Report No. 28233-POL POLAND GROWTH, EMPLOYMENT AND LIVING STANDARDS IN PRE-ACCESSION POLAND Background Papers March 22, 2004 Poverty Reduction and Economic Management Unit Europe and Central Asia Region Document of the World Bank 2 CURRENCY AND EQUIVALENT UNITS (Exchange Rate Effective March 19, 2004) Currency Unit = Zloty (PLN) US$1.00 = 3.859 PLN ACRONYMS AND ABBREVIATIONS ALMPs Active Labor Market Policies BAEL Labor Force Survey (Badana Aktwnoci Ekonomicznej Ludnoci) BGD Household Budgets Survey (Badanie Budetow Gospodarstw Domowych) CAL Local Activities Center (Centra Aktywnoci Lokalnej) CASE Center for Social and Economic Research (Centrum Analiz Spoleczno- Ekonomicznych) CEEC Central and Eastern European Countries CIT Corporate Income Tax COICOP Classification of Individual Consumption by Purpose (Klasyfikacja spoycia Indywidualnego Wedlug Celu) COP Central Industrial Area (Centralny Okrg Przemyslowy) CPI Consumer Price Index CSO Civil Society Organization ECA Europe and Central Asia EPL Employment Protection Legislation EU European Union FDI Foreign Direct Investment FDPA Fundation for the Delopment of Polish Agriculture (Fundacja na Rzecz Rozwoju Polskiego Rolnictwa) FTE Full Time Employment Contract GDP Gross Domestic Product GUS Central Statistical Office (Glówny Urzd Statystyczny) HBS Household Budget Survey HDI Human Development Index IALS International Adult Literacy Survey IMF International Monetary Fund IPISS Institute of Labor and Social Studies (Instytut Pracy i Spraw Socjalnych) ISP Institute of Public Affairs (Instytut Spraw Publicznych) KRUS Old Age and Disability Pensions System for Farmers (Kasa Rolniczego Ubezpieczenia Spolecznego) LFS Labor Force Survey LS Living Standard LSA Living Standard Assessment MARD Ministry of Agriculture and Rural Development (Ministerstwo Rolnictwa i Rozwoju Wsi) MGPIPS Ministry of Economy, Labor and Social Policy (Ministerstwo Gospodarki, Pracy i Polityki Spolecznej) MONAR Center for Drug Addicts' Therapy (Centrum Terapii Narkomanów) MPC Monetary Policy Council (Rada Polityki Pieninej) NAS National Account Statistics NBP National Bank of Poland (Narodowy Bank Polski) NGO Non-governmental Organization OECD Organization for Economic Cooperation and Development OLS Ordinary Least Square OMC Open Method of Coordination OPS Social Aid Centers (Orodki Pomocy Spolecznej) PCK Polish Red Cross (Polski Czerwony Krzy) PCPR Poviat Center for Family Assistance (Powiatowy Orodek Pomocy Rodzinie) PD Poverty Deficit PGR State Farm (Pastwowe Gospodarstwo Rolne) PI Price Index PIS Social Intervention Threshold (Próg Interwencji Spolecznej) PISA Program for International Student Assessment PIT Personal Income Tax PLFS Polish Labor Force Survey PPP Purchasing Power Parity REER Real Effective Exchange Rate SAPARD Special Accession Programme for Agriculture and Rural Development SGH Main Trade School (Szkola Glówna Handlowa) SME Small and Medium Enterprises UNDP United Nations Development Program VAT Value Added Tax WB World Bank ZUS Social Security Office (Zaklad Ubezpiecze Spolecznych) Fiscal Year January 1 to December 31 Vice President: Shigeo Katsu Country Director: Roger Grawe Sector Director: Cheryl Gray Sector Leader: Asad Alam Team Leader: Pierella Paci CONTENTS PART I: POVERTY AND GROWTH IN PRE-ACCESSION POLAND ............................1 1. POVERTY IN POLAND: AN OVERVIEW OF EXISTING LITERATURE...............2 A. Limited Development in the Past..................................................................................3 Neglect of Rural Areas and Agriculture ....................................................................3 Regional Disparities...................................................................................................4 Educational Backwardness ........................................................................................4 B. Poverty From the Research Perspective.......................................................................4 C. Poverty Development .....................................................................................................9 Applied Poverty Definitions ...........................................................................9 D. Poverty Trends ....................................................................................................11 Poverty Profile.........................................................................................................13 E. Policy Response.............................................................................................................14 F. Charity and Nongovernmental Organizations ..........................................................16 G. European Union Integration and Combating Poverty..............................................17 H. Conclusions ...................................................................................................................19 References.............................................................................................................................20 2. MACROECONOMIC DEVELOPMENTS OVER THE LAST DECADE: ECONOMIC GROWTH, INCOME DISTRIBUTION AND POVERTY ............................................25 A. A Successful Macroeconomic Performance...............................................................25 The Factors Behind the Trends in Growth...............................................................26 B. Did Poverty Respond to Growth.................................................................................28 A Statistical Artifact?...............................................................................................28 Inequality ................................................................................................................29 Decomposing Trends in Poverty .............................................................................30 C. What is Behind the Increase in Consumption Inequality.........................................33 Changes in the Distribution of Original Income......................................................33 The Role of Taxes and Social Benefits....................................................................34 The Role of Taxes and Social Benefits....................................................................38 D. Policy Recommendations.............................................................................................39 References.............................................................................................................................42 Annex .................................................................................................................................43 3. THE IMPORTANCE OF PRICES IN MEASURING POVERTY, THE CASE OF POLAND, 1994-2001 ..........................................................................................................47 A. Introduction ..................................................................................................................47 Poland's Macroeconomic Situation.........................................................................48 Poverty in Poland.....................................................................................................48 Direction of Research ..............................................................................................48 B. Data and Methodology.................................................................................................49 Household Budget Surveys and Price Statistics ......................................................49 Consumption Aggregate ..........................................................................................50 Poverty Line.............................................................................................................50 C. Accounting for Prices: Spatial Dimension ................................................................51 Empirical Evidence of Price Differences.................................................................51 Theoretical Background...........................................................................................52 Results......................................................................................................................53 D. Accounting for Prices: The Time Dimension............................................................56 Regional Prices through Time .................................................................................56 Accounting for Food Price Inflation........................................................................57 E. Conclusions ...................................................................................................................59 References.............................................................................................................................60 4. POVERTY IN POLAND: PROFILE, 2001 AND CHANGES, 1994-2001...................61 A. Data Source and Methods............................................................................................62 