed at better cooperation and coordination of policy decisions among different layers of government, both vertically and horizontally. In this area much remained to be done to enhance macroeconomic stability and public sector efficiency. Managing Fiscal Risks in Public Private Partnerships (PPPs) Sweder van Wijnbergen agreed that PPPs could be a useful tool on the interface between the private and public sector and with the authors’ emphasis on proper accounting and transparency. Improper accounting could lead to avoidable surprises which in turn could bring an otherwise useful tool in disrepute, and worse still, improper accounting for risk exposure incurred in PPP projects could increase vulnerability on a macroeconomic scale. A well known example was the Mexican crisis in 1994/95 when failing toll road companies held the government hostage through their impact on commercial bank capital, forcing the government to take on almost 5% of GDP in private debt.

The key message in the chapter, to integrate PPPs and the risks they entailed properly into the fiscal accounts, was the more timely as the EU8 were on the one hand under pressure to increase investment in an effort to accelerate the convergence process, and on the other hand were faced with troublesome fiscal deficits. The temptation to shift risks off-budget through the use of improperly-accounted-for PPPs was a very real one.

Mr. van Wijnbergen felt that the authors might have been a bit harsher on the use of PPPs to “create fiscal space”, i.e. to allow essentially public investment in excess of what was permissible under existing fiscal rules. To suggest that PPPs might help in this respect begged the question of why to adhere to fiscal rules to begin with if their sting was taken out through recourse to what amounted to off-budget financing. PPPs were useful to the extent that they brought the advantages of private entrepreneurship to the execution of public investment projects, but they should not degenerate into loopholes in fiscal rules necessary for macroeconomic stability. The fact that public investment contributed to economic growth was no reason to keep it outside the framework of responsible financial planning for the public sector. This was the more an issue because public investment projects, by their very nature, often did not generate revenues in line with their social returns, which was of course why they were public projects to begin with.

Accounting for PPPs raised a number of issues that merited further research. Valuing future claims was usually not straightforward. For example the authors’ claim that the Black-Scholes (B-S) option pricing formula could be used to price guarantees was doubtful: in many cases the B-S assumption that risk-free arbitrage was possible between the option and a particular set of marketed financial instruments was not valid for the particular guarantee at hand. And once risk-free pricing formulas could not be used, the issue of which discount rate to use had to be addressed. The same issues arose of course in an earlier stage, when public investment projects, whether implemented as PPP or not, were evaluated.

There were also institutional issues when considering PPPs, and the Dutch experience in this regard was thought to be instructive. In the Netherlands, many such projects had been negotiated between lower levels of Government and private parties, but only rarely did the expertise to design, negotiate and implement the complicated contracts necessary in PPPs exist at for example the municipal level. In the Netherlands, an expertise center established at the Ministry of Finance had therefore been involved in such projects. Further, the suggestion to establish an explicit fund to cover contingent liabilities was seen as controversial. 164

Automatic recourse to such a fund, or even the mere existence of such a fund, could improperly encourage risk-taking in the design and/or operation of PPPs.

Vladimír Bezděk congratulated the World Bank for providing an excellent piece of work on PPPs, which could shed new light on benefits and risks of such arrangements. He appreciated the balanced approach to PPPs, noting that they could help close the investment gaps in the presence of fiscal constraints and reduce the whole-of-life investment costs, while at the same time there were risks in that PPPs usually involved long- term government liabilities that were not always transparently recorded in the fiscal statistics. He argued that the net wealth effect of PPPs was unclear due to the long-term character of these and uncertainty about the key economic assumptions. There was still a lack of data and experience that would allow making an ex-post assessment of PPPs, while an ex-ante analysis or forward-looking approach could not avoid being subjective.

