40541 W O R L D B A N K W O R K I N G P A P E R N O . 1 1 7 Social Assistance in the New EU Member States Strengthening Performance and Labor Market Incentives Dena Ringold Leszek Kasek THE WORLD BANK W O R L D B A N K W O R K I N G P A P E R N O . 1 1 7 Social Assistance in the New EU Member States Strengthening Performance and Labor Market Incentives Dena Ringold Leszek Kasek THE WORLD BANK Washington, D.C. Copyright © 2007 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First Printing: June 2007 printed on recycled paper 1 2 3 4 5 10 09 08 07 World Bank Working Papers are published to communicate the results of the Bank's work to the development community with the least possible delay. The manuscript of this paper therefore has not been prepared in accordance with the procedures appropriate to formally-edited texts. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the International Bank for Reconstruction and Development/The World Bank and its affiliated organizations, or those of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank of the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will normally grant permission promptly to reproduce portions of the work. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, Tel: 978-750-8400, Fax: 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, Fax: 202-522-2422, email: pubrights@worldbank.org. ISBN-10: 0-8213-7159-2 ISBN-13: 978-0-8213-7159-6 eISBN: 978-0-8213-7160-2 ISSN: 1726-5878 DOI: 10.1596/978-0-8213-7159-6 Dena Ringold is a Senior Economist in the Social Protection unit of the Latin America & Caribbean department of the World Bank. Leszek Kasek is a Consultant in the Poverty Reduction & Economic Management unit of the Europe and Central Asia department of the World Bank. Library of Congress Cataloging-in-Publication Data has been requested. Contents Acknowledgments vii Abbreviations ix Executive Summary xi 1. Introduction 1 Social Assistance in the EU8 1 Social Assistance and the Lisbon Social Inclusion Process 2 Scope of this Report 3 Data and Methodology 3 2. The Context: Poverty and Social Exclusion in the EU8 7 Poverty and Social Exclusion 7 Implications for Safety Nets 10 3. Social Protection in the New Member States 13 Social Protection Expenditures and Financing 13 4. Performance and Effectiveness 21 Coverage: How Many People Receive Social Protection? 21 Targeting: Does Social Assistance Reach the Poor? 23 Adequacy: Does Social Protection Alleviate Poverty? 27 5. Design of Social Assistance Programs 29 Administration of Targeted Benefits 29 Decentralization 30 Labor Market Incentives 31 6. Making Work Pay: Work and Social Protection 35 Unemployment Traps 35 In-Work Benefits 41 7. Conclusions 43 Appendixes 45 A: Family Benefits in Hungary 47 B: Methodology: The Consumption Aggregate 51 References 67 iii iv Contents LIST OF TABLES 1. Poverty Rates in Comparison xii 2. Summary of Household Budget Surveys 5 3. Poverty Rates in Comparison 8 4. Key Labor Market Indicators in the EU8 Countries 10 5. Estimated Roma Populations in the EU8 10 6. Distribution of Benefits by Quintile (contributory and non-contributory benefits), 2004 24 7. Marginal Effective Tax Rates for Various Types of Households in the EU8 Countries in 2004 (percent) 40 A.1. Expenditure on Family (Maternity and Child) Benefits (millions of Hungarian forints) 49 A.2. Social Protection Spending in the EU8, 2003 52 A.3. Minimum Income Guarantee Programs (percent of GDP) 53 A.4. Minimum Income Guarantee Programs, Beneficiaries (percent of population) 53 A.5. Spending on Child Benefits (percent of GDP) 53 A.6. Coverage of Child Allowances (percent of population receiving) 54 A.7. Summary Indicators, Social Protection Benefits 55 A.8. Summary of Programs included in the Household Budget Surveys 58 A.9. Child Allowances, Main Features, 2005 61 A.10. Guaranteed Minimum Income Programs, Main Features, 2005/6 63 LIST OF FIGURES 1. Aging Populations in Europe and Central Asia (percent of population 65 and over) 8 2. Total Social Protection Expenditures 2001 and 2003 (percent of GDP) 14 3. Composition of Social Protection Spending, 2003 (percent of total) 15 4. Pension Spending and Average Exit Age from the Labor Force in the EU 15 5. Share of Cash and In-Kind Benefits 16 6. Spending on Guaranteed Minimum Income Programs (percent of GDP) 17 7. Coverage of Guaranteed Minimum Income Programs (percent of population, latest year) 18 8. Changes in Total Fertility Rates in Europe and Central Asia, 1950 and 2000 19 Contents v 9. Child Allowances (percent of GDP), 2004 20 10. Coverage of Social Protection Benefits, 2004 (percent of individuals in households receiving a benefit) 22 11. Coverage of Social Assistance and Family Benefits, 2004 (percent receiving benefits) 22 12. Share of Means-Tested Benefits, 2003 23 13. Targeting of Family Benefits and Social Assistance, 2004 (percent of benefits going to the bottom and top quintiles) 25 14. Targeting of Social Assistance in Comparison (CGH Index) 26 15. Adequacy of Social Protection Benefits: Bottom Quintile (average benefit as a share of average consumption for each beneficiary group) 27 16. Riga Residents Have a Disproportionately High Likelihood of Receiving Local Government Benefits 32 17. Benefits and Minimum Wages, 2004 (average monthly benefit as percent of minimum wage) 33 18. Unemployment trap in the EU8 in 2001 and 2004 (percent; earners of 67 percent of the APW) 37 19. Low-Wage Trap in the EU8 in 2001 and 2004 (percent; earnings increase from 33 to 67 percent of the APW) 38 20. Marginal Effective Tax Rates for Various Types of Single-earner Households in Visegrad Countries in 2004 (percent) 39 LIST OF BOXES 1. Definitions 4 2. Differences between the ESSPROS and National Accounts 6 3. Slovakia's New Social Policy 18 4. The Slovakia Social Development Fund 34 5. Measuring Marginal Effective Tax Rates on Labor 36 Acknowledgments his report was prepared by a team consisting of Dena Ringold (Team Leader), Leszek T Kasek, Elena Rydvalova, and Dorota Holzer-Zelazewska. A team of experts collected data at the national level, including: Ondrej Schneider (Czech Republic); Reelika Leetmaa and Andres Vork (Estonia); Orsolya Lekes (Hungary); Tana Lace (Latvia); Aurelija Jakaitiene (Lithuania); Anna Ruzik (Poland); Tine Stanovnik (Slovenia); and Katarina Kotovova (Slovakia). Taras Pushak provided indispensable help with the household budget survey database. Peer reviewers Ruslan Yemtsov, Emil Tesliuc, and Sándor Sipos (The World Bank, Washington, DC) provided excellent comments and guidance. The draft also benefited from the comments and discussion at the workshop on the EU8 Social Inclusion Study held May 8-9, 2006 in Warsaw, Poland. The authors are grateful to Nicholas Barr (The London School of Economics), Bernard Funck (The World Bank, Washington, DC), Thomas Laursen (The World Bank, Warsaw), and Lucian Pop (The World Bank, Bucharest) for their valuable comments during the preparation of this paper. The report is part of a series of studies on current issues in public finance reform in the Central European and Baltic countries (Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, and Slovenia--the "EU8") which joined the European Union on May 1, 2004. These studies have been undertaken since 2005 and coordinated and edited by Thomas Laursen, Lead Economist for Central Europe and the Baltic States in the World Bank. Marta Michalska provided excellent administrative and logistical support through- out the process of preparing these studies. vii Abbreviations ALMP Active labor market program APW Average production wage in manufacturing (APW) CSO Central Statistical Office DG Directorate General (of the European Commission) ECA Europe and Central Asia ESSPROS European System of Integrated Social Protection Statistics EU European Union EU8 The 8 new EU member states in Central Europe and the Baltics (Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, and Slovenia) EU-15 The 15 EU member states pre-2004 enlargement EU-25 EU member states after 2004 enlargement HBS Household Budget Survey IWB In-work benefit GMI Guaranteed minimum income MISSOC Mutual Information System on Social Protection MOF Ministry of Finance NAP National Action Plan on Social Inclusion OECD Organisation for Economic Co-Operation and Development OMC Open Method of Coordination PPP Purchasing power parity SILC Survey on Income and Living Conditions ix Executive Summary ocial protection and social assistance are important elements of the EU's Social Inclusion S process. Within a very short period of time, the new member states of the EU have put in place programs and institutional arrangements for delivering income support and other benefits to the poor. Social assistance programs and policies for tackling poverty and exclusion exist across countries and are central elements of each of the countries National Action Plans for Social Inclusion. Although means-tested social assistance benefits make up a relatively small share of GDP and total spending on social protection, they are important sources of income for those who receive them. This report looks at social assistance in the new member states from the perspective of fiscal impact and effectiveness. Given the limited budgets for social assistance, ensuring that resources are spent well, reach those who need them, and optimize welfare gains and labor market outcomes are important tasks. Attention is also required to the balance between spending on means-tested anti-poverty programs and family benefits and social insurance programs. Managing the long-term costs of these programs to meet emerging needs is a priority. Despite the importance of social assistance, there is little comparative analysis of programs across countries. This is needed to assess effectiveness at the country and regional levels, as well as for capturing experience for subsequent waves of EU accession countries and for sharing with the EU15. The EU's Lisbon process for Social Inclusion is based on policy coordination and experience-sharing. However, as the EU noted in a recent review of the National Action Plans for Social Inclusion in the new member states, comparable information on coverage of social assistance and adequacy of benefits is not available. This report draws from national and European level databases to present a picture of national level spending on social protection programs, with a specific focus on social assistance and family benefits. It reviews the performance of social assistance and social protection based on household survey data for six of the countries, as well as information on program design collected at the national level. Poverty and Social Exclusion in the EU8 Social protection systems in the new member states need to meet the challenge of responding to changing and diverse needs within the countries. Socioeconomic conditions in all of the EU8 countries transformed rapidly across the transition period and continue to be dynamic. From a social protection perspective, rapidly ageing populations will influence the size of the sector over the longer term. Based on projections, between 20 and 25 percent of the populations of the EU8 countries will be 65 and over by 2025. Poverty remains a serious concern in the EU8 countries. Poverty rates vary depending on the methodology--whether it is measured in absolute or relative terms (Table 1). Using an international line of $4.30 per day in PPP terms, poverty ranges from 11 percent in Hungary to 21 percent in Poland. The EU measures poverty using a relative measure of 60 percent of the national median equivalised income. Using this line, poverty ranged from 8 percent in the Czech Republic to 18 percent in Estonia in 2003. xi xii Executive Summary Table 1. Poverty Rates in Comparison Laeken at-risk of International poverty National poverty rate (2003)1 line, $4.30 PPP poverty rate3 Czech Republic 8 -- -- Estonia 18 18 -- Latvia 16 14 5.9 Lithuania 15 22 -- Hungary 12 11 -- Poland 17 21 14.8 Slovakia 13.32 -- -- Slovenia 10 -- -- Sources: Laeken indicator from Eurostat; $4.30PPP based on HBS with OECD PPP rates. Notes: (1) At-risk of poverty rate is 60% of national median equivalised income. (2) Preliminary data from EU-SILC. (3) Latvia data from World Bank (2006b); Poland from World Bank (2004). There are severe pockets of poverty among certain groups. Poverty is closely correlated with labor market status, educational attainment, and household size. Other risk groups identified in World Bank poverty studies and in the National Action Plans for Social Inclusion include people in lagging regions (poorly served by basic services and poorly integrated into markets); linguistic and ethnic minorities including Roma, but also Russian speakers in Latvia and Estonia; the disabled; homeless; and those with serious health risks. The specific characteristics of poverty in the EU8 have significant implications for the types of social assistance programs and policies which are required. The dynamic environment calls for programs which can be flexible and responsive to changing conditions. Labor market conditions continue to fluctuate in the EU8 countries. While EU integration has been a source of stability, the EU8 countries remain at risk of shocks which can affect poverty and welfare. Thecloselinksbetweenpovertyandthelabormarket,andparticularlytheserious issues of long-term unemployment and working poverty have implications for policy design. Social assistance benefits are of critical importance for the long-term unemployed, as unemployment benefits in all countries have time limitations. Once insurance benefits are exhausted, beneficiaries become eligible to apply for social assistance. Many countries also have a range of active labor market measures targeted at the long-term unemployed. However, as evaluations have found that some active labor market measures are not effective, these programs need to be closely monitored to ensure their effectiveness. High numbers of working poor call for attention to benefits which help to"make work pay" and provide support to individuals and households even if they have a job. The EU8 countries have adopted a wide range of options to support the work- ing poor including, tax credits for working households (Estonia), social assistance benefits which taper off with wages (Slovakia, Hungary) and child care benefits and Executive Summary xiii other support to families. This is an area where monitoring and evaluation of early lessons and experience will be essential, especially to ensure that they are cost-effective. Increasing regional disparities underscores the need for programs which can respond to local needs. The EU8 countries have adopted significant reforms toward decentralization of their intergovernmental fiscal frameworks and public adminis- tration systems to increase the roles of local governments in social services-- including financing and delivering cash and in-kind social assistance benefits. Transfer arrangements that include earmarked resources for social assistance (including for local co-financing of central government allocations) may be a use- ful approach. Similarly,designofsocialassistanceneedstoensurethatprogramsmeettheparticular needs of vulnerable groups. Roma and other ethnic minorities need services which take into account their distinct needs and preferences. Outreach and design of services need to tackle exclusion and discrimination faced by these communities. Linguistic groups require language support to facilitate their access to services, while the disabled will need other types of support depending on their condition. Social Protection in the New Member States Social protection, encompassing pensions, unemployment benefits, family benefits, and targeted social assistance programs have an important impact on the welfare of the popu- lations in the new member states. Nearly all households across countries receive at least one type of benefit. Social protection is also significant from a fiscal perspective. Total spending on social assistance and social insurance, comprises a significant share of government resources across countries. Based on spending levels, countries clustered into three groups in 2003. The Baltic States (Estonia, Latvia, and Lithuania) spent about 13 percent of GDP, the Visegrad countries (Hungary, Czech Republic, and Slovakia) between 20 and 21 percent, and Slovenia over 24 percent of GDP. In all cases, pension spending makes up the largest share of the budget, ranging from 42 percent of total social protection in Slovakia, to 68 percent in Poland. Total social protection spending in the EU8 countries remains below the EU15, which stood at 27 percent of GDP in 2003. Between 2001 and 2003, social protection spending contracted alongside total government expenditures in the Baltic States, Slovakia, and Slovenia, and expanded in the Visegrad countries, particularly Hungary. Across the EU8, guaranteed minimum income (GMI) programs have emerged to form the core social assistance program of last resort for the poorest households. Minimum income benefits are targeted means-tested benefits which are paid monthly, or on a one-off emergency basis, to poor households. GMI programs cover between 1 and 4 percent of the total population and comprise 0.1 to 0.5 percent of GDP. Family benefits are another important element of social safety nets in the EU8. Family benefits have wider coverage than social assistance programs and are well targeted because of the close connection between poverty and household size across countries. Coverage.CoverageofsocialprotectionishighintheEU8.In2004coverageranged from 90 percent in Hungary to 62 percent in Lithuania. The poor (the bottom quintile xiv Executive Summary of the population) receive social protection at a higher rate. Coverage is driven by pensions, which are received by 30 to 45 percent of the population. Benefits targeted at the poor are much less prevalent. Family benefits reach about half of the population in Estonia, Hungary and Latvia. Social assistance programs targeted at low-income households are very small across countries. The share of the population receiving means-tested social assistance is similar across countries--ranging from 4 to 9 percent of the bottom quintile, or 2 to 5 percent of the total population. Targeting.Contributoryinsurance-basedbenefitsaccruemoretotheupperincome deciles, while non-contributory benefits, such as family benefits and social assistance have higher incidence among the poor, except in Poland. Social assistance and family benefits appear to be well targeted in the new member states, although there is variation across countries, and room for improving targeting to the poorest households. In Lithuania and Hungary, 42 to 56 percent of social assistance benefits went to the bottom 20 percent of the population in 2004. In contrast with the other countries, in Latvia and Lithuania there is a higher level of leakage to the top quintile--24 and 15 percent in 2004 respectively. In both countries, this leakage represents a deterioration in targeting performance since 2002. Leakage is also high in Slovenia, with 20 of resources going to the top quintile. Adequacy.Average social protection benefits amount to between 27 (Latvia) and 65 (Poland) percent of average household consumption for those who receive social