Data Source..............................................................................................................62 Measures of the Living Standard.............................................................................63 Measures of Poverty ................................................................................................68 B. Poverty in 2001 .............................................................................................................69 Overall Picture of Poverty .......................................................................................69 Where Do the Poor Live? ........................................................................................71 What Drives People into Poverty.............................................................................79 Economic Factors: Work Status and Main Source of Income................................80 Demographic Factors: Health, Family Status and Age...........................................83 Social Factors: Gender and Education....................................................................86 Poverty Factors Combined.......................................................................................87 C. Changes in Poverty in 1994 - 2001..............................................................................89 Overall Picture.........................................................................................................89 Changes in the Poverty Profile ................................................................................92 D. Temporal Incidence and Mobility of Poverty............................................................93 Temporal Incidence of Poverty................................................................................93 Poverty Mobility......................................................................................................94 Summary of Panel Analysis.....................................................................................94 Concluding Remarks ..........................................................................................................95 Methodological Annex..........................................................................................................98 PART II: IDENTIFYING PARTICULARLY VULNERABLE GROUPS 5. REGIONAL INEQUALITIES IN LIVING STANDARDS..........................................103 A. Introduction: Historical Heritage of Polish Spatial Patterns................................103 B. The Current State f the Polish Transformation ......................................................104 C. Territorial Patterns in Poland...................................................................................108 Demographic Patterns............................................................................................108 Economic Territorial Differences ..........................................................................110 The Labor Market ..................................................................................................118 D. Regional Dynamics in the Transformation Period..................................................120 Demography and the Settlement System ...............................................................120 Regional Dynamics................................................................................................121 E. Conclusions: The Possible Future Developments...................................................130 References...........................................................................................................................134 6. POVERTY AND RURAL DEVELOPMENT................................................................135 A. Overview .....................................................................................................................135 B. Rural Population in a Regional Context ..................................................................136 C. Main Developments in Agriculture in the 1990s .....................................................137 Production and Employment..................................................................................137 Land Tenure and Productivity ...............................................................................138 D. Trends in Rural Poverty and Unemployment in the Transition Period................140 E. Importance and Structure of Rural Employment...................................................142 F. Structure of Rural Income.........................................................................................145 Summary................................................................................................................148 G. WelfareandInvestmentLinkages: AvailabilityofInfrastructureandServices................148 Education ...............................................................................................................148 Health.....................................................................................................................149 Social Assistance ...................................................................................................150 Physical Infrastructure ...........................................................................................151 H. The Need for Redefining the Goals of Rural Policy ................................................151 References...........................................................................................................................154 Annex ...............................................................................................................................155 7. THE ECONOMIC STATUS OF WOMEN....................................................................160 A. Stereotypical Perceptions of Female and Male Roles..............................................161 B. The Legal and Institutional Framework..................................................................162 C. Business Initiatives Among Women..........................................................................164 D. Barriers and Limitations to Women's Career Paths: The "Glass Ceiling Phenomenon................................................................................................................168 E. Women in the Retirement Pension System ..............................................................170 F. Socioeconomic Situation of Women Living in Rural Areas ...................................171 G. Conclusions .................................................................................................................172 PART III: A CASE OF "JOB-LESS GROWTH" 8. HOW FAR IS WARSAW FROM LISBON ...................................................................175 A. Introduction ................................................................................................................175 B. Employment Performance Over the Last Ten Years: The Key Facts..................176 C. Why is the Distance from Lisbon Increasing...........................................................182 Poland: A Case of Job-less Growth ......................................................................183 D. The Role of Labor Market Institutions ....................................................................185 Employment Protection Legislation ......................................................................186 Real Wage Resistance............................................................................................189 Collective Bargaining and Industrial Relations .....................................................192 Decline of Trade Union Density............................................................................193 E. Conclusions and Policy Implications ........................................................................194 The Labor Market Effects of Partial Deregulation in the Goods Market ..............195 References:..........................................................................................................................196 9. DEALING WITH AN INCREASINGLY STAGNANT UNEPLOYMENT POOL...198 A. Unemployment Dynamics..........................................................................................199 B. Long-Term Unemployment and the Matching Process..........................................200 C. Characteristics of Long-term Unemployment.........................................................202 D. Effects of Duration on Wage Aspirations.................................................................204 E. The Scope for Active Policies.....................................................................................206 F. Conclusions .................................................................................................................211 References:..........................................................................................................................212 10. THEEVOLUTIONOFREGIONALUNEMPLOYMENTINPOLAND,1992-2002..........213 