In summarizing the discussions, Drahomira Vaskova noted that while PPPs could influence positively the net worth of the government and thus allow for long-term savings, managing the associated risks required a high-quality and on-time institutional and legal framework. Sufficient information and analysis of the long- term fiscal implications of projects were crucial for making the right decisions in the budgetary process. This required detailed information and analysis of potential risks and long-term liabilities which went beyond what was usually carried out. It was necessary to raise general knowledge about the aforementioned issues, not only at the state administration level but also among the decision-makers and general public. It was equally important to increase staff for surveillance and management of fiscal risks. Financing Higher Education The discussion supported the idea that private funding, prices and other market allocation mechanisms played a key role in improving the quality of education in the presence of school autonomy. Variable tuition fees could increase equity and incentives for both students and schools. It was generally agreed that the public-private distinction was no longer valid and that instead it was more relevant to compare “good quality” and “poor quality” institutions with the same levels of budgetary financing.

Jaak Aaviksoo stated that the chapter on financing higher education was an important review of experience in EU8 countries and framed the issues for policy makers very clearly. The need to develop a higher education system that could compete with the US and Asia was at the forefront of public debates and European governments were increasingly focusing on how to ensure that private finance would contribute to the development of a flexible and innovative system accountable to stakeholders. This debate about how to develop a sustainable higher education system in Europe continued in the environment where mass enrollments and a rapid increase in private and, in some countries, overseas provision was not being matched by comparable investments in higher education institutions. Expansion without investment also threatened quality. The lack of transparency in how public finances were allocated and used made it hard to ask value-for-money and was a grave concern. Progress had been made in the UK, experimenting with transparent allocation and financial accounting systems to increase accountability while ensuring autonomy. Mr. Aaviksoo also discussed the necessity for tuition fees and performance based financing, topics that were further expanded by the other discussants.

Tom McCarthy commented that the report was correct in highlighting the problem that when government policies dictated that no fees could be charged, the funding deficit would inevitably grow. EU8 countries needed to find efficient and equitable allocation mechanisms for existing state financing, linked to public policy objectives, which could ensure transparency and foster competition linked to performance. It was not going to be feasible for the private sector to take up the excess demand because private institutions would not be able to meet the high investment costs to develop the scientific and research capacity needs of a knowledge and innovation society. He went on to make the point that, even in the US, there were virtually no totally private institutions and that most US higher education institutions were in receipt of substantial public subsidies, mostly in the form of research grants while also receiving substantial private flows from fees and endowments.

A variety of public funding mechanisms was desirable so that universities could both develop talent and foster innovation. These mechanisms could include: competitive funding for new initiatives; experimental funding for innovation; research funding to encourage institutional competition; and specific projects Current Issues in Fiscal Reform in Central Europe and the Baltic States 2005 165 and grants to support individuals. Mr. McCarthy noted that the report might usefully have discussed the role of vouchers as well as fees as a way of attracting more private finance and increasing competition among higher education institutions. He agreed with the need to create stronger higher education institutions, differentiated by institution and type of course, but stressed that EU8 governments should ensure real as opposed to apparent autonomy for institutions. He also agreed that there was no single right answer: institutions should, in all cases, be free to select their expenditure priorities subject to transparent budgeting practices with well defined rules to channel extra-budgetary resources.

In summarizing the discussions, Stepan Jurajda noted that there was broad concensus that private funding, prices and other market allocation mechanisms should play a key role in improving the quality of education in the presence of school autonomy. Tuition fees could increase equity and incentives for both students and schools. Insufficient transparency in financial accounting of school systems (not comparable across countries, nor clear even within countries) was a problem, there were extreme cross-school differences in the content of „tertiary education” in countries that experienced dramatic expansion; and if schools were controlled neither by government nor by student choice (tuition), incentives would suffer. Mr. Jurajda added that student loan schemes were indeed technically feasible in the EU8, that tuition fees could help reduce high drop out rates, that not only dual-track but also single-track all-free systems penalized students from non-priviledged families, and that regional location of colleges in countries where these were and all were in excess demand could have consequences for access (fairness) and lead to increasing concentration of college educated in a few areas (human capital externalities). Controlling Health Expenditures Andrzej Rys provided a more positive assessment of health care reforms in the region than that presented in the paper, noting that some of these reforms had been the most innovative in Europe (including general practitioner reforms in Poland and Slovenia). He acknowledged that some hospital debt problems were serious, but at the same time pointed to the need for new capital investment in the sector. In particular, there were serious regional imbalances in many countries, which needed to be addressed.