protection. For those individuals in the poorest quintile, benefits make up as much as 91 and 92 percent of average consumption in Estonia and Hungary. Although their main objective is consumption smoothing, pensions have the greatest welfare impact. Even among the bottom quintile, pensions make up over 80 percent of the average consumption of those who receive them. However, the majority of the population in the bottom quintile does not receive pensions, and relies on social assistance, which is much less generous. Labor Market Incentives A critical issue for the design of GMI programs is the need to avoid welfare dependency and ensure that beneficiaries have incentives to participate in the labor market. Labor mar- ket outcomes are a critical concern for social protection systems as a whole. Labor market participation can be encouraged by limiting the duration of benefits. While the EU8 coun- tries have limitations on how long beneficiaries can receive benefits before reapplying (ranging from one to six months) none of the countries appear to enforce a fixed cut-off. Activation measures aimed at supporting labor market integration have been an increas- ing focus within social assistance policies. Benefit Design.Within benefit programs,labor market incentives can be increased by keeping benefits lower than wages, however, there is a tradeoff with adequacy, and the need to make sure that benefits are also effective sources of poverty relief. Another mechanism is to link social assistance beneficiaries with labor offices. Unemployment Traps. In addition to benefit design issues, current tax and social policy regulations appear to discourage formal sector employment in most of the Executive Summary xv EU8. Unemployment traps--measured as the share of gross earnings which is taxed away through higher tax and social insurance contributions and the withdrawal of social benefits--in Latvia, Poland and Slovenia are even higher than in the EU15. In Latvia and Poland they have even increased in recent years. In these countries, the effective marginal tax rate for an unemployed person returning to work is around 80 percent of gross earnings--including the effect of taxes and reduced benefits. In-WorkBenefits.Thehighunemploymenttrapsforlowwageworkersandfamilies suggest the need for tax credits and other in-work benefits (IWBs) which can support low-income households and facilitate the transition into employment without reinforcing dependency. Estonia has introduced tax credits for families with children, and Hungary and Slovakia have benefits which taper off as individuals enter the labor market. Conclusions and Policy Recommendations Within a very short period of time, the new member states of the EU have put in place programs and institutional arrangements for delivering income support and other benefits to the poor. Social assistance programs and policies for tackling poverty and exclusion exist across countries and are central elements of each of the countries National Action Plans for Social Inclusion. Although means-tested social assistance benefits make up a relatively small share of GDP and total spending on social protection, they are important sources of income for those who receive them. Policy priorities emerging from this analysis are as follows. Further efforts will be needed to reduce the overall envelope for social protection, especially given demographic pressures, social exclusion and poverty challenges. On aggregate, social protection consumes a significant share of GDP in the EU8. High spending has costs, including crowding out fiscal space and limiting employment growth. Pension reforms are important first steps, although the fiscal impact has not been substantial to date. The Baltic states and Slovakia have made substantial strides towards consolidation and streamlining of social protection systems, while Poland, Hungary, the Czech Republic and Slovenia have much larger systems with significant potential for efficiency gains. The challenge across countries is ensuring that social protection reaches those who need it and does not distort employment incentives. However,reformstocutexpendituresandstreamlinesocialprotectionshouldnotcome at the expense of well-targeted social assistance programs which provide essential sup- port to the poor. There may be a rationale for increasing spending on effective social assistance programs to reach more of the poor and provide increased support. Cov- erage of means-tested social assistance among the poor in 2004 ranged from 4 to 9 percent of the bottom quintile, or 2 to 5 percent of the population. Spending on minimum income programs--the main social assistance program in all countries-- ranged from 0.2 to 0.5 percent of GDP in 2003. In general, many social assistance programs are relatively well targeted, with between 30 and 60 percent of resources xvi Executive Summary going to the poorest quintile of the population. For those who receive them, benefits can make up as much as 37 percent of average consumption of the poor. However, there is variation across countries and substantial scope for improving targeting, effectiveness and adequacy of programs. More can be done to reduce leakage and expand coverage among the poor. Social assistance in Latvia stands out as particularly weak in terms of targeting, with one-quarter of resources going to the richest quintile of the population. Benefit levels are also significantly lower than in other countries. Recent declines in targeting performance in Lithuania and Estonia-- which are among the top performers in targeting of social assistance--merit monitoring and further analysis. Familybenefitscanmakeanimportantcontributiontotacklingchildpoverty because of the tight link between poverty and household size. All of the National Action Plans for Social Exclusion identify reducing child poverty as a priority. Although programs are only means-tested in Poland, Slovenia, and the Czech Republic, benefits are highly progressive. Family benefits also make up an important source of income for poor households, especially in Hungary, where the network of family benefits is more extensive than in other countries. The important role of family benefits as income support for the poor underscores the potential for greater coordination of social assistance and family benefits programs--possibly through further consolidation of programs and/or harmonizing of eligibility criteria to reduce administrative costs. Attention to implementation arrangements is critical for improving targeting and reducing administrative costs. Further program-specific analysis is needed to assess what works in each country context, but existing analysis and experience of reforms in the region point to the importance of implementation and delivery mechanisms for good targeting. Well-designed information systems, which link social assistance to other databases on social protection and taxation, can sharpen targeting through better information on beneficiary eligibility. Public information and outreach are essential for informing potential beneficiaries about programs, as well as reducing the stigma surrounding social assistance programs among the wider public. For difficult to reach excluded populations--such as geographically isolated commu- nities, or linguistic minorities, specially designed information campaigns an be important. Decentralization of social assistance needs to be carefully phased and managed to ensure effectiveness and equity. While decentralization of social assistance may improve targeting, if local officials have better information and eligibility: it also entails risks if poorer municipalities are less able to finance social assistance. In Latvia, while the decentralized GMI program is well targeted at the national level, local governments in relatively richer regions, such as Riga, have considerably more funds available to spend on these programs than those in poorer regions. Central- ized financing with earmarked transfers are important to ensure a safety net for the poor, both by making sure that resources reach the areas that need them and by ensuring that funds are not diverted for other purposes. Poland has plans to decen- tralize social assistance starting in 2007, and the possibility has also been discussed in Slovakia. Buildinglabormarketincentivesintosocialassistanceprogramsisessential,especially because of low employment levels and pervasive long-term unemployment in the EU8. Executive Summary xvii At the national level, social assistance benefits appear low in relation to minimum wages. However households may receive more than one benefit, and aggregate ben- efit levels may create poverty traps. Variations in wage levels across regions may also create poverty traps in some geographic areas. The recent social assistance reforms in Slovakia reformed benefit levels to eliminate poverty traps which were particularly high in some regions. All of the countries require that social assistance beneficiaries register with employment offices, and some go further (for example, Slovenia) in requiring participation in an active labor market program. Activationmeasuresareotherimportantinstrumentsforgettingpeopleintowork.As in the EU15, active labor market policies are gaining popularity in the EU8, particularly given resources made available through the European Social Fund. Slovakia has introduced a wide range of programs, including transportation subsidies for attending job interviews, small business support for the unemployed, child care, and training. The focus on activation measures will require greater coordination between social assistance and employment offices. In Slovakia, this was addressed through actual merging of offices. This approach will be important for other countries. In-work benefits which support the transition from welfare to work and alleviate unemployment traps can help address significant levels of working poverty. Recent poverty analysis found that--with the exception of Hungary--there are more working adults among the poor than non-working adults in the EU8. Disincentives to enter into employment in the first place are also high--especially for low-wage workers in Poland. Estonia has introduced tax credits for families with children, and Hungary and Slovakia have benefits which taper off as individuals enter the labor market. In the EU15, these programs include tax credits and lump sum benefits. Thedatabaseforanalysisofsocialassistancehasimprovedgreatlyinrecentyears,but further efforts are needed. The availability and quality of administrative and survey data for analysis of social assistance are vastly improved from the early transition period. The EU8 countries also benefit from being included in cross-country databases and data collection efforts coordinated by Eurostat including ESPROSS for expenditure data and MISSOC for program data. The introduction of the EU-SILC surveys will similarly improve the database on programs and outcomes. However, because of the small-size and diversity of programs, intense efforts are needed to ensure that country-level data are available for rigorous outcome evalu- ation of the targeting and effectiveness of programs overtime, particularly given dynamic changes in the landscape of poverty and exclusion in the EU8 countries. CHAPTER 1 Introduction Social Assistance in the EU8 From a fiscal perspective, social safety nets in the new member states appear almost insignificant. Spending on targeted social assistance benefits for the poor ranged from 0.1 to 0.5 percent of GDP in 2003.1 However, while small in size, these programs are important for the households that receive them. A recent regional study on poverty found that at least 40 percent of poor households in Poland and Hungary received social assistance benefits (World Bank 2005a). Benefits can also be an important contributor to the income of poor households. Safety nets have a critical role to play for the poor, and have a potentially wider impact on welfare beyond income support. Targeted assistance can help low income households meet expenses to send their children to school, pay for health care services, and can provide a springboard into employment, as a temporary safety net for the unemployed and those isolated from the labor market. They also provide assistance and consumption smoothing for the working poor and households facing temporary income shocks such as illness. Safety nets do not operate in isolation, rather they fit into a broader context of social protection and labor market policies. All of the EU8 countries have fully developed social protection systems which do consume a significant amount of resources--ranging from 13 to 24 percent of GDP in 2003. Systemic reforms prior to accession transformed the EU8 welfare states from "cradle to grave" socialist systems, where nearly all income support was tied to the workplace, to European-style welfare states including social insurance (pensions and unemployment benefits), and social assistance cash benefits (such as guaranteed minimum income programs) and in-kind services to protect the poor (Barr 2003). 1. Spending on guaranteed minimum income programs, based on national data. 1 2 World Bank Working Paper Targeted social assistance is still new in the EU8 countries. In a remarkably short period of time, countries have put in place programs and institutional arrangements for delivering income support and other benefits to the poor. There are three main approaches to targeting: First, means testing which includes individual or household income-testing, and may or may not include some type of assets test; Second,proxymeans-testing,orindicator-targeting,whereeligibilityisbasedupon one or more indicators correlated with poverty which are easier to observe than income, such as household size and geographic location. Categorical benefits, such as child allowances which are granted to households with children may be regarded as a form of indicator targeting; Finally, self-targeting in which individuals self-allocate benefits based upon their own decisions and choices. Social assistance benefits may be self-targeted if the opportunity costs (for example, time spent on paperwork, waiting) are high and wealthier households do not bother to apply. Social Assistance and the Lisbon Social Inclusion Process Social protection and social assistance are important elements of the EU's Social Inclusion process. The acquis communautaire, the EU's body of legislation, which formed the "rule book" for accession, as well as the Copenhagen accession criteria, did not include specific criteria in the area of social assistance policies. Rather, social assistance, and social protection policies of member states are addressed through the Open Method of Coordination (OMC) process which was developed following the 2000 Lisbon Summit. Through the Lisbon process, each member state is required to submit a National Action Plan (NAP) for Social Inclusion to the Commission every two years, detailing how it intends to fulfill the Lisbon poverty reduction objectives. The NAPs outline the major challenges, strategies, objectives, policy measures, indicators, and good practice examples. The EC reviews the NAPs and, in consultation with the member states, drafts a Joint Inclusion Report on poverty and social exclusion, which is adopted by the EU Council. In the area of social protection, the NAP's detail how existing policies facilitate social inclusion and outline future challenges. A number of common themes emerge which are also discussed here including: (i) balancing passive income support with activation measures; (ii) supporting working families through policies which make work pay and make it possible for women with children to participate in the labor market; (iii) integrating excluded groups through targeted initiatives, including those for the disabled and linguistic and ethnic minorities; (iv) introducing preventative measures which aim to avert exclusion from risks including alcoholism, drug abuse, and domestic violence, criminality, and other risks; and (v) focused efforts to reduce and prevent child poverty. The National Action Plans detail how each of the countries plans to make use of the EU Structural Funds (SFs), which are intended to provide additional financing to the member states to assist them in meeting the objectives included in the NAPs. The amount of incoming resources is substantial. In total, 15 billion were programmed for the EU8 Social Assistance in the New EU Member States 3 countries between 2004­06. The European Social Fund is the main SF for financing social inclusion-related programs (3.6 billion total). However, the European Regional Development Fund as well as the Agriculture, and Fisheries SFs are also noted as sources in the NAPs. In addition to the SFs, financing is also available through Community Initia- tives such as EQUAL, which countries are already accessing to support equal opportunities initiatives in employment. Scope of this Report This report looks at safety nets in the new member states from the perspective of fiscal impact and effectiveness. Given the limited budgets for social assistance, ensuring that resources are spent well, reach those who need them, and optimize welfare gains and labor market out- comes. This also requires attention to the balance between spending on means-tested anti- poverty programs and family benefits and social insurance programs. Because of high levels of pension spending and rapidly aging populations, there is a risk that contributory social insurance spending will crowd out spending on social assistance. Managing the long-term fiscal sustainability of these programs to meet emerging needs is a priority. Despite the importance of social assistance, there is little comparative analysis of programs across countries.2 This is needed to assess effectiveness at the country and regional levels, as well as for capturing experience for subsequent waves of EU accession countries and for shar- ing with the EU15. The EU's Lisbon process for Social Inclusion is based on policy coordina- tion and experience-sharing. However, as the EU noted in a recent review of the National Action Plans for Social Inclusion in the new member states, comparable information on cov- erage of social assistance and adequacy of benefits are not available (EC 2005). This paper looks at social assistance within the broader context of social protection (Box 1). The main focus of the policy discussion is on the guaranteed minimum income programs, which comprise the centerpiece of the social assistance systems in each of the EU8 countries. Child allowances are also discussed, which are the most important programs from the perspective of spending levels. This first Chapter reviews the datasets and approach. Chapter 2 looks at the context of poverty and social exclusion in the EU8 and the implications for safety nets. Chapter 3 analyses social protection spending. Chapter 4 reviews the performance of social assistance and social protection based on household survey data for six of the countries. Chapter 5 discusses policy design issues including administration, decentralization and activation measures. Chapter 6 looks at the labor market incentives of combined taxes and benefits, and discusses in-work benefits. Chapter 7 concludes. Data and Methodology Data on social protection, and particularly social assistance programs are notoriously difficult to collect and analyze across countries because of the myriad of programs and the lack of 2. A recent review of the National Action Plans for the 10 new EU Member States stated: "Given the key role of social protection systems, one would have expected the NAPs/inclusion to present a much more developed account of how existing and envisaged policies in this area are impacting or are expected to impact on poverty. However, evidence in this area is scarce." 