A. Introduction ................................................................................................................213 B. Context: Poland's Macroeconomy and Labor Market..........................................215 Macroeconomic Context........................................................................................215 Changes in Labor Institutions................................................................................218 C. Unemployment and Structural Change Across Poland's Regions.........................220 The Pattern of Regional Unemployment over the Period......................................220 The Inverse Relationship and the Big Shift in Skills.............................................221 Regional Variations in Skill and Unemployment ..................................................225 D. Theoretical Issues .......................................................................................................225 Regional Natural Rate Differences ........................................................................226 Regional Mismatch................................................................................................228 Aggregate Labor Demand and Endogenous Populations ......................................229 E. Econometric Results...................................................................................................230 F. Conclusions .................................................................................................................233 References:..........................................................................................................................235 Annex: ...............................................................................................................................237 11. THE DISTRIBUTION OF WAGES IN POLAND, 1992-2002.....................................241 A. Background.................................................................................................................241 B. Macroeconomic and Institutional Context...............................................................241 Macroeconomic Overview of the Labor Market....................................................241 Changes in Labor Institutions................................................................................242 C. PLFS DATA................................................................................................................244 D. Wage Equation Estimation........................................................................................248 The Evolution of Pay Determination over the Transition Period ..........................248 Disaggregating by Gender .....................................................................................253 Quantile Regression Analysis................................................................................254 Controlling for Participation Bias..........................................................................255 What Generates Low Pay?.....................................................................................259 E. Conclusions .................................................................................................................259 References:..........................................................................................................................261 Annex 1...............................................................................................................................263 Annex 2...............................................................................................................................264 PART IV: VULNERABILITY, SAFETY NETS AND INCENTIVES.............................265 12. INFORMAL NETWORKS OF SUPPORT OF POOR PEOPLE IN POLAND.........266 A. Introduction ................................................................................................................266 B. Informal Networks of Support of the City Poor (Lód and Katowice) .................267 Potential Networks of Support...............................................................................267 Perceiving a Network as a Support Group.............................................................268 The Need for Support and the Received Support ..................................................269 The Family as the Basis Group of Real Support....................................................270 C. Informal Support Networks of Poor People from Post-State Farm (PGR) Areas and Small Towns .............................................................................................272 D. Final Remarks.............................................................................................................275 E. Conclusions .................................................................................................................276 References:..........................................................................................................................277 Annex 1...............................................................................................................................280 Annex 2...............................................................................................................................282 13. CIVIL INVOLVEMENT AND CIVIL SOCIETY ORGANIZATIONS AS INFORMAL SAFETY NETS .................................................................................................................288 A. Civil Society Organizations (CSOs) in Poland.........................................................288 B. Civil Activity Development........................................................................................290 Civil Organizations for the Prevention of Poverty and Social Exclusion: Number of Organizations, Scope of Activities and Structure................................292 Funding..................................................................................................................295 Cooperation with Other Entities ............................................................................298 Review of the Activities Undertaken by Organizations for Poverty and Social Exclusion Prevention .............................................................................................301 Beneficiaries ..........................................................................................................303 C. Conclusions and Recommendations .........................................................................305 References:..........................................................................................................................308 14. SOCIAL BENEFITS AS ALTERNATIVE TO WAGE INCOME IN POLAND .......310 A. Benefits for the Unemployed .....................................................................................311 Financing of Unemployed Policies ..............................................................................311 Unemployment Benefits...............................................................................................312 Pre-retirement Transfers...............................................................................................314 B. Social Assistance and Income Support for the Poorest...........................................316 Social Assistance for the Unemployed.........................................................................317 Selected Information on the Provision of Social Assistance........................................318 Projected Changes in Social Assistance and Family Benefits......................................320 C. Social Security System Disability Pensions.............................................................321 Institutional Framework ...............................................................................................321 Eligibility Criteria.........................................................................................................321 Basic Information about Disability Pensions ...............................................................322 D. The Role of Social Benefits in Replacing Wage Income..........................................325 Assumptions for the Simulation...................................................................................325 Results of the Simulation .............................................................................................326 E. Income of Individuals Versus Their Labor Market Situation ...............................328 Sources of Income of the Unemployed ........................................................................329 Changes in the Labor Market .......................................................................................330 F. Conclusions .................................................................................................................334 References:..........................................................................................................................336 