Miroslov Zamecnik focused his remarks on the various forces opposing reform in the health sector. These included drug companies, providers (including hospitals), “insurers”, and politicians. The stated grounds for their opposition were that market-based reforms could not ensure equity and universal access to health services, since the market would necessarily respond to price signals and preferences expressed by consumers and thus allocate resources according to ability to pay rather than according to need. Competition among insurers would result in aggressive marketing and high non-clinical costs and eventually be wasteful. Likewise, a rational market response would encourage providers to adopt a more business-like stance and use monetary scales to make clinical decisions. Consumers were unlikely to ever be able to process the kind of information necessary to make the best decisions, and would exercise their choice of providers and insurers based on superficial offerings. However, given the poor performance of the state-managed health care system in the financing and delivery of health services, huge productivity differences between insurers, the continuous bailing out of the health system by the state, and uncontrolled increases in pharmaceutical expenditures, he argued that perhaps the introduction of competitive forces in the health sector was necessary and unavoidable. This should, of course, be accompanied by strong regulation to ensure that the market failures usually associated with the health sector did not have a negative impact on consumer welfare.

In summarizing the discussions, Lucie Antosova noted that participants had stressed the issue of fiscal sustainability and that it was therefore a concern that in most of the EU8 countries EU8 the stock of healthcare sector debt was still rising and that healthcare budgets in general were not ready to deal with medium-long term challenges. While to some extent the issues faced were similar to those of the EU15 (population ageing, new medical technology, rising pharmaceutical costs, etc.), the EU8 countries still had to deal with some of their own, inherited problems (excessive hospital infrastructure, rising salaries of medical personnel, etc.). There was also an issue of system fragmentation in terms of care integration, payment mechanisms, and often unsystematic capital investment. While there was widespread consensus at the technical level about what needed to be done, mustering the necessary political will was key to addressing these challenges. Concluding and Other Remarks In the concluding panel discussion, Ondrej Schneider noted that the EU8 countries could be divided into two groups: those with a large government (the Czech Republic, Hungary, Poland, and Slovenia) and those with 166 small(ish) governments (Estonia, Latvia, Lithuania, and Slovakia since 2003). Along a different dimension, there were those in breach of the SGP (the Czech Republic, Hungary, Poland, and Slovakia) and those compliant (the Baltic countries and Slovenia). On both scores, the “bad” cases were the Czech Republic, Hungary, and Poland, and the “good” cases the Baltic States, while Slovenia and Slovakia were intermediate cases.

Based on data for 2000-04, he calculated sustainability gaps (difference between actual and debt stabilizing primary surpluses) ranging from more than 5 percent of GDP in the Czech Republic to a negative gap in Estonia. He also calculated “fiscal effort” (netting out the impact of growth and monetary policy on fiscal adjustment) during this period – positive in Estonia, Slovenia, and Latvia, and negative in the other EU8 countries (notably Poland and Hungary). Further, he examined successful (large and sustained) fiscal consolidations in the region, finding that the successful episodes reduced the deficit more, and did not rely on revenue increases but rather “painful” expenditure cuts (less on investment). The most reform-minded countries had been Estonia and Slovakia, while there had been almost no reforms in Hungary and the Czech Republic. Main reforms had been in health care, tax systems, fiscal decentralization, and budgetary rules and procedures.

Andreas Woergoetter and Witold Orlowski highlighted the importance of political economy constraints to reform and noted that building effective fiscal institutions was key in this regard. Reforming public sectors, where not much had been achieved yet, was a fundamental part of this process. Mr. Woergoetter noted that it was an irony that reducing wasteful spending was considered “painful”.

The representative from the International Monetary Fund warned that the EU8 countries should not be satisfied with stabilizing debt around the 60 percent of GDP Maastricht level, but rather should aim for considerably lower debt levels given their persisting vulnerability to external shocks and higher macroeconomic volatility. In the same vein, the representative from the ECB noted that other emerging markets might be better benchmarks than the old EU or OECD countries. The representative from the EIB felt that the discussion of PPPs focused excessively on fiscal risks while most new projects relied on availability payments which were fully known, although – as the Portuguese experience demonstrated – it was crucial to plan for these future payments.