4 World Bank Working Paper Box 1: Definitions This report relies on a number of key terms for describing the set of policies and programs, these are as follows: Social assistance refers to income-tested cash benefits targeted to poor households. Child protection refers to cash benefits and services for children, including child allowances and social services intended for children-at-risk, such as disabled children and children without parental care. Social safety nets and social welfare are used to describe the full range of social assistance and child protection cash and in-kind benefits and services intended for low income individuals and house- holds (e.g. the sum of the previous two categories). Social insurance refers to contribution-based benefits, including old age, disability, survivors' pen- sions and unemployment insurance. Social protection refers to all social insurance and social assistance programs and benefits as well as active labor market policies.3 comparability of systems across countries. This report draws from a collection of data sources to describe the systems and assess their impact on welfare, including: (i) household budget surveys; (ii) national administrative data; and (iii) cross-country databases on social protection. Household budget surveys are used for benefit incidence analysis. Household data collection for living standards analysis is in transition in the new member states as the countries are in the process of adopting Eurostat's comparable Survey on Income and Living Conditions (SILC) methodology which is common across Europe. Analysis of the first round of SILC surveys was underway at the time of writing. For the purpose of this report, national level household budget surveys implemented by the country statistical offices were used as the basis for analysis of welfare impact. Sample sizes and questionnaires of the surveys vary across countries (Table 1). How- ever, efforts were made to ensure comparability. A comparable consumption aggregate was constructed as the welfare measure.4 Country coverage was limited by data quality and availability. Surveys were analyzed for six of the eight countries: Estonia, Hungary, Latvia, Lithuania, Slovenia and Poland. The Czech Republic and Slovakia were not included in the analysis because of omissions in the sample frame which limited their usefulness as instru- ments for analysis of poverty and social assistance.5 For each country attempts were made to analyze two surveys to look at developments over time--especially given the significant reforms to social protection programs over the period. However, because of changes to the household surveys in some countries, the years selected for each country differ. Because of differences in social protection systems and programs, as well as in household survey questionnaires, the cross-country analyses of programs are not strictly comparable. 3. For more information on social protection refer to www.worldbank.org/sp 4. The methodology is described in Appendix B. For information on the measure refer to World Bank, 2005a, Appendix, p. 220. 5. Among other issues, the Czech survey leaves out households with heads who are unemployed. The Slovak survey also leaves out the unemployed (European Commission 2004). Social Assistance in the New EU Member States 5 Table 2. Summary of Household Budget Surveys Country Years Analyzed Sample Sizes Estonia 2003, 2004 3317, 3165 Hungary 2000, 2004 10191 Latvia 2002, 2004 3949, 3913 Lithuania 2000, 2004 7111, 7969 Poland 2000, 2004 36163, 32214 Slovenia 2003 3794 For the purpose of analysis programs itemized in the income sections of the HBS ques- tionnaires were aggregated into five main groups: Pensions comprise contributory pensions including old age, disability and sur- vivors' pensions. Unemployment benefits refers to unemployment insurance benefits, severance pay and unemployment assistance for the long-term unemployed. Family benefits includes family and child benefits, birth grants, and maternity benefits. Targetedsocialassistanceincludesincome-testedcashbenefitsforpoorhouseholds-- mainly the guaranteed minimum income programs and other local cash transfer programs. The collection of programs under each category differs by country. Details on what is included by country are presented in Appendix Table A.8. Relatively small sample sizes limit the extent to which the smaller social protection programs can be analyzed in depth. For example the number of individual beneficiaries of targeted social assistance captured in the surveys is quite small.6 This limits the extent to which data can be disaggregated and benefit incidence can be analyzed by population group (e.g. by gender, region or house- hold size). Such detailed analysis is also constrained by the HBS questionnaires themselves which do not include questions on characteristics which may be correlated with exclusion such as disability and ethnicity.7 Administrative data on social protection programs were collected by a team of local experts in each of the countries from Ministries of Finance, Social Affairs and country sta- tistical offices. This included data on expenditures and beneficiaries. The local teams also collected institutional information on the structure of the systems and the regulations. Efforts were made to ensure comparability of information across countries, but there are limitations because of differences in national programs and definitions. 6. In the 2004 surveys there were 68 households receiving social assistance in Estonia, 193 in Lithuania, and 731 in Poland. 7. For example, Roma ethnicity is not captured in the survey. For a more detailed discussion of the methodological constraints see Ringold, Orenstein, and Wilkens (2005). 6 World Bank Working Paper Box 2: Differences between the ESSPROS and National Accounts Despite a broad harmonization of the ESSPROS and the national accounts (SNA 1993 or ESA 95) there are some differences between them. In the scope of social protection, the national accounts definition of social benefits includes the func- tion Education, while the ESSPROS does not. The ESSPROS definition of social benefits covers both current and capital transfers, while the national accounts definition refers to current transfers only. Moreover, the ESSPROS records certain reductions on taxes and other obligatory levies payable by households.8 The second difference refers to statistical units. The SNA deals with social insurance schemes with- out turning them into units of classification, while ESSPROS social protection transactions are described in schemes. Such schemes can be run by one institutional unit or a group of units and they include non-contributory social protection schemes as well as SNA-defined social insurance schemes. Both systems differ in the way they distinguish social benefits in cash and in kind. In the national accounts social benefits in kind refers only to benefits provided by government units or non-profit organizations serving households, while in ESSPROS social benefits in kind may be granted by any unit. As far as the way transactions are recorded is concerned, ESSPROS does not split contributions by covered persons into payments for insurance services and net contributions available to finance benefits. In contrast, NAs makes it split when autonomous private funded schemes are involved. Moreover, ESSPROS does not impute property income and contribution supplements in respect of private funded schemes. Source: ESSPROS Manual. Cross-country European databases including the European System of Integrated Social Protection Statistics (ESSPROS) database on social expenditures and the Mutual Infor- mation System on Social Protection (MISSOC) are drawn upon in the report. The new member states benefit from being part of these cross-country databases which aim to har- monize data collection across the EU-25. ESSPROS includes detailed data on social protection expenditures. It is important to note that the ESPROSS methodology differs from national accounts data collected at the country level. In this report the ESSPROS data is used for aggregate analysis of social pro- tection programs, while national data is used for analysis of individual program types (e.g. guaranteed minimum income programs). Sources are indicated in the text. The MISSOC database includes detailed descriptive information on the characteris- tics of social protection programs in all EU-25 countries, compiled by national focal points and coordinated by DG Employment, Social Affairs, and Equal Opportunities. 8. They must meet all the following conditions: they answer to general definition of social protection: are granted as flat rate allowances, and the benefits are paid in cash where the taxable income of eligible households is too low to benefit from a reduction. CHAPTER 2 The Context: Poverty and Social Exclusion in the EU8 Poverty and Social Exclusion Social protection systems in the new member states need to meet the challenge of responding to changing and diverse needs within the countries. Socioeconomic conditions in all of the EU8 countries transformed rapidly across the transition period and continue to be dynamic. From a social protection perspective, rapidly ageing populations will influence the size of the sector over the longer term. Based on projections, between 20 and 25 percent of the populations of the EU8 countries will be 65 and over by 2025. Looking across the 25 countries of the Europe and Central Asia Region, the EU8 countries are among the top twelve countries in the region with the oldest and most rapidly aging populations (Figure 1). Absolute poverty is defined as the share of the population below a threshold needed to meet basic subsistence.9 For international comparisons of absolute poverty, the 2005 regional World Bank Study, Growth, Poverty and Inequality, used a common poverty line of $4.30 PPP per day for cross-country comparisons. At this level, the poverty rate (head- count) ranged from 22 percent in Lithuania in 2004 to 11 percent in Hungary (Table 3). At the national level, national poverty lines provide a more meaningful point of analysis based on specific country circumstances. National poverty lines can be developed based on different methodologies, including consumption baskets or thresholds used for benefit eligibility. The World Bank's 2004 study of living standards in Poland used a national poverty line equivalent to the social assistance threshold to define "moderate poverty" and a subsistence basket to define "deep 9. For more on poverty measurement refer to Ravallion (1994). 7 8 World Bank Working Paper Figure 1. Aging Populations in Europe and Central Asia (percent of population 65 and over) 25 20 15 10 5 0 av us M H y ia keyruT & akiav aine & ech Russia Latvia eniav Cz ajikistanT erbaijan Albania Moldo Armenia Belar olandP Georgia Croatia Romania Slo Estonia Ukr Hungar Bulgar Slo Uzbekistan kmenistan yrgyzstanK Lithuania Az urT Kazakhstan Mecedonia Serbia Bosnia FYR 2000 2025 Source: World Bank projections. World Bank, 2006c Forthcoming. poverty." At the moderate line poverty was 14.8 percent and at the deep line, 8.9 percent. The 2006 World Bank Latvia living standards study estimated the poverty headcount at 5.9 percent, based upon an absolute poverty line (World Bank, 2006b). The EU measures poverty using a relative measure of 60 percent of the national median equivalised income. Using this measure, poverty ranged from 8 percent in the Table 3. Poverty Rates in Comparison Laeken at-risk of International poverty National poverty poverty rate (2003)1 line, $4.30 PPP rate3 Czech Republic 8 -- -- Estonia 18 18 -- Latvia 16 14 5.9 Lithuania 15 22 -- Hungary 12 11 -- Poland 17 21 14.8 Slovakia 13.32 -- -- Slovenia 10 -- -- Sources: Laeken indicator from Eurostat; $4.30PPP based on HBS with OECD PPP rates. Notes: (1) At-risk of poverty rate is 60% of national median equivalised income. (2) Preliminary data from EU-SILC. (3) Latvia data from World Bank (2006b); Poland from World Bank (2004). Social Assistance in the New EU Member States 9 Czech Republic to 18 percent in Estonia in 2003. In addition to the choice of poverty line, an important distinction between the EU and World Bank approaches is that the World Bank approach relies on consumption as the main welfare indicator, while the EU mea- sures poverty based upon income. The rationale for using consumption, rather than income, is the significant level of informal sector activity in the new member states and the difficulty of measuring income accurately. Absolute poverty measures may differ consid- erably from relative measures, as relative poverty rates measure individual welfare in rela- tion to the rest of the population, instead of to a fixed benchmark. In some ways the profile of poverty in the EU8 countries has converged toward com- mon characteristics. Poverty is closely correlated with labor market status, educational attainment, and household size (World Bank 2005a). Other risk groups identified in World Bank poverty studies and in the national action plans include people in lagging regions (poorly served by basic services and poorly integrated into markets); linguistic and ethnic minorities including Roma, but also Russian speakers in Latvia and Estonia; the disabled; homeless; and those with serious health risks. Long-term unemployment is a particularly serious concern. According to a recent World Bank employment study, 40­60 percent of the unemployed across the EU8 have been out of work for more than one year (World Bank 2006a). Long-term unemployment puts people at risk of further spiraling into exclusion and poverty. People out of work for long spells face increasingly steep barriers to reentering employment. This is a result of declining productivity, as skills and morale erode over time. The long-term unemployed are also likely to have declining contacts with others through social networks, further low- ering their employment opportunities. Long-term unemployment also carries a stigma which discourages employers, as they may perceive the long-term unemployed as low in productivity and risky to hire. Another critical issue is the significant number of working poor in most EU8 coun- tries. The regional poverty study mentioned above found that--with the exception of Hungary--there are more working adults among the poor than non-working adults (World Bank 2005a). This is an issue of concern across Europe. The EU recently estimated that 7 percent of the employed population in the EU-25, or an estimated 14 million people, live in a household with equivalized income below the national poverty line (Bardone and Guio 2005).10 The same study found that the in-work poverty rate for the new member states was 9 percent on average.11 All of the countries also have unique poverty and exclusion challenges which have implications for policy design. Hungary, the Czech Republic, and Slovakia, for example, have sizeable Roma populations, which face higher risks of poverty and exclusion than other groups (Table 4). High Roma poverty is multidimensional, linked to high unemployment-- formal unemployment can range up to 90 percent in some Roma settlements, low educa- tion levels and poor living conditions and health status (Ringold, Orenstein, and Wilkens 2005). Regional poverty is another area where countries face specific hurdles. For example, in Poland, the largest of the EU8 countries, there is increasing divergence in regional 10. Based on the EU's relative poverty line of 60 percent of equivalized median income. 11. In-work poverty rate includes Malta and Cyprus. 10 World Bank Working Paper Table 4. Key Labor Market Indicators in the EU8 Countries Employment rate, Unemployment Long-term 15­64 years rate unemployment rate 2000 2004 2005 2000 2004 2005 2000 2004 2005 EU15 63.4 64.7 65.1 7.7 8.1 7.9 3.4 3.4 3.3 Czech Republic 65.0 64.2 64.8 8.7 8.3 7.9 4.2 4.2 4.2 Estonia 60.4 63.1 64.5 12.8 9.7 7.9 5.9 5.0 4.2 Latvia 57.5 62.4 63.4 13.7 10.4 8.9 7.9 4.6 4.2 Lithuania 59.1 61.2 62.7 16.4 11.4 8.3 8.0 5.8 4.3 Hungary 56.3 56.9 57.0 6.4 6.1 7.2 3.1 2.7 3.2 Poland 55.0 51.7 52.8 16.1 19.0 17.7 7.4 10.3 10.2 Slovenia 62.8 65.3 65.9 6.7 6.3 6.5 4.1 3.2 3.0 Slovakia 56.8 57.0 57.7 18.8 18.2 16.3 10.3 11.8 11.8 Source: Eurostat. development, with poverty reduction in rural areas and secondary cities, particularly in the North, South-West, and South-East falling behind that of Warsaw (World Bank 2004). Regional disparities in unemployment in Slovakia are the second highest in the EU after Italy. Even in the smaller Baltic countries, regional differences are prevalent. The recent World Bank living standards study for Latvia found significantly higher poverty in the regions of Vidzeme and Latgale, than in Riga (World Bank 2006b). Implications for Safety Nets The context of welfare developments in the EU8 has significant implications for the types of social assistance programs and policies which are required. In the first place, the dynamic environment calls for programs which can be flexible and responsive to changing conditions. Labor market conditions continue to fluctuate in the EU8 countries. While EU integration Table 5. Estimated Roma Populations in the EU8 Official data (census) Unofficial Estimates Czech Republic 12,000 160,000­275,000 Estonia na na Hungary 190,000 500,000­800,000 Latvia 8,200 na Lithuania 2,000 na Poland 12,700 15,000­50,000 Slovakia 89,920 350,000­500,000 Slovenia 3,246 7,000­10,000 Sources: Census Data, NAPs, ERRC, World Bank. Social Assistance in the New EU Member States 11 has been a source of stability, the EU8 countries remain at risk of shocks which can affect poverty and welfare. For example, the recent oil price shocks are expected to have an impact on household welfare, particularly in the winter heating season. The close links between poverty and the labor market, and particularly the serious issues of long-term unemployment and working poverty have implications for policy design. Social assistance benefits are of critical importance for the long-term unemployed, as unemployment benefits in all countries have time limitations. Once insurance benefits are exhausted, beneficiaries become eligible to apply for social assistance. Many countries also have a range of active labor market measures targeted at the long-term unemployed, some of which may be connected with social assistance (for exmaple, public works programs). However, as evaluations have found that some active labor market measures are not effective, these programs need to be closely monitored to ensure their effectiveness. High numbers of working poor call for attention to benefits which help to "make work pay" and provide support to individuals and households even if they have a job. The EU8 countries have