Annex 1...............................................................................................................................337 Annex 2...............................................................................................................................342 Statistical Appendix ................................................................................................................344 Tables Table 1.1 Comparison 1. Selected Quantitative Analyses of Poverty.................................6 Table 1.2 Selected Qualitative Research on Poverty in Poland ..........................................8 Table 1.3 The Rate of Unemployment Among Groups Vulnerable to Unemployment....13 Table 1.4 Selected Social Exclusion Indicators in the EU, Average for 15 Member States .................................................................................................................18 Table 2.1 Selected Economic Indicators, 1997-2003........................................................26 Table 2.2 Between Group Inequality as Percent of Total Inequality, 1994-2002 .............30 Table 2.3 Contribution of Various Inc. Sources to Total Inequality, 1994-2002..............30 Table 2.4 Pro-poor Growth ...............................................................................................33 Table 2.5 Income Inequality and Decomposition by Income Sources, 1994-2002 ..........34 Table 2.6 The Role of Taxes and Social Benefits in Inequality Trends, 1994-2002 ........35 Table A.2.1 Distributional Characteristics, 1994-2002.........................................................45 Table 3.1 Results for the Poverty Rate..............................................................................54 Table 3.2 Estimates of Relative Poverty Risks for Various Groups..................................56 Table 4.1 Consumption Aggregates ..................................................................................66 Table 4.2 Imputed Consumption by QuartileGroups and Place of Residence, 2001 ...................66 Table 4.3 Poverty Measures (indices) ...............................................................................69 Table 4.4 Poverty According to Various LS*Concepts: Selected Indices ........................70 Table 4.5 Voivodship Ranking by Poverty Rate...............................................................70 Table 4.6 Poverty of Post-PGR Population.......................................................................76 Table 4.7 Poverty in Rural and Urban Areas ....................................................................77 Table 4.8 Work Status and Poverty...................................................................................80 Table 4.9 Poverty by Unemployment Duration ................................................................83 Table 4.10 Health, Age, Family Status and Poverty ...........................................................84 Table 4.11 Gender, Family, Size, and Poverty ...................................................................86 Table 4.12 Education and Poverty.......................................................................................87 Table 4.13 Logit Models for Medium and Hard Poverty....................................................88 Table 4.14 Poverty in 1994-2001: An Overview ...............................................................90 Table 4.15 Poverty Temporal Incidence, 1994-2000, Number of Years in Poverty...........94 Table 4.16 Poverty Mobility, 1994-2000, Movements In and Out of Poverty....................95 Table A.4.1 Legal Title to Dwelling, Poland 2001 ...............................................................98 Table A.4.2 Utilities and Rents Paid by Co-op Members, Poland 2001...............................98 Table A.4.3 Potential Rental Payments' Explanatory Factors ..............................................99 Table A.4.4 Estimation Results, Poland 2001.....................................................................100 Table A.4.5 Durable Consumption Imputation Values, Poland 2001.................................101 Table A.4.6 Average Number of Durables per 100 Households, Poland 2001...................102 Table 5.1 Dynamics of Basic Economic Categories, Poland, 1989-2001.......................104 Table 5.2 Dynamics of GDP in Post-Socialist Countries, 1989-2001.............................105 Table 5.3 Regional Differences in GDP per Inhabitant in Selected European Countries, 1998 ...............................................................................................111 Table 5.4 GDP and Personal Incomes per Inhabitant, 1999 (extreme values for 16 voivodships, current prices in Zlotys ..............................................................111 Table 5.5 Gross Value Added per Employee, 1999 (current prices, in Zlotys) ..............111 Table 5.6 Shares of Leading Four Academic Centers in Number of Publications and Citations in the Total Number Noted in the "Philadelphia List," in the Total Number of Publications and Citations of the Respective Type of Polish Academic Establishments, 1999......................................................................114 Table 5.7 Domestic Migrations in Poland, 1971-2001....................................................121 Table 5.8 Regional Differentiation of GDP per Inhabitant in Selected Regions of Poland=100......................................................................................................122 Table 6.1 Unemployment Rates in Urban and Rural Areas, 1997-2003.........................141 Table 6.2 Rural Shares of Consumption Quintiles and Poverty Head Count, 1994-2001........................................................................................................141 Table 6.3 Share of Food Consumption Own Produced, by Quintiles,*Rural and Urban........................................................................................................148 Table 6.4 Educational Level Reached by Shares of the Population (of those aged 15 and above) .......................................................................................................149 Table 6.5 Distribution of Household Expenditures for Health Services by Income Quintiles ..........................................................................................................150 Table A.6.1 Comparative Rural and Agricultural Employment and Productivity Characteristics by Region................................................................................155 Table A.6.2 Income Sources for Rural and Urban Households ­ Shares, by Quintiles......155 Table 7.1 Employment Rate in Poland in the Years 1995-2002 .....................................162 Table 7.2 Average Gross Remuneration of Full-Time Employees According to Ownership Sectors, Gender, and Level of Education, October 2001..............162 Table 7.3 Employers and the Self-employed Outside of Individual Farming, by Gender, 1985 and 1998 (in thousands) ...........................................................165 Table 7.4 Women in the Total Population of the Self-employed in Poland and Selected EU Countries (average data for years 1990-97) ...............................165 Table 7.5 Size of Future Retirement Benefit for Women and Men Depending on Retirement Age, in Relation to Average Benefit Size (assuming that both the first and the second pillar use universal lifespan expectancy charts) ........170 Table 8.1 Average Employment by Type of Contract, Ownership Status and Employment Status, 1992-2002 ......................................................................177 Table 8.2 Contribution to Employment Growth by Ownership Structure, Type of Contract and Employment Status, 1998-2002.................................................179 Table 8.3 Employment Rate by Age and Gender, 1992-2001.........................................180 Table 8.4 Employment Structure and Employment Growth by Broad Economic Sector...............................................................................................................181 Table 8.5 Contribution to Employment Growth by Sectors, 1995-2002.........................181 Table 8.6 Structure of Employment and Employment Growth by Educational Attainment: 1992-2002...................................................................................182 Table 8.7 Employment Protection Legislation in Poland, Czech Republic and Hungary ....................................................................................................187 Table 8.8 Collective Dismissal in Poland, the Czech Republic and Hungary.................188 Table 8.9 Nominal and Real Wage Growth Across Sector: 1997-2000..........................189 Table 9.1 Matching and Unemployment Duration..........................................................202 Table 9.2 The Changing Profile of Polish Long-term Unemployment, 1992-2002........203 Table 9.3 Estimates of the Reservation Wage of Polish Unemployed Workers, 2002 ...205 Table 10.1 Unemployment in the New Voivodships, 1995, 