adopted a wide range of options to support the working poor including, tax credits for working household s (Estonia), social assistance benefits which taper off with wages (Slovakia, Hungary) and child care benefits and other support to families. Again, this is an area where monitoring and evaluation of early lessons and experience will be essential, especially to ensure that they are cost-effective. The significant level of working poverty indicates that net wages may not be high enough to keep people out of poverty. This in turn highlights the risk of poverty traps. If social assistance benefits are close to wages, some people may opt to stay out of the labor market to remain eligible for benefits, leading to serious issues of welfare dependency. Slovakia most recently tackled the issue of poverty traps in its 2004 reform (Box 3). Increasing regional disparities underscores the need for programs which can respond to local needs. The EU8 countries have adopted significant reforms of their intergovernmental fiscal frameworks and public administration systems to increase the roles of local governments in social services--including financing and delivering cash and in-kind social assistance benefits. Decentralization poses clear trade-offs in this area. While local involvement has the potential to improve targeting of benefits--since local officials are better able to identify the poor, local governments may lack resources for financing social assistance, or the incentives to allocate resources to these programs which tend to be less politically attractive than education, health and other competing priorities. Transfer arrangements that include earmarked resources for social assistance (including for local co-financing of central government allocations) may be a useful approach. Similarly, design of social assistance needs to ensure that programs meet the particular needs of vulnerable groups. For example, qualitative fieldwork conducted in 2002 in Slovakia found that social workers rarely visited Roma communities (Ringold et al., 2005). More recently, the Slovak government has made use of EU funds to increase the number of social workers working in local communities and encourage greater outreach. Other groups will have specific needs. Linguistic groups require language support to facilitate their access to services, while the disabled will need other types of support depending on their condition. CHAPTER 3 Social Protection in the New Member States Social Protection Expenditures and Financing To varying extents, social assistance reforms in the EU8 have aimed to address the issues outlined above. Countries have restructured their systems to meet changing needs and to adjust to structural changes, including decentralization and the availability of EU structural funds for certain types of social programs. The following section looks at social assistance within the context of the overall social protection system. Total social protection spending, including social assistance and social insurance, comprises a significant share of government resources across countries. Based on spending levels, countries clustered into three groups in 2003.12 The Baltic States (Estonia, Latvia, and Lithuania) spent about 13 percent of GDP, the Visegrad countries (Hungary, Czech Republic, and Slovakia) between 20 and 21 percent, and Slovenia over 24 percent of GDP. This pattern is consistent with that of overall public expenditure in the region (Laursen 2005). In all cases, spending on pensions makes up the largest share of the budget, ranging from 42 percent of total social protection in Slovakia, to 68 percent in Poland (Appendix Table A.2). Total social protection spending in the EU8 countries remains below the EU15, which stood at 27 percent of GDP in 2003. Between 2001 and 2003, social protection spending contracted alongside total government expenditures in the Baltic States, Slovakia, and Slovenia, and expanded in the Visegrad countries, particularly Hungary. Composition of Social Protection Spending Pensions. Social insurance, especially pensions, are contribution based benefits intended to smooth consumption over the lifecycle, provide income security during old 12. The most recent year available in ESPROSS. 13 14 World Bank Working Paper Figure 2. Total Social Protection Expenditures 2001 and 2003 (percent of GDP) 30.0 25.0 20.0 15.0 10.0 5.0 0.0 EE LV LT SK CZ HU PL SI EU15 Non-Pensions Pensions Source: ESPROSS Database, authors' calculations. age, and protect beneficiaries from poverty. Pensions, including old age, disability and survivors' pensions, make up the largest share of social protection spending, ranging from 6 (Estonia) to 14 (Poland) percent of GDP (Appendix Table A.3). Spending on pensions has remained relatively constant, despite significant reforms in all countries to improve the long-term sustainability and efficiency of the PAYG systems. Latvia is the only country where pension expenditures have fallen notably--from 9.5 to 7.5 percent of GDP between 2001 and 2003. High pension spending is maintained by low effective retirement ages (Figure 4) driven by early retirements and generous benefit levels.13 Sickness Benefits. Following pensions, health and sickness benefits comprise the second largest component of social protection spending. The ESPROSS accounting methodology includes health care as a component of social protection spending, which is in contrast with national accounts practices. Health care is not analyzed here.14 Sickness benefits are short- term insurance-based benefits paid to employees in the event of illness.15 The category can 13. Further information on recent pension reforms in the EU8 can be found in World Bank (2005c) and Schwarz (2005). 14. Analysis of health financing in the EU8 is included in Chapter IV of Laursen (2005). 15. Long-term disability pensions are included with pensions. Social Assistance in the New EU Member States 15 Figure 3. Composition of Social Protection Spending, 2003 (percent of total) 100% 80% 60% 40% 20% 0% EU15 CZ EE LV LT HU PL SI SK Administrative/other Social exclusion n.e.c. Family/Children Unemployment Sickness/Health Pensions Note: "Administrative/other" includes housing expenditures. Source: ESPROSS. include other short-term benefits such as maternity leave, or care for ill relatives. Costs of sickness benefits are shared between the employer and the health or social insurance fund. These benefits are likely to be an increasing source of fiscal pressure in the EU8 as formal employment increases. Sickness benefits have been escalating in OECD countries. In the Figure 4. Pension Spending and Average Exit Age from the Labor Force in the EU 15 AT PL DE GR GDP FR NL EU15 SE of 12 EU25 % BE PT SI FI LU DK UK 2003, in MT ES HU 9 CZ SK LV LT CY xpendituree EE 6 ensionP IE 3 55 56 57 58 59 60 61 62 63 64 Average exit age from the labor market in 2003, years Source: Eurostat based on ESPROSS. 16 World Bank Working Paper EU15, spending on sickness and health benefits climbed from 7.1 to 7.7 percent of GDP between 2001 and 2003. In the EU8, expenditures have been increasing in all countries except Estonia, Lithuania and Slovakia (Appendix Table A.3). In 2003 spending ranged from 7.8 percent of GDP in Slovenia to 3 percent of GDP in Latvia. Eligibility for sickness benefits can range from 4 months to over a year. Unemployment Benefits. With the escalation of open unemployment through the 1990s, all countries put in place unemployment insurance benefits financed through pay- roll tax contributions. Other unemployment related benefits include unemployment assis- tance, which are designated non-contributory benefits for the long-term unemployed distinct from social assistance, and active labor market programs. In 2003, spending on unemployment benefits ranged from 1 percent of GDP in Slovakia to 0.2 percent of GDP in Estonia and Latvia. Social Assistance. All countries have guaranteed minimum income programs (GMI) as the core element of their social safety net. GMI programs provide means-tested income support to poor households who fall below a set threshold. These benefits are discussed further below. Benefits can either be monthly, or paid on an emergency, one-off, basis. Other benefits included in the social assistance category include targeted scholarships and school meal programs. On balance, social protection expenditures in the EU8 countries are largely focused on cash transfers (Figure 5). The share of cash benefits ranges from 64 percent of total social protection in Hungary, to 83 percent in Poland. In-kind benefits accounted for in ESPROSS include health services (the largest share), as well as hous- ing, active labor market programs and social services. The high share of transfers in Poland reflects the high social insurance spending, particularly on old age and disability benefits. Figure 5. Share of Cash and In-Kind Benefits 100 90 80 70 60 50 40 30 20 10 0 HU LT SK CZ SI EU15 EE LV PL In-kind benefits Cash benefits Social Assistance in the New EU Member States 17 Figure 6. Spending on Guaranteed Minimum Income Programs (percent of GDP) 1.20 1.00 0.80 0.60 0.40 0.20 - 1999 2000 2001 2002 2003 2004 Czech Republic Estonia Latvia Lithuania Poland Slovakia Slovenia Sources: National budget data; and authors' calculations. Guaranteed Minimum Income (GMI) Programs Across the EU8, guaranteed minimum income (GMI) programs have emerged to form the core social assistance program of last resort for the poorest households. These programs were introduced during the 1990s in response to growing poverty. Minimum income ben- efits are targeted means-tested benefits which are paid monthly, or on a one-off emergency basis, to poor households. While precise benefit formulas vary across countries, the amount of the benefit is generally the difference between household income and the ben- efit threshold. The design features of GMI programs in the EU8 countries are summarized in Appendix Table A.10. GMI programs cover between 1 and 4 percent of the total population and comprise 0.1 to 0.5 percent of GDP (Figure 6).16 Programs are the smallest, in terms of spending, in the Baltic states and in Poland. While spending on the GMI programs has remained relatively stable in most countries, the decline in spending in Slovakia, from 1.13 percent of GDP in 1999 to 0.48 percent of GDP in 2004 is striking. This reflects the significant systemic reforms that Slovakia implemented starting in 2002 (Box 3). Despite the con- traction, coverage of the program in Slovakia remains higher than in the other countries (Figure 7).17 16. Coverage rates presented here are based on administrative data and vary from the survey-based data on coverage presented later in the report. 17. Note that differences in coverage between Figure 7 and Figure 11 on coverage of social protection based on the household survey data differ based on the differences in the survey instruments and cover- age of programs. Figure # includes all social assistance programs, while Figure # is restricted to the GMI program. 18 World Bank Working Paper Box 3: Slovakia's New Social Policy In 2002, as part of its overall structural reform agenda, Slovakia introduced one of the most aggres- sive set of reforms to social benefits in the region. The aim of the reforms was to modernize and consolidate benefit programs to eliminate the poverty and employment traps which had plagued the system. The value of the benefits received by some workers exceeded the wage they could have received from employment. The high tax wedge on wages was also associated with low job cre- ation, particularly for lower wage workers in low productivity/employment regions. In response, the government launched a "New Social Policy" with the three main aims of: (i) making work pay, by increasing work incentives; (ii) streamlining social benefits administration and improving targeting through refined eligibility criteria, and (iii) strengthening assistance to those in need through the guaranteed minimum income program supplemented by active labor market programs and outreach by social workers. Specific reforms to the social benefits included: ­ Introduction of a basic social benefit for the poor, with supplementary allowances for health care, housing, activation and poverty relief, based on a consistent means-testing methodology. ­ Lowered and simplified calculation of the base benefit, and increases in benefits for families with children. ­ New ALMPs, including training, services for families with children, the activation allowance, and allowances for disabled workers. These programs all require regular registration at the Labor Offices. ­ Merging of social assistance and labor offices in to Offices of Labor, Social Affairs and Family. These measures were part of a broader set of reforms including the introduction of a flat tax. Early assessment suggests that while the reforms have potentially increased the disposable income slightly for most low-income households, they appear to have had a negative effect on the income level of poor households with unemployed adults and large numbers of children. On the employ- ment side, a 2005 study by the Centre for Research on Work and Family found that 90 percent of previous beneficiaries had left the social assistance system following the reforms and that 74 per- cent left because they had entered the labor force. Further analysis on outcomes is important to assess the impact on welfare and employment outcomes. Source: World Bank (2005b), Tomova (2006). Figure 7. Coverage of Guaranteed Minimum Income Programs (percent of population, latest year) 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Hungary Poland Lithuania Estonia Slovenia Latvia Slovakia Czech Republic Notes: Data are for 2004 except Poland and Hungary which refer to 2003. Sources: National budget data; and authors' calculations. Social Assistance in the New EU Member States 19 Figure 8. Changes in Total Fertility Rates in Europe and Central Asia, 1950 and 2000 7 6 Rate 5 4 tility erF 3 2 otalT1 0 ia y us M av H Rep Latvia eniav aine & akiav & FYR Estonia keyruT Bulgar Belar Croatia Russia olandP ech Georgia erbaijan Albania Lithuania Hungar Slo Ukr Romania Moldo Slo Armenia Cz Serbia Kazakhstan yrgyzstanK ajikistanT kmenistan Bosnia Az Uzbekistan urT 1950 2000 Source: World Bank projections. World Bank, 2006c Forthcoming. Family Benefits Family benefits are an important element of social safety nets in the EU8. Family benefits have wider coverage than social assistance programs and are well targeted because of the close connection between poverty and household size across countries. Family benefit sys- tems generally have two main objectives. First, because of declining fertility rates in the region, family benefits have sought to encourage families to have children (Figure 8). This is despite inconclusive research on the impact of benefits on birth rates.18 Second, family benefits are viewed as an efficient form of income support because of the close link between welfare and household size. Family benefits are a complex mix of programs, including inherited programs from the socialist period as well as new benefits and services. Like the EU15, each of the coun- tries has a monthly child benefit paid per child which comprises a large share of total spend- ing on family benefits. Benefits are means-tested in the Czech Republic, Slovenia and Poland, and not in the other countries. In 2003 total family benefits--including child allowances and other benefits--ranged from 2.7 percent of GDP in Hungary, which has the most extensive system, to 1 percent in Lithuania. Family benefits include a wide range of targeted and universal benefits. Hungary has the most complex system, with a plethora of fragmented benefits and programs, many of which have no connection with each other. A detailed description of the Hungarian sys- tem is included in Appendix A. The main program in all countries is the monthly child allowance. With the exception of Slovenia, Poland and the Czech Republic, child 18. For example, see Moffitt, 1998 for a summary of US research. 20 World Bank Working Paper Figure 9. Child Allowances (percent of GDP), 2004* 1.0 0.9 0.8 0.7 0.6 GDP 0.5 of %0.4 0.3 0.2 0.1 - Lithuania Hungary Poland Czech Latvia Slovakia Estonia Slovenia Republic Note: *Data for Hungary and Slovenia refer to 2004. Source: National budget data, authors' calculations. allowances are universal for all children. Spending ranges from 0.9 percent of GDP in Slovenia, to 0.2 in Lithuania and Hungary. Monthly benefit levels vary. While some coun- tries have flat benefits per child (Slovakia), other countries set monthly amounts based on income and number of children. Amounts range from the equivalent of 9 Euros per month, the lowest possible benefit in the Czech Republic, to 108, the upper range in Slovenia (Appendix Table A.9). CHAPTER 4 Performance and Effectiveness T his chapter reviews the performance of social protection programs using the household survey data. The first section looks at the full impact of social protec- tion programs, while the second section discusses particulars related to the effectiveness of social assistance programs. It focuses on three key indicators. Coveragelooksatthereachofthesystemandisdefinedastheshareofhouseholds receiving a particular type of benefit. Targetingassesseswhetherbenefitsarereachingthosewhoneedthemandrefersto the share of benefits received by each consumption quintile. Targeting also mea- sures the amount of leakage to the non-poor. Finally, adequacy looks at the generosity of the system and is defined as the ratio of the average benefit amount for recipients to average consumption of that group. Results of this analysis are summarized in Appendix Table A.7. The results focus on the six countries for which HBS data were available: Estonia, the Czech Republic, Latvia, Lithuania, Hungary, Poland and Slovenia. Coverage: How Many People Receive Social Protection? Coverage of social protection is high in the EU8, with the majority of households receiving at least one benefit. In 2004 coverage ranged from 90 percent in Hungary to 62 percent in Lithuania (Figure 10). Coverage of social protection is progressive. The poor (defined as the bottom quintile of the population) receive social protection at a higher rate, reaching nearly 100 percent of the population in Hungary in 2004. 