2001, 2002 ..........................220 Table 10.2 The Big Shift in Skills: Changes in the Distributions of Employment, Population and Unemployment by Level of Completed Education, 1992-2002........................................................................................................222 Table 10.3 The Inverse Relationship, 1998 and 2002.......................................................223 Table 10.4A Labor Market Attachment by Completed Education Level, Spring 2002.......223 Table 10.4B Average Uncompleted Unemployment Duration by Educational Attainment, 1998 and 2002 .................................................................................................223 Table 10.5 The Structure of Employment by Major Industrial Sector and Education, 1992, 1998 and 2002 .......................................................................................224 Table 10.6 Voivodship Variations in Changes in Unemployment, Industry and Skill, Percentage Point Changes, 1994-98................................................................225 Table 10.7 The Slowdown of Migration, Gross Migration Flows (`000) .........................227 Table 10.8 Estimates of the Influence of Structural Supply and Demand Features on Voivodship Unemployment Rates, Annual Panel of 49 Voivodships, 1994-98............................................................................................................231 Table 10.9 Estimates of the Influence of Structural Supply and Demand Features on Voivodship Unemployment Rates, Half-yearly Panel of 16 New Voivodships, Autumn 1999-Spring 2002........................................................232 Table 10.10 Probit Analysis of Unemployment Status, November 1998 ...........................233 Table A.10.1 Macroeconomic Data ......................................................................................237 Table A.10.2 Polish Labor Market........................................................................................240 Table 11.1 Monthly Wage Distribution, 1994-2002 .........................................................245 Table 11.2 Summary Statistics for the Population of Working Age, 1992-2002..............245 Table 11.3 Descriptive Statistics for Employees, 1994-2002 ...........................................246 Table 11.4 OLS Modeling of (log) Monthly Earnings in the Private Sector, 1994 -2002 249 Table 11.5 OLS Modeling of (log) Monthly Earnings in the State Sector, 1994 - 2002 ..251 Table 11.6 Quantile Wage Regressions for Spring 2002 ..................................................254 Table 11.7 ML Heckman Wage Equation Estimates, 1994, 1995, 2001, 2002.................255 Table 11.8 Accounting for the Gap between Wages in the Lowest Decile and the Mean Wage, Full-time workers, Autumn 1994 and Spring 2002 .............................259 Table A.11.1 Estimated Wage Impact of Professional Status (relative to skilled manuals), 1994-2002........................................................................................................263 Table A.11.2 Estimated Wage Impact of Tenure over 10 Years (relative to tenure of 1 year or less), 1994-2002...............................................................................263 Table A.11.3 Estimated Wage Impact of Experience over 20 Years (relative to less than 5 years experience), 1994-2002..........................................................................263 Table 13.1 NGOs Operating in the Area of Social Aid with FTEs...................................294 Table 13.2 Social and Professional Status of the Volunteers............................................294 Table 13.3 Volunteers in NGOs........................................................................................295 Table 13.4 Income of NGOs Operating in the Area of Social Aid ...................................296 Table 13.5 Sources of Income for NGOs..........................................................................297 Table 13.6 Frequency of Contacts with Other Entities .....................................................299 Table 13.7 Assessment of Cooperation with Various Institutions ....................................300 Table 13.8 Forms of Actions Undertaken by Organizations.............................................302 Table 14.1 Individuals Entitled to Pre-retirement Benefits and Allowances in 1997-2002 (end of year)..................................................................................315 Table 14.2 Social Assistance Recipients in Crisis Situations, 1999-2002 ........................319 Table 14.3 Most Frequently Granted Benefits to Families with Unemployed Applying for Social Assistance, 2000-2002....................................................................320 Table 14.4 Income Test (Per Family) for Benefits by Household Type in 2002 ..............326 Table 14.5 Income Substitution Ratio by Family Type and Income Level.......................327 Table 14.6 Change in Labor Market Status After One Year Based on LFS, 2002 ...........331 Table 14.7 The Structure of Persons Outside the Labor Market by Source of Income, Age, Gender, and Labor Market Status in the Previous Year .........................333 Table A1.14.1 Replacing the Average Income of a Single Person with Social Transfers (in PLN), 2002.................................................................................................337 Table A1.14.2 Replacing the Minimum Income of a Single Person with Social Transfers (in PLN), 2002.................................................................................................338 Table A1.14.3 Replacing the Average Income of a Four-person Family with Social Transfers (in PLN) 2002..................................................................................338 Table A1.14.4 Replacing the Minimum Income of a Four-person Family with Social Transfers (in PLN) 2002..................................................................................339 Table A1.14.5 Replacing the Average Income of a Family with Four Children with Social Transfers (in PLN) 2002..................................................................................339 Table A1.14.6 Replacing the Minimum Income of A Family with Four Children with Social Transfers (in PLN) 2002..................................................................................340 Table A1.14.7 Replacing the Average Income of a Single Parent with Two Children with Social Transfers (in PLN) 2002) .....................................................................340 Table A1.14.8 Replacing the Minimum Income of A Single Parent with Two Children with Social Transfers (in PLN) 2002.......................................................................341 Table A2.14.1 Supporting Oneself from Non-earning Sources of Income (logit estimates) ..343 Table A2.14.2 Supporting Oneself from a Disability Pension (logit estimates).....................343 Figures Figure 1.1 Poland: Poverty Rates in 1994-2001, According to Different Measures ..........12 Figure 1.2 Poland: Poverty Rates in 2001 by Family Type ...............................................12 Figure 2.1 Growth, Employment and Poverty....................................................................26 Figure 2.2 Fiscal and Monetary Developments..................................................................27 Figure 2.3 GDP Growth, Export, and Industrial Performance ...........................................28 Figure 2.4 Inequality, 1994-2001 .......................................................................................29 Figure 2.5 Decomposition of Changes in Poverty, 1994-2001...........................................31 Figure 2.6 Composition & Growth in Mean Consumption by Percentiles/Quintiles, 1995-2002..........................................................................................................31 Figure 2.7 Growth Incidence Curve ...................................................................................32 Figure 2.8 Progressivity of Tax and Benefit System in 1998 ............................................35 Figure 2.9 Comparison of Income Composition: by Decile in 2002 (left) and over 1994- 2002 for the Poor (right)....................................................................................36 Figure 2.10 Poverty and Social Transfers, 1994-2002 .........................................................37 Figure 2.11 Targeting of Social Transfers (excluding old age pensions).............................37 Figure 2.12 Composition of Food and Non-food Expenditures and CPI and Food Price Developments, 1998-2002.................................................................................39 Figure 2.13 Composition of the Expenditures of Tradables versus Non- tradables and REER Development, 1998-2001 ................................................................39 Figure A.2.1 Private Consumption Growth, Discrepancies between NSA and HBS, 1995-2001..........................................................................................................43 Figure A.2.2 Distribution of Consumption and Log Consumption, 2002 ..............................45 Figure 3.1 GDP Growth and Poverty Rates, 1994-2001 ....................................................47 Figure 3.2 Cumulative Distribution of Consumption .........................................................51 Figure 3.3 Relative Prices: by Voivodships (left) and by Town Size (right)......................52 Figure 3.4 Poverty Rates vs. Regional Price (2001)...........................................................52 Figure 3.5 Distribution of a Household Specific Price Index.............................................54 Figure 3.6 Impact of Regional Prices on the Distribution of Consumption .......................54 Figure 3.7 Regional Prices Correction to Poverty Depth ...................................................55 Figure 3.8 RegionalPrices,1994versus2001(left)andtheStandardDeviation(right)..............................56 Figure 3.9 Food Share in Consumption by Half-decile......................................................57 Figure 3.10 "PoorPI"versusCPI(left)andFoodPriceInflationversusCPIInflation(right).........................58 Figure 3.11 Properly Deflated Social Assistance Threshold versus Actual .........................58 Figure 3.12 Poverty in Poland (adjusted) .............................................................................58 Figure 4.1 Medium Poverty by Place of Residence, Poland 2001......................................79 Figure 4.2 Hard Poverty by Place of Residence, Poland 2001...........................................79 Figure 4.3 Medium and Hard Poverty (WB2) by Age, Poland 2001 .................................85 Figure 4.4 Medium Poverty, Poland 1994-2001.................................................................90 Figure 4.5 Hard Poverty, Poland 1994-2001......................................................................91 Figure 4.6 Unemployment Rate, Poland 1994-2001 ..........................................................92 Figure 5.1 GDP Dynamics in Central and Eastern Europe, 1989-2001 ...........................106 Figure 5.2 Different Restructuring Trajectories, J-curve Patterns....................................106 Figure 5.3 Population in Pre-productive Age (0-17) in Polish Municipalities, 2001.......108 Figure 5.4 Population in Productive Age (men: 18-65, women: 18-60) in Polish Municipalities, 2001........................................................................................109 Figure 5.5 Population in Post-productive Age (men:65+, women:60+) in Polish Municipalities, 2001........................................................................................109 Figure 5.6 GDP per capita, NUTS 3, Poland=100, 2000..................................................110 Figure 5.7 Share of Working Population in Agriculture, Forestry, and Fishery in the Total Number of Working Population, 2000, by Powiats...............................112 Figure 5.8 Share of Working Population in Industry in the Total Number of Working Population, 2000, by Powiats..........................................................................113 Figure 5.9 Share of Working Population in Services in the Total Number of Working Population, 2000, by Powiats..........................................................................113 Figure 5.10 Quintiles of Polish Municipalities according to the Value of Own Revenues and Shares in State Taxes (PIT and CIT) in Zlotys per Inhabitant, 2001........115 Figure 5.11 Lowest (blue) and Highest (red) Quintiles of Polish municipalities according to the Value of Own Revenues and Shares in State Taxes (PIT and CIT) per Inhabitant, 2001...............................................................................................116 Figure 5.12 Density of Companies with Foreign Capital in Polish Municipalities, 2001 .117 Figure 5.13 Shares of Votes Cast for T. Mazowiecki in the First Round of Presidential Elections, November 1990 ..............................................................................118 Figure 5.14 Unemployment Rates (registered unemployment) in Powiats, December 2002 ...............................................................................................119 Figure 5.15 Migration Balance in Selected Polish Metropolitan Regions, 1994-1999 (in per cent) .....................................................................................................121 Figure 5.16 Growth of GDP in NUTS 3, in percent, constant prices (national deflator) ...124 Figure 5.17 Changes in the Lowest Quartiles of the Arrangement of Municipalities, on the Scales of Own Revenues and Shares in PIT and CIT per Inhabitant, 1998-2001........................................................................................................126 Figure 5.18 Dynamics of Own Revenues and Shares in State Taxes in Relation to the Average Change in the Whole Country among the Poorest Municipalities, 1998-2001 (in nominal terms).........................................................................127 Figure 5.19 Increase in Unemployment Rate 1998-2002, in Percentage Points (in gray- powiats not existing in 1998) ..........................................................................128 Figure 5.20 Relation of Rate of Unemployment in 1998 and its Growth in the Period 1998-2002........................................................................................................129 Figure 5.21 Typology of Powiats According to the Relation of Rate of Unemployment in 1998 and its Growth 1998-2002..................................................................129 Figure 6.1 Share of Population in Rural Areas in CEE Countries (2000-01)...................137 Figure 6.2 Trends in Agricultural ValueAdded and Employment, 1995-2000 ..........................138 Figure 6.3 Distribution of Ownership of Agricultural Land ............................................139 Figure 6.4 Rural Poverty Head Count by Quintilesof Agric. Land, 1994-2001 .........................142 Figure 6.5 Share of Rural Off-Farm and Rural Agricultural Employment in Total Regional Employment, 2001 .................................................................143 Figure 6.6 Trends in Rural Non-Agricultural Employment and Total Agricultural Employment, 1995-2000.................................................................................143 Figure 6.7 Distribution of Population and Employment Type by Degree of Urban Concentration ..................................................................................................144 Figure 6.8 Sources of Average Monthly Per Capita Income of Households, by Region in 2001, PLN ...................................................................................................146 Figure 6.9 Value of Monthly Consumption, PLN Per Capita...........................................146 Figure 6.10 Rural Shares of Employment and Earned Income by Employment Type, 2001.................................................................................................................147 Figure 6.11 Private Household Expenditures on Hospitals and Sanatoria per Adult .........150 Figure 6.12 Total Area (km2) and Infrastructure (km) .......................................................151 Figure A.6.1 Average Monthly Wages, PLN .......................................................................156 Figure A.6.2 Number of Enterprises, Total and with Foreign Capital, and Persons Conducting Economic Activity.......................................................................156 Figure A.6.3 Number of Rural Enterprises per 1000 Population .........................................157 Figure A.6.4 Enterprise Assets, Investments, and Gmina and Powiat Revenues.................157 Figure A.6.5 Rural Enterprise Assets and Investments per Off: Farm Employee, PLN...................158 Figure A.6.6 Tourist Accommodations Provided per1000 Population ................................158 Figure A.6.7 Doctors per 10,000 Population........................................................................159 Figure A.6.8 Employment Shares by Level of Urbanization ...............................................159 Figure 7.1 Employment Rate in Poland in the Years 1995-2002 .....................................163 Figure 8.1 The Dynamics of Total, Private and Public Employment...............................177 Figure 8.2 Polish Distance from Lisbon Target, 1992-2001 ............................................179 Figure 8.3 Employment Reallocation across Broad Economic Sectors, 1994-2002 ........180 Figure 8.4 Quarterly Growth Rate in Industrial Production and Unemployment Rate, 1992-2002........................................................................................................183 Figure 8.5 GDP and Employment Growth in Poland and EMU Countries, 1993-2002...184 Figure 8.6 Productivity Growth in Poland, 1993-2002.....................................................185 Figure 8.7 Wage Growth and Productivity Growth across Sectors..................................191 Figure 