21 22 World Bank Working Paper Figure 10. Coverage of Social Protection Benefits, 2004 (percent of individuals in households receiving a benefit) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Estonia Hungary Poland Latvia Lithuania Slovenia Total Bottom Quintile Top Quintile Source: World Bank Household Survey Database. Coverage is driven largely by pensions, which are received by 30 to 45 percent of the population across countries. In contrast, benefits targeted at the poor are much less preva- lent (Figure 11). Family benefits have a wide reach covering around half of the population in Estonia, Hungary and Latvia. Although family benefits are only income-related in Figure 11. Coverage of Social Assistance and Family Benefits, 2004 (percent receiving benefits) 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Estonia Poland Lithuania Hungary Latvia Slovenia % receiving social assistance % of bottom quintile receiving social assistance % receiving family benefits % of bottom quintile receiving family benefits Source: World Bank Household Survey Database; authors' calculations. Social Assistance in the New EU Member States 23 Poland, the Czech Republic and Slovenia, they are progressive across countries, with much greater coverage among the bottom 20 percent of the population. Hungary stands out as having the most extensive and progressive set of family benefits programs, with 86 percent of the bottom quintile receiving benefits in 2004. On the other hand, social assistance programs targeted at low-income households are very small in coverage across countries. Despite the differences in programs and the survey data, the share of the population receiving means-tested social assistance is similar across countries-- ranging from 4 to 9 percent of the bottom quintile, or 2 to 5 percent of the total population. Targeting: Does Social Assistance Reach the Poor? Overall, means-tested and targeted benefits are more common in the EU15 than in the EU8--with the exception of Slovenia (Figure 12). This is not surprising, given that these are new programs and countries have had to build the institutional capacity for delivering income-and means-tested benefits from scratch. Targeted benefits encompass social assis- tance transfers and other programs directed at low income households and individuals. Family benefits in the Czech Republic, Poland and Slovenia are income-tested, while in the other countries all residents are eligible. The picture on targeting of social protection as a whole across the countries is mixed. Not surprisingly, contributory insurance-based benefits accrue more to the upper income deciles, while non-contributory benefits, such as family benefits and social assistance have higher incidence among the poor, except Poland (Table 6). Because of the close link between family size and poverty, child allowances and other family benefits are more likely to reach poor households. Social insurance benefits in the Baltic countries show a more equal distribution across quintiles, while in Hungary and Poland, social insurance is more regressive. Non-contributory benefits are more progressive, particularly in Hungary and Latvia. These benefits stand-out as regressive in Poland--where a greater share of non-contributory benefits accrue to the upper decile than to the bottom, indicating significant inefficiencies in one of the most costly systems in the region. Figure 12. Share of Means-Tested Benefits, 2003 100.0 98.0 96.0 94.0 92.0 90.0 88.0 86.0 84.0 EU15 CZ EE LV LT HU PL SI SK Means Tested Non-Means Tested Source: ESPROSS. Table 6. Distribution of Benefits by Quintile (contributory and non-contributory benefits), 2004 Poland Lithuania Latvia Estonia Hungary Slovenia C* NC** C* NC** C* NC** C* NC** C* NC** C* NC** 1 (poor) 7 20 17 23 17 30 19 24 14 31 13 32 2 12 18 21 22 23 20 21 24 19 22 16 24 24 3 18 18 21 16 23 17 21 20 22 17 19 17 4 26 19 23 18 20 18 20 19 22 17 24 15 5 (rich) 36 25 19 21 17 16 19 13 24 14 28 13 Total 100 100 100 100 100 100 100 100 100 100 100 100 * Contributory; ** Noncontributory. Source: Household Budget Surveys. Notes: Contributory includes insurance (pensions, unemployment); non-contributory are non-insurance based benefits. Quintiles are based on post-transfer con- sumption. Social Assistance in the New EU Member States 25 Figure 13. Targeting of Family Benefits and Social Assistance, 2004 (percent of benefits going to the bottom and top quintiles) 60% 50% 40% 30% 20% 10% 0% Q1 Q5 Q1 Q5 Family Benefits Social Assistance Estonia Hungary Poland Latvia Lithuania Slovenia Source: ECA Household Survey Database. Authors' calculations. As a whole, the household survey data show that social assistance and family benefits appear to be well targeted in the new member states, although there is variation across the countries, and room for improving targeting to the poorest households. In Lithuania and Hungary, 42 to 56 percent of social assistance benefits went to the bottom 20 percent of the population in 2004 (Figure 13). In contrast with the other countries, in Latvia and Lithuania there is a higher level of leakage to the top quintile--24 and 15 percent in 2004 respectively. In both countries, this leakage represents a noticeable deterioration in targeting performance since 2002, meriting fur- ther analysis. Leakage is also high in Slovenia, with 20 of resources going to the top quintile. The relatively weak targeting of social assistance in Latvia is driven by regional factors. Latvia has one of the most decentralized social assistance systems in the EU8. While reforms introduced in 2004 strengthened the central GMI program, municipalities can also provide additional local benefits. This may explain the variation across municipalities, if poorer municipalities are less able to provide benefits. Data from 2002 indicate weak targeting across consumption deciles and a noticeable deterioration in targeting. While the reforms may have strengthened targeting of the GMI, the local benefits (captured together with the GMI in the household survey) may distort the picture. Targeting of family benefits is also progressive, despite the fact that most benefits are not income tested. Child benefits are targeted in Poland, the Czech Republic and Slovenia. 26 World Bank Working Paper Figure 14. Targeting of Social Assistance in Comparison (CGH Index) 3.5 3 2.5 2 1.5 1 0.5 0 a nia'04 0 0 4 ssia azil '02 Ru Poland'04LatviaSlove '04 nia'03 entin Br Arg nia'03 nia'04 mania nia'00 Esto Poland'0Hungary'0 ngary'0Latvia sto ua Ro ua Hu E Lith Lith Notes: Data for Romania are for 2003. Data include country-specific definitions of social assistance as included in the surveys and are not strictly comparable across countries. Sources: Russia and Romania, from Tesliuc, 2004; ECA household budget survey database; and authors' calculations. This is evident in the data which show that Poland has the highest level of family benefits going to the bottom quintile in 2004 (40 percent), and the lowest leakage to the rich (3 per- cent). In Poland and Latvia, family benefits appear better targeted than social assistance. In all countries, the share of family benefits received by the upper quintile is under 16 per- cent. This may be a result of low take-up through self-targeting, if richer households opt not to apply for the benefit because of stigma or the bureaucratic obstacles (opportunity costs) of applying for benefits. Across countries, family benefits have greater incidence among the middle of the income distribution (quintiles 2­4) than social assistance, con- firming their role as a transfer program for middle income households. Another measure for analysis of targeting of transfer programs is the Coady-Grosh- Hoddinott index (CGH Index), which is used increasingly in international comparisons. The index is a simple measure, based on a comparison of actual targeting performance to a common reference outcome of neutral targeting. In other words, it compares current tar- geting to the level of targeting that would result from equal allocation of resources across quintiles--20 percent to each.19 The higher the CGH, the greater the targeting efficiency. Social assistance in many of the EU8 countries is better targeted than in other countries and regions. Figure 14 shows CGH rates for the EU8 and other countries. Targeting in 19. The CGH indicator is constructed by dividing the actual outcome by the neutral outcome for the bottom quintile of the population. In other words, if the poorest 20 percent of the population received 30 percent of the benefits, then the CGH indicator is equal to (30/20) = 1.5, which means that targeting has led to the poorest quintile receiving 50 percent more than they would have received under neutral tar- geting (see Lindert, Skoufias, and Shapiro 2006; Coady, and others 2005). Social Assistance in the New EU Member States 27 Figure 15. Adequacy of Social Protection Benefits: Bottom Quintile (average benefit as a share of average consumption for each beneficiary group) 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Estonia Hungary Poland Latvia Lithuania Slovenia Pensions UI FB SA Note: UI = Unemployment insurance; FB = family benefits; SA = social assistance. Source: ECA Household Survey Database. Authors' calculations. Lithuania is particularly high relative to other countries. However, as mentioned, there was a decline in targeting performance over time, with lower rates in 2004. This merits further investigation to understand the sources of this decline. Adequacy: Does Social Protection Alleviate Poverty? Social protection is an important source of income for many households, especially those in the bottom quintile of the population. Average social protection benefits amount to between 27 (Latvia) and 65 (Poland) percent of average household consumption for those who receive social protection. For those individuals in the poorest quintile, benefits make up as much as 91 and 92 percent of average consumption in Estonia and Hungary (Figure 15). Although their main objective is consumption smoothing, pensions have the greatest welfare impact. Even among the bottom quintile, pensions make up over 80 percent of the average consumption of those who receive them. However, the majority of the population in the bottom quintile does not receive pensions, and relies on social assistance, which is much less generous. Benefit levels are quite low in Latvia in particular. CHAPTER 5 Design of Social Assistance Programs he differences in performance of social protection programs reflect contrasting systems T and program design. During the transition period and leading up to accession, the EU8 countries introduced important reforms to improve the effectiveness of social assistance programs. A first step for many countries was to consolidate and simplify the complex web of inherited programs. For example, in 2004 Poland consolidated its family benefits into a single benefit. Countries have both introduced new means-tested social assistance benefits and cri- teria to target previously un-targeted benefits such as child allowances. Targeting is espe- cially difficult in Central Europe because of the significant size of the informal sector, which complicates measurement of income for accessing benefit eligibility since beneficiaries may be reluctant to report income if they are not paying taxes or social security contributions. To address this, countries, including Lithuania and Slovenia, have introduced criteria beyond income. Administration of Targeted Benefits The major challenge for social assistance systems is getting targeting right by minimiz- ing both errors of inclusion (benefits go to the non-poor) and errors of exclusion (the poor not getting benefits). As indicated in the survey analysis above, there is substan- tial scope for improving targeting of social assistance benefits to reduce leakage. Coun- tries have already made significant efforts to improve targeting by refining eligibility criteria to include property and financial assets, modernizing information systems, increasing the role of social workers to provide outreach services, and administrative reforms. 29 30 World Bank Working Paper Lithuania is one of the countries which has made significant efforts to improve tar- geting. In 2004, 70 percent of benefits went to the poorest two quintiles, the highest of any of the countries for which data were available. In April 2004, further reforms were intro- duced to improve targeting including adding property as a means-testing criteria along with income, and adding additional qualification criteria. A recent assessment conducted for a forthcoming World Bank study on targeting in the ECA region identified a set of fac- tors which has contributed to better targeting in Lithuania (Zalimiene 2005). These included: Inter-agency sharing of databases for cross-checking eligibility of beneficiaries, including with the State Social Insurance Fund; Public information and outreach about benefit programs and eligibility criteria through written materials and national radio broadcasts; Regulations on social benefits which specify instructions on implementation, including means-testing procedures; and Administrativearrangementsinsocialassistanceoffices,includingdivisionoflabor between professionals responsible for customer service and document processing, additional verification of eligibility by specialized staff, and random auditing by unit managers. However, there is a need for further analysis of the outcomes of the reforms. As noted in the previous section, although the targeting of overall social assistance programs in Lithuania has decreased, Lithuania remains the best performer across the EU8. However, this may not be reflective of the individual GMI program which cannot be isolated in the data. Administrative arrangements for GMI programs vary significantly across countries, implying different levels of administrative costs. For example, in Latvia and Lithuania, means-testing is rechecked every three months, while in Slovenia it is done every six months. In Estonia, beneficiaries are required to reapply monthly. Decentralization Decentralization to lower levels of government has been an ongoing process in the EU8 countries (Dillinger 2005). Decentralization has the potential to improve social service delivery, by bringing services closer to clients and strengthening the capacity of local gov- ernments and providers to plan and deliver services and to monitor outcomes. However, it also entails risks, particularly for social welfare. In the absence of sufficient attention to local resources and capacity, decentralization can lead to an erosion of services. It can also deepen geographic inequalities--if poorer municipalities with the least resources are unable to maintain services. Lessons from other countries in the region underscore the significant risks of decentralizing the financing of social welfare programs without a well-functioning intergovernmental fiscal system that ensures adequate, sustainable and equitable financing of benefits and programs across municipalities. Without earmarked financing sources there is a high risk that funds designated for social assistance programs will be diverted Social Assistance in the New EU Member States 31 elsewhere. There is also a risk that the poorest municipalities with the greatest need will have the most limited resources available for social assistance, leading to a vicious circle in which funds are not available to those who need them the most. Experience from Romania and Bosnia and Herzegovina shows that premature decentralization of financing responsibility for social assistance benefits can undermine their provision, as institutional and financing arrangements are unable to cope with the requirements of administering cash benefits. After delivery of cash transfers collapsed in the late 1990s, Romania recentralized social assistance financing in 2002. Mean- while, in Bosnia and Herzegovina, decentralized financing mechanisms have resulted in substantial inter-regional disparities in coverage, with poorer localities providing the least amount of services. In the face of resource constraints, eligibility criteria for most benefits in Bosnia are ad hoc, as local welfare offices use discretion when rationing available resources. The situation in Latvia illustrates the risks of a decentralized system.20 Latvia's Guaranteed Minimum Income program is the most decentralized in the EU8 countries, as it is both financed and administered by local governments. Although the program is well targeted at the national level, it is less so at the regional level. In 2004, around one-half of program beneficiaries came from the poorest quintile. However, local governments in relatively richer regions, such as Riga, have considerably more funds available to spend on these programs than those in poorer regions. This results in a perverse outcome. More than 40 percent of total social assistance transfers in Latvia go to people living in Riga (Figure 16), even though they are, on average, considerably richer than those residing in other parts of the country. This illustrates the dilemma in local government financing of social assistance benefits. On the one hand, the principle of subsidiary decision-making in the provision of social assistance is desirable. It helps ensure that municipalities plan their budgets and expenditures carefully, and that they do not end up lobbying the central government for large transfers. On the other hand, however, municipalities with the most funds are not necessarily the ones whose residents are most in need of social assistance. Labor Market Incentives A critical issue for the design of GMI programs is the need to avoid welfare dependency and ensure that beneficiaries have incentives to participate in the labor market. Labor market outcomes are a critical concern for social protection systems as a whole. This section discusses specific issues that arise in the design of GMI programs, while the fol- lowing section broadens the discussion to consider labor effects of social protection and taxation. Within benefit programs, labor market incentives can be increased by keeping benefits lower than wages, however, there is a tradeoff with adequacy, and the need to make sure that benefits are also effective sources of poverty relief. Another mechanism is to link social 20. This discussion draws on the forthcoming World Bank report "Latvia: Sharing the High Growth Dividend, A Living Standards Assessment" (June 2006 version). 