9.1 The Evolution of Polish Unemployment since the Inception of Transition, 1990-2002........................................................................................................199 Figure 9.2 Monthly Outflows to Jobs as a % of Unemployment, 1991-2002 ..................200 Figure 9.3 Share of Long-Term Unemployment and UB Coverage.................................208 Figure 9.4 A ALMPS versus Unemployment Rates, 2002...................................................209 Figure 9.4 B ALMPS versus Vacancy Rates, 2002..............................................................209 Figure 9.4 C ALMPS versus LTU, 2002..............................................................................210 Figure 9.4 D ALMPs versus Unskilled Unemployment, 2002.............................................210 Figure 10.1A Unemployment rates (%) in the old voivodships, 1992 and 1998...................221 Figure 10.1B Unemployment rates (%) in the new voivodships, 1995 and 2002.................221 Figure 10.2 Skill Mismatch with Rigid Wages...................................................................229 Figure A.10.1 Poland's Phillips Curve...................................................................................238 Figure A.10.2 Poland's Beveridge Curve...............................................................................239 Figure 14.1 Structure of Labor Fund Expenses, 1999-2003...............................................312 Figure 14.2 Percentage of Unemployed Eligible for the Unemployment Benefit in Various Categories, 1996-2002.....................................................................................313 Figure 14.3 Number of Disability Pensioners, 1990-2002 .................................................323 Figure 14.4 Number of Newly Granted Benefits, 1990-2002 ............................................323 Figure 14.5 Structure of Newly Granted Disability Pensions by Age in 2002...................324 Figure 14.6 Structure of Disability Pensioners by Age in 2002 .........................................324 Figure 14.7 Main Sources of Income of People Unemployed at the Time of the Survey and Unemployed 12 Months before this Date.................................................330 Boxes Box 5.1 Specialization of Temporary Migrations............................................................119 Box 5.2 "Soft" Benefits and Motivation for Work ..........................................................119 Box 5.3 Trade and Border Crossings...............................................................................123 Box 7.1 Key Factors Motivating People to Start a Business ...........................................166 Box 10.1 Summary of Some Relevant Previous Studies...................................................214 Box 14.1 Institutional Setting for Unemployment Programs.............................................311 Box 14.2 Eligibility Criteria for Pre-retirement Benefits...................................................315 Box 14.3 Institutional Framework of social Assistance.....................................................317 Box 14.4 The Need for Further Changes in the Disability Pension System......................325 Maps Map 4.1 Medium Poverty 2001 (CSO)...............................................................................72 Map 4.2 Medium Poverty 2001 (CSO reg) ........................................................................73 Map 4.3 Medium Poverty 2001 (WBO).............................................................................74 Map 4.4 Medium Poverty 2001 (WB2)..............................................................................75 CONTRIBUTORS Tito Boeri Professor of Economics, Bocconi University, Milan, Italy, and Director, field of Labor Market and Social Policy Reforms, Fondazione Rodolfo Debenedetti, Milan, Italy. Boguslawa Budrowska Research Unit on Women and Family of the Institute of Philosophy and Sociology of Polish Academy of Sciences, Warsaw, Poland. Agnieszka Chlon-Dominczak Director of the Department for Economic Analyses and Forecasting at the Ministry of Economy, Labor and Social Policy, Warsaw, Poland. Edwarda Dabrowska Head of the Unit of Social Protection, Department for Economic Analyses and Forecasting, Ministry of Economy, Labor and Social Policy, Warsaw, Poland. Pietro Garibaldi Associate Professor of Economics, Bocconi University, Milan, Italy, and Head of Labor Studies for the Fondazione Debenedetti, Milan, Italy. Piotr Gliski Professor of Sociology, Head of Civil Society Unit, Institute of Philosophy and Sociology, Polish Academy of Sciences, Warsaw; Head of Social Structure Unit, University of Bialystok, Poland. Katarzyna Górniak Assistant Lecturer, Institute of Social Sciences and Administration, Warsaw University of Technology, Warsaw, Poland; Member (professional bodies), Polish Sociological Association. Grzegorz Gorzelak Professor of Economics and Director of the Centre for European Regional and Local Studies (EUROREG), Warsaw University, Poland. Stanislawa Golinowska Professor in Economics and lecturer, Jagiellonian University, Poland. Jolanta Grotowska-Leder Associate Professor, Department of Sociology, University of Lodz, Poland. Alexandre Kolev Economist, Poverty Reduction and Economic Management Sector, Europe and Central Asia Region, The World Bank, Washington, D.C., United States. Karol Kuhl Teaching Assistant (Mathematical Statistics and Econometrics); and PhD student, Department of Economics, Warsaw University, Poland. Ewa Lisowska Associate Professor, Women's Entrepreneurship, Warsaw School of Economics, Warsaw, Poland. Mark Lundell Lead Agricultural Economist, Environmentally & Socially Sustainable Development Sector Unit, Europe and Central Asia Region, The World Bank, Washington, D.C., United States. Mauro Maggioni Junior Researcher, Fondazione Rodolfo Debenedetti, Milan, Italy. Eugenia Mandal Associate Professor, Institute of Psychology, University in Katowice, Poland. Andrew Newell Professor, University of Sussex, Brighton, U.K. and Research Fellow at IZA (Institute for the Study of Labor), Bonn, Germany. Hanna Palska Associate Professor, Civil Society Unit, Institute of Philosophy and Sociology, Polish Academy of Sciences, Warsaw, Poland. Marcin Sasin Economist, Poland Country Office, Europe and Central Asia Region, The World Bank Office in Warsaw, Poland. Janina Sawicka Professor, Agricultural Economics and Agrarian Policy, Warsaw Agricultural University, Warsaw, Poland. Mieczyslaw W. Socha Professor, Department of Economics, Warsaw University, Poland. Piotr Stronkowski Head of Labor Market Analyses Section, Department for Economic Analyses and Forecasting, Ministry of Economy, Labor and Social Policy, Warsaw, Poland. Irena Topiska Associate Professor, Department of Economics, Warsaw University, Poland. Katarzyna Tyman-Koc Economic Advisor to the Plenipotentiary for Equal Status of Women, Government of Poland. Jos Verbeek Lead Economist, Poverty Reduction & Economic Management 1, Africa Technical Family, The World Bank, Washington, D.C., United States. Wielislawa Warzywoda-Kruszyska Professor, Department of Sociology, University of Lodz, Poland. PART I: POVERTY AND GROWTH IN PRE-ACCESSION POLAND 1 1. POVERTY IN POLAND: AN OVERVIEW OF EXISTING LITERATURE Stanislawa Golinowska 1.1 Poverty is a good indicator of the social development accompanying the contemporary modernization process in any country. The social development of recent decades indicates that impoverishment and the growth of poverty are connected to two main processes. One process is the exposure of the agricultural population to competition on the part of an aggressive market, while the other process is the economic collapse of small farmers whose existence was based on an almost purely subsistence economy. 1.2 The second process is connected to changes involving industrialization and urbanization. What is taking place is the downfall of traditional industry. The collapse of traditional branches of industry, the mass restructuring of many industries, and the subsequent difficulties in adaptation on the part of numerous worker groups all constitute yet another source of the intensification of poverty--this time of urban poverty (Mingione 1996). The ongoing process of change in industry is smoothed by nonpartisan social programs for workers leaving their traditional workplaces. The poverty of these groups is not directly connected with reductions, but, rather, with later adaptation difficulties. 