32 World Bank Working Paper Figure 16. Riga Residents Have a Disproportionately High Likelihood of Receiving Local Government Benefits 100% Pieriga Pieriga 80% Riga Riga 60% Vidzeme Vidzeme 40% Latgale Latgale 20% Zemgale Zemgale Kurzeme Kurzeme 0% Share of Total Population Share of Social Assistance Beneficiaries Source: World Bank estimates based on 2004 HBS data. assistance beneficiaries with labor offices. All of the EU8 countries require that beneficiaries are registered with local labor offices as unemployed. Some countries require that individuals demonstrate a willingness to work by accepting available jobs which the labor offices deem appropriate. In Slovenia, beneficiaries must be actively seeking work or participating in an active labor market program before getting benefits. Social assistance and family benefits in the EU8 are generally low in comparison with minimum wages. The average level of social assistance benefits received by beneficiaries ranged from 15 (Estonia) to 2 (Latvia) percent of the minimum wage, and family benefits from 11 (Hungary) to 4 (Lithuania and Slovenia) percent (Figure 17). These levels suggest that labor market incentives are not a serious concern. However, this basic analysis only provides a partial picture as households may receive more than one benefit, including social assistance, family benefits, short-term benefits, as well as social insurance if a household member is eligible, therefore aggregate benefit levels may create poverty traps. Variations in wage levels across regions may also create poverty traps in some geographic areas. Labor market participation can be encouraged by limiting the duration of benefits. While the EU8 countries have limitations on how long beneficiaries can receive bene- fits before reapplying (ranging from one to six months) none of the countries appear to enforce a fixed cut-off. Data for 2003 indicated that 18 percent of total beneficiaries received benefits every month of the year. There is some evidence in Estonia of house- holds who have received social assistance permanently for over ten years (Trumm 2006). Social Assistance in the New EU Member States 33 Figure 17. Benefits and Minimum Wages, 2004 (average monthly benefit as percent of minimum wage) 18% 16% 14% 12% 10% 8% 6% 4% 2% 0% Estonia Hungary Latvia Lithuania Poland Slovenia Social Assistance Family Benefits Notes: Slovenia refers to 2003. Average benefit levels are average monthly benefits for beneficiaries. Source: Household Budget Surveys; minimum wages from Eurostat. Activation measures aimed at supporting labor market integration have been an increasing focus within social assistance policies. The EU has actively supported such mea- sures through pre-accession and structural fund programs. The European Social Fund, in particular, has provided substantial financial support for activation measures. Active mea- sures can include various types of active labor market programs including public works, training programs, and support for small businesses, as well as community development programs (World Bank 2006a). In the EU15, various activation programs are in place. The Netherlands outsources reintegration activities to the private sector, while the UK makes the unemployed enter into individual labor market reintegration contracts. Belgium has a program focused on reintegration of the elderly. Slovakia introduced a wide range of programs in 2004, including transportation subsidies for attending job interviews, small business support for the unemployed, child care, and training. The focus on activation measures will require greater coordination between social assistance and employment offices. In Slovakia, this was addressed through actual merging of offices. This experiment will be important for other countries. Active social assistance measures are of critical importance both for avoiding dependency traps and for reaching excluded households and communities that may even be missed by social assistance benefits. Slovakia has pioneered an innovative approach through its Social Development Fund (SDF), which was established to support demand driven community development projects in excluded and marginalized communities (Box 4). The SDF provides support for sub-projects which are identified, designed and 34 World Bank Working Paper Box 4: The Slovakia Social Development Fund An early assessment of Slovakia's tax-benefit reforms concluded that even under the best case sce- nario, the reforms would not be sufficient to address severe pockets of unemployment and social exclusion. While absolute poverty at the national level is low in Slovakia, there are significant regional disparities. Slovakia also has the largest share of Roma population of any of the EU8 coun- tries. An estimated one-fourth to one-third of Roma in Slovakia live in settlements, many of which are geographically isolated and cut-off from public utilities, basic social infrastructure and social services (Ringold, et al., 2005). In 2004, Slovakia set up the Social Development Fund (SDF) as an innovative approach to tackling social exclusion through demand-driven, community-based projects. The idea of the Slovakia SDF was to complement the structural reforms of social benefits and taxation with active measures tar- geted at the most excluded communities and to serve as a mechanism for targeting EU Structural Funds to poor communities (World Bank, 2005b). The design of the SDF combined aspects of the social investment approach supported by the World Bank in other countries of the world with European Social Inclusion Partnerships (SIPs) which have been set up in Scotland and Ireland, among other areas in the EU, to bring diverse stakeholders together to develop regional strategies for social inclusion.21 The SDF was designed to target poor communities and groups including: (i) geographically isolated poor communities, and particularly Roma settlements; (ii) unemployed in disadvantaged regions; and (iii) other vulnerable groups including the disabled, homeless, single elderly and other groups requiring services traditionally provided in residential care institutions. In contrast with the other social assistance programs in Slovakia, the SDF aims to involve commu- nities in preparing, assessing, implementing, monitoring and evaluating projects. Since its estab- lishment in 2004, the SDF has targeted 8 regions in Slovakia with the most severe problems of exclusion and poverty. A total of 27 SIPs have been created, and 97 microprojects were approved for funding during the first Call for Proposals, affecting 88,000 beneficiaries. The three main pro- grams supported by the SDF are capacity building, small social infrastructure, and social work. The role of the SIPs is to include all possible stakeholders and target groups at the local level and enable them to take part in preparation and implementation of sub-projects. The SIPs develop strategies for social inclusion for their regions, assist in preparation of projects to ensure that they are in line with the strategy, establish the mechanisms of project assessment and recommend and evaluate the results of the projects and their impact on the strategy. SIPs have broad participation and must include representatives from four areas: (i) public administration (the SDF, local Offices of Labor, Social Affairs and Family, the Office of the Plenipotentiary for Roma Communities, local governments and other institutions); (ii) the private sector; (iii) community organizations (NGOs, schools, churches); and (iv) target groups. SIP members form a Partnership Board (10 to 15 mem- bers) to make policy decisions in consultation with the local community. Source: Pojarski, 2006. implemented by beneficiaries. It is also unique in that it serves as an instrument for targeting EU Structural Funds to poor communities. In the absence of such a mechanism, there is a risk that Structural Funds--such as the ESF--will not reach poor communities in need of investment and support. This is a model that Romania and Bulgaria are also considering prior to their accession to the EU. 21. See www.worldbank.org/socialfunds for further information. CHAPTER 6 Making Work Pay: Work and Social Protection inding a balance between social protection and incentives to work is a challenging F task. Beyond the specific design issues within social assistance programs discussed above, the aggregate effects of the social protection system can influence labor market outcomes. Social insurance contributions and taxes can create disincentives to work in the formal sector, particularly for low wage workers. Similarly, concerns about dependency traps, growing numbers of working poor, and efforts to meet the Lisbon employment targets, have place a renewed focus on measures to encourage employment and make work pay across Europe. Employment conditional benefits, or "in-work benefits," which aim to stimulate employment and provide social protection for low wage workers, are common in the EU15 and are gaining currency in the EU8. These include tax credits for low-wage workers, as well as lump sum and other benefits which taper off as wages rise. This section first analyzes the aggregate impact of taxes and social insurance contributions on labor market incentives, and then discusses considerations of in-work benefits for the new member states. Unemployment Traps Current tax and social policy regulations appear to discourage formal sector employment in most of the EU8. Unemployment traps--measured as the share of gross earnings which is taxed away through higher tax and social insurance contributions and the withdrawal of social benefits--in Latvia, Poland and Slovenia are even higher than in the EU15 (Box 5). In Latvia and Poland they have even increased in recent years. In these countries, the effective marginal tax rate for an unemployed person returning to work is around 80 percent of gross earnings--including the effect of taxes and reduced benefits. 35 36 World Bank Working Paper Box 5: Measuring Marginal Effective Tax Rates on Labor The EU and OECD define marginal effective tax rates and unemployment traps as falls. Marginal effective tax rates (METR) show what part of a change in earnings is "taxed away" by the combined operation of taxes, social security contributions and benefit thresholds22, and any with- drawal of earnings-related social benefits paid in cash (unemployment, family, housing, and min- imum income benefits23). The METRs allow for comparisons across countries and family types. High METRs indicate that workers face little financial incentive to increase work effort or to take up a job, instead of staying on welfare benefits. The METR results, presented below, were calculated within the joint European Commission-OECD project. METR 1 ( ynet)/( ygross) (1) where ygross denotes the additional earnings and ynet is the change in net income obtained after taxes and benefits so that the change in gross earnings between labor market states A and B is ygross ygrossB ygrossA (2) and the net earnings change is ynet ynetB ynetA (ygrossB ­ tB + bB) ­ (ygrossA ­ tA + bA) (3) where t denotes total taxes and b stands for total benefits. The OECD tax/benefit model was applied to calculate gross and net incomes for a set of different "hypothetical" family types. The results were scaled in relation to observed earnings levels across countries ranging from 0 to 200 percent of Average Production Worker in manufacturing (APW). The unemployment trap refers to a situation where benefits paid to the unemployed and their families are high relative to potential net earnings. Here, it measures the percentage of gross earnings which is taxed through income tax and social security contributions and the withdrawal of unemployment and other cash benefits when an unemployed person returns to employment. This indicator is available only for single persons without children earning 67 percent of the APW when in work. The low wage trap (or poverty trap) is related to the financial consequences of increasing working hours (or work effort) for low-paid workers. It refers to a situation where an increase in gross in-work earnings fails to translate into a net income increase that is felt by the individual to be a sufficient return for the additional effort. Here, it measures the percentage of gross earnings which is taxed away through the combined effects of income taxes, social security contributions and any withdrawal of social benefits paid in cash when gross earnings increase from 33 to 67 percent of the APW. The measures above cannot be considered as general statistical indicators, but rather as tax-benefit indicators (they describe features of tax-benefit systems), work incentive indicators (they relate to the trade-off applying to individuals' work decisions), and income indicators (they focus on current, disposable and cash incomes). They provide information for selected sets of circumstances (earnings level, household type). This is because they need to be interpreted with caution. Source: Carone G., Salomäki A., Immervoll H., Paturot D., Indicators of unemployment and low-wage traps (marginal effective tax rates on labor), Economic Papers No. 197, EC, December 2003. 22. Taxes which aretaken into account include national and local income taxes with only standard tax reliefs (those available to all taxpayers irrespective of family status; relief depending on their marital status; relief granted to families with children; and the relief for work-related expenses), and own social security contributions paid by employees and benefit recipients excluding voluntary contributions made to either private or public insurance institutions. The calculations relate to current incomes and therefore do not take into account any longer-term effects of today's labor market status on future earnings, pension entitlements etc. Moreover, all income measures were computed at the household level. 23. Disability benefits, private-, occupational- or state old-age pension payments as well as any income from capital are not considered. If the possibility of non means-tested and in-kind benefits was taken into account, the METR could be much higher at the bottom than the top end of the earnings distribution. Hence, the low-skilled are more likely to confront unemployment or poverty traps. Social Assistance in the New EU Member States 37 Figure 18. Unemployment trap in the EU8 in 2001 and 2004 (percent; earners of 67 percent of the APW) 100 90 87.1 85.8 83.0 80.5 80.5 81.0 80 77.0 79.178.3 75.0 70 67.0 66.0 65.0 60 54.1 50.4 48.8 50 48.2 43.0 40 30 20 10 0 LV PL SI HU CZ EE LT SK EU15 2001 2004 Source: Eurostat. However, some countries have made impressive strides in this area. Between 2001 and 2004, Slovakia impressively halved the taxation rates through its aggressive reform pro- gram, reducing the combined effects of taxes and benefits to the lowest in the region, at 43 percent. Hungary managed to reduce the unemployment trap significantly--although it remains relatively high, at 68 percent. Slight declines were also recorded in the Czech Republic and Lithuania. Low-wage traps (or poverty traps) are particularly high in Poland and lowest in Slovakia for single persons without children and highest in Latvia and lowest in Hungary for single-earner couples with two children (assumed ages are 4 and 6). A single person without children in Poland faces limited incentives--as she/he loses 65 percent of the gross wage to higher taxes and lower benefits if she/he starts to earn two-thirds instead of one-third of the average wage. This proportion is three times lower in Slovakia and equals to around 30 percent in the remaining EU8 countries.24 Poor couples with one wage earner and two children in the Baltic States and Poland have practically no incentive to increase their wages. If they earn two-thirds of the average wage instead of the one-third, their marginal effective tax rate is 80 to 100 percent of gross earnings. Slovenia and Hungary provide the strongest incentives to increase wages in order 24. It is important to note that average public wages (APW) used for this analysis are significantly higher than minimum wages in many countries. Minimum wage rates range between 35 percent (Slovakia) to 60 percent (Slovenia) of APW. Hence, the disincentives for low wage workers may be even more severe than discussed. 38 World Bank Working Paper Figure 19. Low-Wage Trap in the EU8 in 2001 and 2004 (percent; earnings increase from 33 to 67 percent of the APW) Single Persons Without Children 80 70 67.0 65.0 60 57.0 54.553.9 50 46.7 39.4 40 39.0 38.0 36.0 36.0 34.934.9 33.0 32.0 30 28.2 26.0 22.0 20 10 0 PL LV LT SI CZ HU EE SK EU15 2001 2004 One Income-earner Couples With Two Children 140 125 120 100 100 100 91 94 82 80 78 79 80 69 57 61 60 55 39 40 26 26 20 20 16 0 LV PL EE LT CZ SK SI HU EU15 2001 2004 Source: Eurostat. to exit from poverty for low wage single-earner couples with children, with a combined tax rate of 26 and 20 percent, respectively. The system of marginal effective tax rates, adjusted upwards by income tax and social security contributions and by the decline in social benefits, reflects relatively low progressivity Social Assistance in the New EU Member States 39 Figure 20. Marginal Effective Tax Rates for Various Types of Single-earner Households in Visegrad Countries in 2004 (percent) 40 40 35 32 3232 33 28 30 31 30 26 25 22 22 19 20 14 15 10 5 0 -5 -10 -8 earner of earner of 67 earner of earner of earner of earner of single 50% of APW % of APW 80% of APW 100% of 125% of 167% of earner of APW APW APW 67% of APW with two children CZ HU PL SK Source: Eurostat. in Poland and Latvia (and Table 7). This is rather paradoxical that the effective tax rates are most linear in Poland, and relatively more progressive in Slovakia and the other Baltic States which introduced flat income taxes in recent years. A simple explanation is the fact that even if flat income taxation exists, there is an income-tax threshold and social security contributions and social benefits may be progressive. The differences in tax rates between single parents earning 50 percent of the APW and those earning 167 percent ranges from only 4.3 percentage points in Poland and 4.4 percentage points in Latvia to a high of 26.4 percentage points in Hungary. A single parent with two children earning two-thirds of the AWP benefits in net terms from social protection/assistance in the Czech Republic, Hungary, Slovakia, and Estonia. In other words, the amount of cash social assistance would exceed the sum of income taxes and social security contributions that the individual would owe. In Poland and Slovenia, her/his income is still taxed by a rate of over 20 percent. Fortunately, they are better off in every country if compared her/his status with a person earning the same, but without chil- dren. In such a case, the effective tax rate declines the most in Estonia ( 33.5 percentage points), the Czech Republic ( 30.8 percentage points) and Hungary ( 26.9 percentage points), but only slightly in Poland ( 7.7 percentage points) and marginally in Lithuania ( 3.0 percentage points). Although the combination of taxes and social benefits generally favors married couples with children than single adults without children in the EU8 countries, the family-oriented Table 7. Marginal Effective Tax Rates for Various Types of Households in the EU8 Countries in 2004 (percent) Household type Single parent Married couple 1st Earner of % of APW 50 67 80 100 125 167 67 100 100 100 100 100 100 2nd Earner of % of APW -- -- -- -- -- -- -- 0 33 67 100 33 100 Children No children Two children No children Country Czech Republic 20.3 21.6 22.4 23.9 25.1 27.1 9.2 4.9 13.7 18.2 19.9 22.3 23.9 40 Estonia 15.2 18.5 20.1 21.7 23.0 24.3 15 12.5 11.0 13.8 15.9 19.1 21.7 Hungary 13.5 19.0 22.5 25.9 31.8 39.9 7.8 6.1 7.9 11.3 16.0 22.8 25.9 Latvia 25.5 27.1 27.8 28.6 29.2 29.9 10.7 16.1 18.8 21.4 23.1 27.1 28.6 Lithuania 16.3 21.3 23.7 26.1 28.1 30.1 18.3 24.2 19.7 23.0 25.1 21.2 26.1 Poland 28.5 30.0 30.8 31.5 32.1 32.7 22.3 29.5 30.0 30.9 31.5 30.0 31.5 Slovenia 30.5 31.6 32.2 33.4 36.4 39.5 24.1 27.7 27.8 29.3 30.6 32.1 33.4 Slovakia 13.4 17.1 19.2 21.3 23.0 24.7 1.2 1.2 4.3 12.3 15.2 13.4 21.3 EU15 19.1 24.1 27.1 29.8 32.2 34.8 1.7 14.8 18.5 21.9 25.1 25.3 29.8 Source: Eurostat. Social Assistance in the New EU Member States 41 incentives are marginal in Poland and very small in Lithuania and Slovenia. Married couples with children, in which one spouse earns 100 percent of the AWP and the other one-third of the average wage, face more favorable combined tax rates on labor than single adult without children earning 100 percent of the AWP. The difference is most significant in Hungary ( 17.9 percentage points) and Slovakia ( 17.1 percentage points), and smallest in Poland ( 1.5 percentage points) and Slovenia ( 5.6 percentage points). Moreover, Poland is the only EU8 country where the status of a two-earner married couple, both earning 100 percent of the APW, does not improve their financial position at all when they have two children compared to a single adult without children earning the average wage. In this respect Hungary (a fall of combined tax rate by 9.9 percentage points) and Slovakia ( 6.1 percentage points) can be regarded as most family-friendly. In-Work Benefits25 The high unemployment traps for low wage workers and families suggest the need for tax credits and other in-work benefits (IWBs) which can support low-income households and facilitate the transition into employment without reinforcing dependency. Estonia has introduced tax credits for families with children, and Hungary and Slovakia have benefits which taper off as individuals enter the labor market. This section discusses some implications of IWBs in the EU8 countries, based on experience in other European countries. In-work benefits (also called employment-conditional benefits) are efforts to simultaneously provide income support to low income households and boost employment. These benefits are paid only to persons who have taken up a full-time or part-time job. IWB programs are established primarily with a view to reduce benefit dependency and increase employment: they are designed to increase the net income from employment and the difference between in-work income and out-of-work benefits, thereby increasing employment incentives. Further aims of these programs include: redistributing income towards low- and middle-income families, thereby reducing poverty of families with children; and tackling high unemployment among less skilled workers. IWBs may take the form of tax credits, wage-related transfers or lump-sum payments. The choice of the program type largely depends on the target group. Tax credits are directed to low income working families, aim at improving the net income of these families and thus contribute to reducing poverty, while at the same time creating work incentives. Wage-related transfers and lump-sum payments are aimed at those currently not in work, seeking to ease the transition from long-term unemployment to work. Normally, tax credit schemes have no time limit, while wage-related transfers are time-limited. There is also potential interplay with activation measures and active labor market policies. IWBs can complement active labor market policies. While some IWBs, like back-to-work-allowances, could be considered part of the labor market policy package directed at the unemployed, tax credits are directed at those already in employment. Therefore, the target groups of IWBs and active labour market policies do not always coincide. 25. This section draws on Leppik (2006). 42 World Bank Working Paper OECD experience suggests that IWBs can be particularly important for countries which have a relatively wide earnings distribution, low minimum wages and where social protection programs create poverty traps. This is particularly relevant for the EU8 countries. It is also argued that compared to the minimum wage, IWBs are better targeted to deal with in-work poverty among families. Another positive feature attributed to IWBs is that they may act as a temporary earnings "insurance" (OECD 1998). IWBs have potential benefits from a political economy perspective. Tax credits, in particular, may be more acceptable to the wider public than increasing social assistance benefits, because of concerns of welfare dependency and stigma. In-work benefits, in particular tax credits, have emerged as the politically most acceptable instruments in tax-benefit reforms of many Anglo-Saxon countries. Administrative costs also merit consideration. In the UK and other countries, tax credits have been introduced into sophisticated income tax schemes which have progressive tax rates and different tax breaks, creating administrative complexity. The UK tax credits have been criticized for high levels of leakage as the system is unable to respond quickly to changes in family circumstances, resulting in high levels of repayments, penalties, and sometimes hardship for low income families. In 2004­05, overpayments comprised nearly one-third of all tax credit awards in the UK, or over 12 percent of total expenditures on tax credits (Leppik 2006). Because tax reforms in several of the EU8 (Estonia, Latvia, Lithuania, Slovak Republic) have introduced flat rate income taxes with simple structures and administrative procedures and very few extra deductions, administrative costs may be less of a concern. CHAPTER 7 Conclusions W ithin a very short period of time, the new member states of the EU have put in place programs and institutional arrangements for delivering income support and other benefits to the poor. Social assistance programs and policies for tackling poverty and exclusion exist across countries and are central elements of each of the countries National Action Plans for Social Inclusion. Although means-tested social assistance benefits make up a relatively small share of GDP and total spending on social protection, they are important sources of income for those who receive them. Given the limited budgets for social assistance, ensuring that resources are spent well, reach those who need them, and optimize welfare gains and labor market outcomes are important tasks. They also requires attention to the balance between spending on means- tested anti-poverty programs and family benefits and social insurance programs. Manag- ing the long-term fiscal sustainability of these programs to meet emerging needs is a priority, including ensuring effective institutional and administrative arrangements, and making decentralization of benefit delivery work within the context of evolving intergov- ernmental fiscal arrangements. 43 Appendixes APPENDIX A Family Benefits in Hungary26 H ungary has one of the most complex family benefit systems in Europe. The development of the system has been fuelled by both pro-natality and income support objectives. In mid-2005 this fragmented system consisted of the following types of benefits: Maternity Allowance (Terhességi-gyermekágyi segély): Mothers are entitled to two types of benefits depending on their previous employment situation. If they were insured for at least 180 days in the two years preceding delivery they are entitled to maternity allowance. It is paid for 24 weeks (4 weeks before and 20 weeks after the planned date of birth, or 24 weeks after the date of birth, depending on the mothers' choice) and amounts to 70 percent of the daily average gross earnings of the previous year. Maternity Grant (Anyasági támogatás): Hungarian residents who give birth, and have previously participated in prenatal care at least 4 times, and have no insurance are entitled to a one-off lump-sum payment of 225 percent of the minimum old-age pension (Öregségi nyugdíj) = Ft 55575 or 300 percent = Ft 74,100 in case of twins. The Child Care Allowance (Gyermekgondozási segély) is a universal entitlement financed from the state budget that provides a flat-rate benefit to parents who stay away from work to care for their children under the age 3 (under the age of 10 in case of permanently ill or severely disabled children) or for grandparents who care for their grandchildren aged between 1­3 years in the household of the parent. 26. This Annex draws from Szivos (2006). 47 48 World Bank Working Paper In case of twins the allowance is paid until the children reach the compulsory schooling age (usually 6 years of age). The monthly amount is equal to the minimum old-age pension of Ft 23,200, in case of twins the amount is doubled. TheChildCareFee(Gyermekgondozásidíj):isacontributorybenefit,whichispaid after the expiry of the Maternity Allowance until the child reaches 2 years of age if the parent does not work. The eligibility criterion is at least 180 days of insurance during the two years preceding delivery of the parent who chooses to take care of the child at home. It amounts to 70 percent of the daily average gross earnings of the previous year with a maximum of Ft 83,000 per month. Parents entitled to the child care fee cannot be eligible for child care allowance. Child Raising Support: (Gyermeknevelési támogatás) is a universal benefit financed from the state budget for parents who raise three or more children in their own home, if the youngest child is between 3 and 8 years old. The monthly amount is equal to the minimum old-age pension, irrespective of the number of children. In Hungary a well-developed child benefit system is operating consisting of universal and means-tested benefits as well. TheFamilyAllowance(Családipótlék)isauniversalbenefitfinancedfromthestate budget. It is paid to the parent from the birth of the child to the termination of the child's studies in the compulsory education system (usually 0­16 years), and then during the secondary school education or vocational training of the child (up to 24 years of age). Its amount depends on the number of children in the family, whether it is a single-parent family or not and whether the child is disabled. In the month of July double amounts are paid in order to support schooling. Monthly amounts: 1 child in family: Ft 5,100, 1 child, single parent: Ft 6,000, 2 children in family: Ft 6,200 per child, 2 children single parent: Ft 7,200 per child, 3 or more children in family: Ft 7,800 per child, 3 or more children, single parent: Ft 8,400 per child, permanently ill or severely disabled child in family: Ft 13,900, permanently ill or severely disabled child, single parent: Ft 15,700, child in foster home/at foster parent: Ft 7,200. Advanceonmaintenancepayments(Tartásdíjmegelolegezése):Thisbenefitispaidto the parent who takes care of a child if the child maintenance is temporarily irrecover- able by the other parent obliged to maintain the child, or if the person who takes care of the child cannot support the child, and if the income per person in the applicant's family does not exceed three times the current minimum amount of old-age pension. Its amount is equal to the amount of child maintenance decided by the court. Regular Child Protection Benefit (Rendszeres gyermekvédelmi támogatás): This benefit is paid to the family if the income per person in the family does not exceed the amount of the minimum old-age pension of Ft 24,700. The local government Social Assistance in the New EU Member States 49 Table A1. Expenditure on Family (Maternity and Child) Benefits (millions of Hungarian forints) Number of beneficiaries, Benefit 2004 (expected) 2004 2005 (planned) Maternity allowance 21 348 28 000 25 340 Maternity grant 5 315 91 673 5 432 Child care allowance 53 102 163 000 54 848 Child care fee 53 019 84 000 57 941 Child raising support 14 436 47 000 14 697 Family allowance 187 887 1 287 000 (families) 195 938 2 10 9000 (children) Regular child 36 905 675 000 (670 167) 39 193 protection benefit Family tax allowance 67 500 1 150 000 65 000 can also review the financial situation of the family. The monthly amount of the regular child protection benefit is 22 percent of the minimum old-age pension = Ft 5,434. The regular child protection benefit can be provided even after the child reaches 18 years of age if he/she is a regular student, until s/he is 23 in the case of secondary education, and until 25 in the case of tertiary education. TheIrregularChildProtectionBenefit(Rendkívüligyermekvédelmitámogatás)is paid to families with temporary cash flow problems or facing emergency situations that seriously threaten their standard of living. The amount is designated by a decree of the respective local government. In 2004, this benefit was paid to 270,000 children with a total amount of Ft 2160 million, which means an average payment of Ft 8000 per child. Family tax allowance (Családi adókedvezmény): Every household with children is eligible for family tax allowance if they have a taxable income. Its amount varies according to the number of children in the family: Onechildfamily:Ft3000, Twochildrenfamily:Ft4000/child, Threeormorechildren:Ft10000/child. A minor simplification of the fragmented system occurred at the beginning of 2006. The main objective of this step was to reach a higher degree of fairness. The family tax allowance was kept only for families with three or more children, the irregular child pro- tection benefit was abolished, but in parallel, the family allowance was doubled--making the system more universal than before. APPENDIX B Methodology: The Consumption Aggregate27 he choice of consumption rather than income to measure for poverty was dictated by T practical considerations. Income data remain particularly difficult to collect in the new member states. In contrast, practice has shown that consumption data can be gathered with a great degree of precision. Survey consumption modules have become more detailed over time, and better capture various dimensions of consumption including informal payments etc. The study relied on these comparable data to construct measures of welfare. Unlike food, consumer durables and housing are consumed over a long period of time. It is customary, therefore, to include the imputed value of the consumption flow associated with the possession of consumer durables (including housing) but exclude the expenditure on the purchase of the these goods. However, for the countries discussed in this report, data availability limited the application of this approach to all countries and thus the second best approach is to exclude both rents and purchases of durables from consumption. Price indices were used all countries and quarterly consumer price indices (from IMF data) were used to compute real values. A consistent approach in assigning a monetary value to these components of consumption was applied to all datasets. To adjust for differences in household composition a per capita scale was used. Finally, the same procedure, which conforms to methods used in other international household survey data programs such as the Luxemburg Income Study, was used to clean the data of outliers across all data sets. As a consistent approach was followed across all data sets, one can be reasonably confident that differences across countries in the final consumption mea- sure are due to differences in the primary data and are not owed to the method of aggregation. 27. This Annex is based on World Bank (2005a). For more details, refer to World Bank (2005a), Appendix, p. 220. 51 Table A.2. Social Protection Spending in the EU8, 2003 % of GDP EU15 CZ EE LV LT HU PL SI SK SOCIAL PROTECTION BENEFITS 27.2 19.8 13.2 13.0 13.1 21.0 20.6 24.0 17.8 Old-age 11.1 8.0 5.8 6.6 6.0 7.5 11.0 10.4 6.8 Disability 2.1 1.6 1.2 1.1 1.3 2.2 2.5 2.0 1.6 Survivor 1.3 0.2 0.1 0.3 0.3 1.1 1.0 0.4 0.2 Sickness/Health care 7.7 7.1 4.2 3.0 3.9 6.2 4.2 7.8 5.8 Unemployment 1.8 0.8 0.2 0.4 0.2 0.6 0.8 0.7 1.0 52 Family/Children 2.2 1.5 1.3 1.4 1.0 2.7 1.0 2.1 1.5 Housing 0.5 0.1 0.1 0.1 0.0 0.5 -- -- 0.1 Social exclusion n.e.c. 0.40 0.59 0.21 0.12 0.44 0.13 0.04 0.62 0.81 ADMINISTRATION COSTS 0.9 0.3 0.2 0.3 0.4 0.4 0.3 0.5 0.6 OTHER SOCIAL PROTECTION EXPEND. 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 SOCIAL PROTECTION EXPENDITURES 28.3 20.1 13.4 13.3 13.6 21.4 20.9 24.6 18.4 Non-means tested to total benefits (%) 90.5 93.3 97.9 98.5 95.4 94.7 95.9 90.2 92.8 Cash benefits to total benefits (%) 69.1 66.4 69.4 73.4 64.2 63.6 82.9 67.6 65.8 Total pensions (% of GDP) 12.7 8.9 6.3 7.5 6.8 9.3 13.9 11.2 7.5 Pensions to total SP expenditures (%) 44.9 43.9 46.9 56.2 50.3 43.3 66.3 45.5 40.4 Source: Eurostat based on ESPROSS. Social Assistance in the New EU Member States 53 Table A.3. Minimum Income Guarantee Programs (percent of GDP) 1999 2000 2001 2002 2003 2004 Czech Republic 0.38 0.45 0.42 0.45 0.48 0.46 Estonia 0.41 0.34 0.34 0.30 0.26 0.17 Latvia 0.03 0.02 0.02 0.02 0.04 0.05 Lithuania 0.17 0.19 0.18 0.17 0.17 0.11 Poland 0.40 0.38 0.38 0.44 0.41 0.19 Slovakia 1.13 1.11 1.13 1.04 0.75 0.48 Slovenia 0.22 0.22 0.24 0.31 0.44 0.48 Sources: National Budget data, authors' calculations. Table A.4. Minimum Income Guarantee Programs, Beneficiaries (percent of population) Country 1999 2000 2001 2002 2003 2004 Czech Republic -- -- -- -- 4.0 3.6 Estonia 5.9 4.8 5.2 5.1 3.8 2.5 Latvia 2.2 2.6 2.7 2.9 2.9 3.2 Hungary 0.3 0.5 0.9 1.2 1.4 -- Lithuania 2.9 3.3 3.4 3.4 3.4 2.4 Poland 4.2 4.2 3.6 1.7 1.5 -- Slovakia 5.5 5.9 6.0 6.0 5.2 3.3 Slovenia 1.6 1.6 1.7 1.9 2.4 2.7 Sources: National data, authors' calculations. Table A.5. Spending on Child Benefits (percent of GDP) 1999 2000 2001 2002 2003 2004 Czech Republic 0.6 0.6 0.6 0.6 0.5 0.4 Estonia 0.8 0.7 0.6 0.6 0.7 Hungary 0.3 0.2 0.3 0.2 0.2 -- Latvia 0.7 0.6 0.6 0.5 0.6 0.5 Lithuania -- -- -- --- -- 0.2 Poland 0.5 0.5 0.5 0.4 0.4 0.3 Slovakia 1.1 0.9 0.8 0.8 0.7 0.6 Slovenia 0.9 1.0 1.0 1.0 0.9 -- Sources: National Budget Data, authors' calculations. 54 World Bank Working Paper Table A.6. Coverage of Child Allowances (percent of population receiving) 1999 2000 2001 2002 2003 2004 Czech Republic 19.8 20.0 19.9 19.8 19.2 18.5 Estonia Hungary 7.9 7.7 7.7 7.5 7.0 Latvia 20.5 13.2 13.0 12.8 12.7 12.6 Lithuania 7.7 Poland 17.8 19.2 19.2 16.8 15.6 Slovakia 10.5 9.9 9.4 11.2 15.0 14.0 Slovenia 13.6 13.5 13.4 13.2 13.1 Sources: National Data, authors' calculations. Table A.7. Summary Indicators, Social Protection Benefits Targeting Average benefit Benefit Coverage (amounts) CGH Adequacy (PPP adjusted) Q5/Q1 Total Q1 Q5 Q1 Q5 For Q1 Total Q1 Q5 TOTAL SOCIAL PROTECTION Estonia '03 928 1.29 78% 81% 67% 18% 19% 0.89 40% 91% 22% Estonia '04 1,053 1.22 79% 84% 63% 20% 18% 0.98 39% 91% 23% Hungary '00 1,375 1.67 85% 93% 70% 17% 22% 0.87 55% 87% 42% Hungary '04 1,864 1.73 90% 99% 73% 17% 22% 0.85 56% 92% 41% Poland '00 1,414 3.87 73% 85% 58% 12% 31% 0.58 60% 65% 57% Poland '04 1,847 5.14 69% 83% 56% 10% 34% 0.49 65% 64% 65% Latvia '02 831 1.52 85% 91% 68% 17% 19% 0.84 30% 62% 18% 55 Latvia '04 839 1.29 84% 93% 68% 19% 17% 0.93 27% 61% 15% Lithuania '00 1,012 2.19 61% 71% 49% 15% 23% 0.76 47% 71% 33% Lithuania '04 1,125 1.81 62% 76% 45% 18% 19% 0.89 44% 78% 29% Slovenia '03 2,490 2.43 85% 95% 67% 16% 26% 0.78 47% 70% 42% PENSIONS Estonia '03 1,492 1.29 38% 36% 31% 16% 18% 0.82 64% 149% 36% Estonia '04 1,709 1.40 39% 44% 31% 19% 19% 0.95 64% 132% 38% Hungary '00 1,931 2.28 45% 43% 41% 12% 26% 0.59 76% 92% 63% Hungary '04 2,565 2.42 48% 46% 39% 12% 25% 0.60 77% 100% 63% Poland '00 1,819 3.32 44% 38% 44% 9% 33% 0.43 71% 82% 63% Poland '04 2,360 4.19 40% 34% 43% 7% 37% 0.36 75% 82% 69% Latvia '02 1,313 1.97 44% 47% 31% 15% 20% 0.75 48% 84% 32% (Continued) Table A.7. Summary Indicators, Social Protection Benefits (Continued) Targeting Average benefit Benefit Coverage (amounts) CGH Adequacy (PPP adjusted) Q5/Q1 Total Q1 Q5 Q1 Q5 For Q1 Total Q1 Q5 Latvia '04 1,321 1.72 42% 50% 28% 17% 16% 0.85 44% 80% 27% Lithuania '00 1,189 2.00 38% 38% 33% 13% 23% 0.65 54% 82% 37% Lithuania '04 1,436 1.66 37% 42% 26% 17% 17% 0.86 55% 101% 35% Slovenia '03 3,947 3.06 44% 53% 39% 13% 29% 0.65 74% 88% 63% UNEMPLOYMENT BENEFITS Estonia '03 309 1.55 4% 6% 2% 21% 9% 1.05 15% 25% 8% 56 Estonia '04 314 1.42 3% 4% 1% 22% 8% 1.11 13% 23% 7% Hungary '00 344 1.56 11% 24% 4% 40% 10% 2.00 15% 26% 11% Hungary '04 438 1.87 10% 24% 2% 43% 7% 2.14 15% 27% 13% Poland '00 1,035 2.20 3% 2% 2% 12% 23% 0.58 41% 62% 31% Poland '04 1,307 2.68 5% 4% 4% 9% 23% 0.45 43% 61% 34% Latvia '02 492 4.13 4% 6% 3% 15% 32% 0.74 20% 26% 20% Latvia '04 543 3.97 4% 6% 1% 22% 19% 1.10 24% 32% 27% Lithuania '00 380 1.93 3% 4% 2% 21% 22% 1.05 18% 32% 12% Lithuania '04 374 3.45 3% 6% 1% 29% 17% 1.43 18% 24% 19% Slovenia '03 899 1.82 6% 9% 4% 24% 20% 1.20 17% 31% 14% FAMILY BENEFITS Estonia '03 179 0.83 47% 53% 39% 25% 15% 1.27 8% 23% 4% Estonia '04 212 0.81 47% 52% 35% 25% 13% 1.25 8% 23% 4% Hungary '00 392 0.61 50% 71% 28% 36% 8% 1.80 15% 40% 7% Hungary '04 437 0.59 53% 77% 33% 37% 10% 1.87 12% 37% 6% Poland '00 144 0.86 31% 56% 8% 39% 5% 1.96 6% 15% 3% Poland '04 212 1.44 27% 58% 3% 40% 3% 2.01 8% 17% 5% Latvia '02 136 0.74 53% 64% 37% 29% 13% 1.45 5% 16% 2% Latvia '04 146 0.79 52% 66% 43% 31% 16% 1.53 5% 15% 2% Lithuania '00 188 0.96 14% 29% 4% 40% 5% 1.98 9% 20% 4% Lithuania '04 143 0.85 20% 32% 10% 38% 10% 1.90 5% 17% 3% Slovenia '03 276 0.51 49% 67% 22% 37% 6% 1.83 5% 15% 2% TARGETED SA Estonia '03 421 2.01 3% 8% 0% 50% 4% 2.49 20% 47% 14% Estonia '04 467 1.01 2% 4% 0% 33% 4% 1.64 19% 37% 7% Hungary '00 317 1.61 4% 8% 1% 37% 8% 1.86 14% 24% 11% Hungary '04 426 1.57 2% 6% 1% 42% 8% 2.10 15% 27% 10% Poland '00 472 1.76 2% 5% 1% 36% 8% 1.80 22% 36% 14% 57 Poland '04 587 2.32 3% 5% 1% 26% 9% 1.31 24% 35% 18% Latvia '02 50 0.84 4% 9% 2% 48% 9% 2.38 2% 5% 1% Latvia '04 47 1.67 5% 8% 4% 27% 24% 1.35 2% 4% 1% Lithuania '00 324 1.10 4% 11% 1% 63% 5% 3.14 15% 38% 10% Lithuania '04 214 2.49 3% 8% 1% 56% 15% 2.82 12% 22% 11% Slovenia '03 690 1.80 9% 14% 5% 28% 20% 1.42 13% 26% 10% Source: Household Budget Surveys. Table A.8. Summary of Programs included in the Household Budget Surveys Lithuania Lithuania Hungary Hungary Poland Poland Estonia Latvia Slovenia 2000 2004 2000 2004 2000 2004 2003 & 2004 2002­2004 2003 CONTRIBUTORY Contributory Old-age Old-age Pension, Pension, Old-age Old-age Old-age Old-age Pensions Pensions pension pension annuity, annuity, pensions pensions pension Disability Supplements Disability Disability disability disability Disability Disability Disablement Service and Pensions from pension pension for pension pension pensions pensions pension special abroad Injuries persons after pension persons state retirement age Foreign pensions Disability pension pension for persons before retirement age Injuries persons state pensions 58 Unemployment Unemployment Unemployment Regular social Regular social Other Other Unemployment Unemployment Unemployment benefits benefit benefit support for support for assistance unemployment benefit benefits benefits Severance and unemployed unemployed from the allowances Retraining termination Unemployment Unemployment Labor Fund Other allowance for pay* benefit benefit assistance the unem- from the ployed Labor Fund Other Social Loss of bread- Loss of bread- Sick-fee** Allowance for Maternity Maternity Sick pay Sickness Maternity leave Insurance winner's pen- winner's pen- Child care fee pregnancy allowances allowances Maternity leave allowance compensation sion (widow, sion (widow, Sick-fee** allowance Maternity Parental orphan) orphan) for Child care fee Survivor's benefits Other allowance Sick payment** persons after (labor, retirement age pension Maternity, accident, paternity bene- Loss of bread- occupational fit** winner's pen- diseases, sion (widow, insurance pre- orphan) for miums etc.) persons before Survivor's retirement age pension Sick payment** Maternity, paternity benefit** NON-CONTRIBUTORY (to be subtracted from total income) Other Pensions Social pension Social pension Old-age benefit Old-age benefit Family Family (incl. Social) Other state for persons Supplement of Supplement of pensions pensions pensions after retire- pension pension ment age Other pensions Social pension for persons before retire- ment age Other state pensions Other pensions Total Social Family benefit Family benefit Child care Child care aid Family Family Maintenance Child care Childbirth Assistance, of Benefit for Benefit for allowance Child care allowances allowances benefit for benefit allowance which Family families with families with Child care support child Child-support Child-support child State family Child care Allowances three and more three and more support child related tax payments from payments from Child benefit, allowance assistance incl. child ben- children children related tax benefit child-support child-support school benefit efits Benefit to Child Extraordinary Extraordinary benefit Family fund fund Child benefit guard for allowance Upbringing 59 single grant for single grant for Family allowance, Child-support for single supporting a birth of a birth of a allowance, schooling aid payments from allowances parents child child* child* schooling aid Child care aid child-support Maternal Remuneration Pregnancy Pregnancy Child care aid cash fund temporary and for doing guard allowance to allowance to cash one-time duties Child care aid Maternal tem- the women the women allowances Child care aid in kind porary and Child birth who study** who study** in kind one-time grant Maternity aid allowances Child birth aid Means tested Social benefit Social benefit Regular Regular Regular Regular Means-tested Cash benefits Financial SA support, aid support, aid benefits from benefits from benefit (incl. to low-income assistance as social welfare social welfare housing families the only source benefit) (including of subsistence housing), food Financial social stamps, and assistance other local govt. cash Social benefits benefits (Continued) Table A.8. Summary of Programs included in the Household Budget Surveys (Continued) Lithuania Lithuania Hungary Hungary Poland Poland Estonia Latvia Slovenia 2000 2004 2000 2004 2000 2004 2003 & 2004 2002­2004 2003 Other Social Child foster Child foster Occasional Occasional Childbirth, Childbirth, Other state or State social Housing Assistance benefit benefit support, aid* support, aid* funeral, funeral, municipal security benefit assistance Burial benefit* Burial benefit* Scholarship Scholarship sickness sickness subsidy (social Disability allowances allowances pensions for Other state Other state Support for Support for allowances (after cessation (after cessation those without benefits and benefits and housing (public housing (public of employment of employment old-age Scholarships compensations compensations utility) utility) relationship), relationship), pensions) Scholarship Other benefits Orphans Orphans rehabilitation rehabilitation Compensation allowance* allowance* for transport Scholarship benefit other benefit other social support social support Housing Housing costs for Allowances for in kind in kind supplementary supplementary disabled disability allowances allowances persons (transport) Temporary Temporary Funeral Nursing benefits from benefits from allowance benefits social welfare* social welfare* Allowance for victims of 60 Other material Other material support, support, Chernobyl including in including in Local govt. the form of the form of in-kind services services benefits (rent, Nursing Nursing electricity, allowances allowances heating, phone bills, fuel, other) Assistance Assistance from non- from non- Benefits for the commercial commercial care of sick, institutions institutions old, or disabled persons Unemployment Other social benefits assistance Health care income benefits in-kind Unemployment Funeral Grants benefits Other local govt. in-kind benefits Scholarship * One time lump sum. ** About six months duration. Table A.9. Child Allowances, Main Features, 2005 Benefit level (monthly Financing Country Name Description Eligibility amount per child) Duration source Czech Republic pridavek na dite Income-tested Family income Depends on income and age Until age 18 or 26 State budget. child benefits. below 3 times of the child. Range is if a student. the minimum 241­797 CZK (9­28) living standard. Estonia lapsetoetuse määr Universal child Residents. 150 EEK (10) State budget. benefit. Hungary családi pótlék Universal child Citizens, legal Depends on number of Until age 16 or 24 State budget. benefit. refugees and children and household if a student. immigrants. characteristics (e.g. with a single parent, disabled child, foster child). Range 61 is from Ft 5,100 to 15,700 (19­59). Latvia G¸imenes valsts Universal child Residents. Depends on number of Until age 15 or 20 State budget pabalsts benefit. children. Range is from 6 LVL if a student. for the first child to 10.8 LVL for the 4th+ child (9­15). Lithuania Universal child Residents. Depends on age and number Age 7 for families State budget benefit. of children. Range is 50 to with 1­2 children; 137 LTL (14­40). until 18 for families with 3+, or 24 if a student. Poland Income-tested Residents with Until age 18 or 24 State budget. child benefit. family income if a student. below 504 PLN per month or 583 for a disabled child. (Continued) Table A.9. Child Allowances, Main Features, 2005 (Continued) Benefit level (monthly Financing Country Name Description Eligibility amount per child) Duration source Slovakia prídavok na diet'a Universal child Residents. 540 SKK (15) Until age 16 or 25 State budget. benefit. if a student or disabled. 62 Slovenia otroski dodatek Income-tested Residents. Depends on income, number Until age 18 or 25 child benefit. of children and household if a student or characteristics (e.g. single disabled. parent, child not in preschool). Range is 3750­25,900 SIT (16­108) Note: Conversions to euros are estimated on current exchange rates. Sources: National experts, MISSOC Database. Table A.10. Guaranteed Minimum Income Programs, Main Features, 2005/6 Minimum income Method of Financing Work Country Name threshhold determination Indexation Duration source Administration incentives Czech Social 2,360 CZK Adjusted Willingness Republic Assistance (83) in annually to work is a Benefit 2005 for (January) requirement. (dávky a single based on sociálne péce) adult. cost of living increases. Estonia Subsistence 750 EEK Difference Adjusted Unlimited, Municipalities Registration benefit (48) in between annually by renewed determine at the labor (toimetulekutoetus) 2006 for income, Parliament. monthly. eligibility. office; the first after housing benefits member of expenses, may be the family. and the refused to Additional subsistence applicants 63 members are minimum. of working 80% of the age. subsistence level of the first family member. Hungary Regular Income There are Monthly Joint Municipalities Social support to 2 groups of benefits financing determine Support 70­80 percent beneficiaries: by state and eligibility. (rendszeres of the (a) adults 18+ who municipal szociális segély) minimum are unable to work budgets. old pension. (lost 67% of ability) 19,760 HUF and monthly (80) in 2005. income is below 80% of minimum old age pension; or (b) active age, unemployed, with income below 70% of minimum old age pension. (Continued) Table A.10. Guaranteed Minimum Income Programs, Main Features, 2005/6 (Continued) Minimum income Method of Financing Work Country Name threshhold determination Indexation Duration source Administration incentives Latvia Guaranteed 24 LVL (34) Difference Adjusted Three State budget. Registration Minimum Income per adult between the annually by months, at the State (Pabalsts garenteŻtaŻ in 2006. minimum and the Cabinet unlimited Employment minimaŻlaŻ income. of Ministers. renewals, Service. 64 ineaŻkuma liŻmen¸a but not to nodrosinaŻsanai) exceed 9 months within a year. Lithuania Social Benefit 155 LTL 90% of the Adjusted by Three State budget. Municipalities (socialné pasalpa) / (45) per difference government months, determine adult in 2006. between actual decision. unlimited eligibility. family income renewals. and the minimum level. Poland Social Assistance 316 (81) Difference Adjusted Unlimited. State budget Local social Registration (Opieka spoleczna) / PLN per adult between the annually (to be assistance with the in 2005 for minimum and based on transferred centers. labour the permanent income, not to the consumer to local office. allowance, exceed 418 PLN price index. budgets zasilek staly. / / per month or in 2007). be lower than 30 PLN. Slovakia Subsistence 1,560 SK Difference Adjusted Paid for 24 Registration Minimum (42) for a between the annually (July 1) months by at the (zivotné single adult. minimum by the Ministry the state Offices of minimum) and income. of Labour, budget, then Labour, Social Affairs transferred to Social and Family, municipalities. Affairs based on cost and Family. of living increases. Slovenia Temporary 46,981 Difference Adjusted Six months, State budget. Centers for Registration Cash Social tolars in between the annually unlimited Social Work at the Assistance 2005 minimum (January) renewals. determine National (denama (196). and income, based on eligibility Employment socialna adjusted cost of Service, pomoc) for family living actively size. increases. seeking work or 65 participating in active programs. Note: Conversions to euros are estimated on current exchange rates. Sources: National experts, MISSOC Database. References Barr, N. ed. 2005. Labor Markets and Social Policy in Central and Eastern Europe: The Acces- sion and Beyond. Washington, D.C.: The World Bank. Carone, G., A. Salomaki, H. Immervoll, and D. Paturot. 2003. "Indicators of unemploy- ment and low-wage traps (marginal effective taxes on labor)." Economic Papers, Euro- pean Commission, Directorate-General for Economic and Fiscal Affairs. European Commission. 2005. "Report on Social Inclusion 2005: An Analysis of the National Action Plans on Social Inclusion (2004­2006) submitted by the 10 new Member States." Directorate-General for Employment, Social Affairs, and Equal Opportunities, Luxembourg. European Commission. 2004. "Household Budget Survey in the Candidate Countries: Methodological Analysis 2003." Eurostat, Luxembourg. Kasek, L., T. Laursen, and E. Skrok. 2006. "Public Finance, Employment and Growth in the EU8," Paper prepared for the 8th Bank of Italy Workshop on Public Finance, Perugia, Italy, March 30­April 1. Kornai, J. 1992. "The Socialist System: the Political Economy of Communism." Princeton University Press. Laursen, T., ed. 2005. Current Issues in Fiscal Reform in Central Europe and the Baltic States, Washington, D.C.: The World Bank. Leppik, L. "In-Work Benefits: Literature Review." Background paper prepared for the EU8 Social Inclusion Study. Lindert, K., E. Skoufias, and J. Shapiro. 2006. "Redistributing Income to the Poor and the Rich: Public Transfers in Latin America and the Caribbean." The World Bank, Washington, D.C. Moffitt, R. 1998. Welfare, the Family, and Reproductive Behavior: Research Perspectives. Washington, D.C.: National Academy Press. Ochel, W. 2001. Financial Incentives to Work--Conceptions and Results in Great Britain, Ireland and Canada. CESifo Working paper No. 627, www.CESifo.de Pietka, K. 2006. "Social Protection in Poland." Background paper for the EU8 Social Inclu- sion Study, CASE, Warsaw. Pojarski, P. 2006. "Rapid Assessment of the Slovakia Social Development Fund." The World Bank, Bratislava. Draft. Ravallion, M. 1994. Poverty Comparisons. Philadelphia: Harvard Academic Publishers. Ringold, D., M. Orenstein, and E. Wilkens. 2005. Roma in an Expanding Europe: Breaking the Poverty Cycle. Washington, D.C.: The World Bank. Sapir A. 2006. "Globalization and the reform of European social models." Background document for the presentation at ECOFIN Informal Meeting in Manchester, Septem- ber 9, Bruegel, Brussels. 67 68 World Bank Working Paper Schwarz, A. 2005. "Pension Reform in the EU8: The Remaining Agenda." The World Bank, Washington, D.C. Processed. Szivos, P. 2006. "Social Policy Review: Hungary." Background paper for the EU8 Social Inclusion Study. TARKI. Budapest, Hungary. Tomova, J. 2006. "New Social Policy in the Slovak Republic: Social Benefits Reform, Moti- vation, Targeting and Assistance." Presentation to the World Bank EU8 Social Inclu- sion Workshop. Warsaw, Poland, May 8­9. Trumm, A. 2006. "EU8 Social Policy Review: National Report of Estonia." Background paper prepared for the World Bank. World Bank. 2004. "Growth, Employment and Living Standards in Poland." Report No. 28233-POL, Washington, D.C. ------. 2005a. Growth, Poverty, and Inequality: Eastern Europe and the Former Soviet Union. Washington, DC. ------. 2005b. "The Quest for Equitable Growth in the Slovak Republic: A World Bank Living Standards Assessment," Report No. 32422-SK, Washington, D.C. ------. 2005c. "World Bank EU8 Quarterly Economic Report: October 2005." Warsaw, Poland. ------. 2006a. Enhancing Job Opportunities: Eastern Europe and the Former Soviet Union. Washington, D.C. ------. 2006b. "Latvia: Sharing the High Growth Dividend, A Living Standards Assess- ment." Washington, D.C. Zalimiene, L. 2005. "Program Implementation Matters for Targeting Performance: Evi- dence and Lessons from the ECA Region, Lithuania Country Study." Background paper prepared for the World Bank. Vilnius. Draft. Databases European System of integrated Social Protection Statistics (ESSPROS). http://europa.eu. int/estatref/info/sdds/en/sespros/esspros_base.htm Mutual Information System on Social Protection (MISSOC), January 1, 2005 version. Database of the European Commission: Directorate-General for Employment, Social Affairs and Equal Opportunities. http://europa.eu.int/comm/employment_social/social_protection/ missoc_tables_en.htm#2005 Eco-Audit Environmental Benefits Statement The World Bank is committed to preserving Endangered Forests and natural resources. We print World Bank Working Papers and Country Studies on 100 percent postconsumer recy- cled paper, processed chlorine free. The World Bank has formally agreed to follow the rec- ommended standards for paper usage set by Green Press Initiative--a nonprofit program supporting publishers in using fiber that is not sourced from Endangered Forests. For more information, visit www.greenpressinitiative.org. In 2006, the printing of these books on recycled paper saved the following: Trees* Solid Waste Water Net Greenhouse Gases Total Energy 203 9,544 73,944 17,498 141 mil. * Pounds Gallons Pounds Co2 Equivalent BTUs 40' in height and 6­8" in diameter Social Assistance in the New EU Member States is part of the World Bank Working Paper series. These papers are pub- lished to communicate the results of the Bank's ongoing research and to stimulate public discussion. This study analyzes the performance of social assistance and family benefit programs in eight new member states of the European Union from the perspective of fiscal impact and effectiveness. It is based on household survey data for six of the countries, as well as budget data and information on program design collected at the national level. The paper finds that, although social assistance programs in the new member states are small in terms of coverage and expenditure levels (reaching 2 to 5 percent of the popula- tion), the programs are an important safety net for the poor. Programs are relatively well targeted, with between 30 and 60 percent of resources going to the poorest quintile of the population. For those who receive them, benefits can make up as much as 37 percent of average consumption of the poor. World Bank Working Papers are available individually or on standing order. Also available online through the World Bank e-Library (www.worldbank.org/elibrary). ISBN 978-0-8213-7159-6 THE WORLD BANK 1818 H Street, NW Washington, DC 20433 USA Telephone: 202 473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org