1.3 There is a third process visible in transforming countries. It is linked to the transformation of a centrally planned economy into a market economy, and with the shift from totalitarianism to democracy. The entrance of freedom into the main spheres of social life brought with it obvious material benefits to dynamic individuals with a modern education who belong to better organized groups. However, for groups outside of the main current of modernization, the fall in social security (radical on the part of the employer and gradual in the case of the state) often leads to helplessness and poverty on an unprecedented scale (as shown in the World Bank report for the year 2000). No public­private forms of social support that might alleviate the pauperization and frustration have coalesced with sufficient speed and scope to take the place of state institutions and workplace social security. The resources and infrastructure, the appropriately qualified staff, and the models of effective action are lacking. Moreover, poverty becomes a "normal" phenomenon as time passes; it then continues without receiving attention and without attracting efforts at special policy action. 1.4 Countries undergoing transformation are subjected to all three processes simultaneously over this relatively short span of time. This results in a political reaction that is neither comprehensive nor of an appropriate scale. It tends to be limited and only has priorities with a short term justification. Ultimately, inequality and social exclusion are growing. 1.5 The present study looks at solutions to the causes, nature, and tendencies of poverty in Poland as well as ways to deal with the problem through the primary actors on the public stage, 2 non-governmental organizations and, finally, the European Union (EU). The basis for this effort encompasses research conducted in Poland over the past few years. 1.6 The study is divided into six sections (parts). In the first, historical part, the study explains how limited modernization in the past influenced the contemporary underdevelopment evident in the economic structure (neglect of agriculture) and in societal inequalities (educational backwardness of agricultural and industrial workers), and visible in the regional disparities. The second section reviews the results of the qualitative and quantitative types of research conducted in and about Poland with a focus on poverty. In the third section, the use of numerous poverty lines applied in the Polish statistics enables the study to present trends in the development of poverty in the 1990s as well as poverty profiles. The fourth section covers the public institutions and programs that are combating poverty, and also assesses policy responses to the poverty phenomenon. A brief fifth section elaborates on the contrary response from the society side. There is a short explanation of the limited development of non-governmental organizations in Poland. The sixth and final section discusses the EU social policy that is oriented towards increasing employment and combating poverty and social exclusion (promoting social inclusion); also discussed is the potential influence of EU policy in that area as regards the development of Polish social policy. A. LIMITED DEVELOPMENT IN THE PAST 1.7 During the period of real socialism Poland failed to attain a level of social development equal to that of Western countries, despite intense industrialization programs and many modernization-oriented ventures. The reasons for this situation were the low efficiency of implemented plans as well as the constant lack of capital in spite of the maintenance of a low level of consumption and the large scale income redistribution. Under such conditions, society's motivation to modernize was poor and was also continuously subjected to further lowering in the face of limitations on liberty and democracy. 1.8 The negligence resulting from such "limited modernization" is mainly visible in three spheres: (i) the disparity between living conditions in the countryside and those in the city; (ii) the strong regional disparities; and (iii) the low educational level and limited capacity to adapt on the part of population groups previously prepared for the efforts of accelerated industrialization but now dismissed due to restructuring processes. Neglect of Rural Areas and Agriculture 1.9 Over the three decades of accelerated industrialization (1950-1980) the private economy of farmers was the source of accumulation and labor resources for industrial growth. Exploitative attitudes towards the farm economy started to change as of the mid­1960s. However, a significant improvement in the rural population's living standard did not appear until the 1970s when the communist authorities changed their policy towards agriculture and the rural areas. Initially, the change consisted of the removal of existing barriers (primarily the elimination of production quotas), but by the second half of the decade input for agriculture was also increased. An increase in the prices of agricultural produce sold to the state was allowed (primarily meat products) and simultaneously, credit policies advantageous for farmers were introduced, as well as subsidized feed and synthetic fertilizers prices. In addition, the year 1977 saw the introduction of a social security scheme for farmers similar to the one already in place for workers; previously, farmers could receive retirement pensions only by bequeathing their farms to the state. 3 1.10 The change in policy towards agriculture, including individual farm agriculture, led to a situation in which, in the 1980s, farmers achieved, on average, identical incomes to those of workers, and even higher at the end of the decade. However, the favorable change in income in agriculture in the 1980s did not bring about any significant improvement in the land subdivision structure, and technical infrastructure in the countryside improved only to a small degree, despite private initiatives (activities of the Church Fund and subsequently the Rural Support Fund). Regional Disparities 1.11 Poland is characterized by a relatively high degree of regional diversification. This is the result of the divergent economic development of various parts of the country, which had developed under different historical conditions and different dynamics of industrialization processes. Contemporary Poland still has traces of the policies pursued by the partitioning powers in the nineteenth century--a time when Poland was divided among Prussia, Austria, and Russia. 1.12 The transformation of the economy has given birth to a new territorial diversity that is connected to the decline of agricultural production and especially the collapse of collective (state) farms. Unemployment indicators illustrate these new differences. The problems of poor regional development were transferred from the eastern part of the country to the northern and northwestern regions. However, poverty indicators also show the traditional regional diversity, which is connected to industrial underdevelopment (in what is known as Poland's "eastern wall"). Educational Backwardness 1.13 The legacy of the educational policy of socialist Poland is the dominance of basic vocational education. The accelerated industrialization of the country that lasted three decades (1949­1978), with an exceptionally intense phase in the first half of the 1950s, demanded a mass increase in the number of highly qualified workers. The model of the basic vocational school and its mass growth was the answer to this demand. Accordingly, the middle-aged working population of today is dominated by people who have received only a basic vocational education (over 50 percent) (Cichomski 1998). The material situation of this group was relatively good. Representatives of the worker avant-garde and members of trade unions came from this group. Many of them entered the elite of government authorities in the 1980s and 1990s. Similarly, most of them became proprietors of small and medium-size private businesses that were created on a mass scale in the years 1989­92 (Domaski 1994). 1.14 The politicization of this group, the incomplete acceptance of the market system (especially of its efficiency requirements), as well as a lack of qualification adjustments to meet the present situation on the labor market, currently constitute a noticeable barrier to development. B. POVERTY FROM THE RESEARCH PERSPECTIVE 1.15 The consequences of the limited modernization efforts of real socialism are the still present diversity of living conditions among regions and between major cities and rural areas where the population makes its living from farming. The annual UNDP report on social development in the year 2000 (UNDP 2000), devoted to the rural population, proposed the conclusion that what we are facing in a single country is actually two different situations and two different developmental conditions. We have two Polands in Poland. Using the HDI indicator, on the one hand there is the developed "urban country" (HDI = 0.828), and on the other hand there is the poorly developed "rural country" (HDI = 0.794). 4 1.16 The development of the urban population is less diverse. Over the 1980s, the Gini index amounted to 0.242 for cities and 0.277 for rural areas. The urban population earned its living mainly through work in the public sector as what statistics call the "working population." One might ask whether poverty was present in this group during the period of real socialism (and if so, what kind of poverty). 1.17 Studies from the 1970s and 1980s prove that poverty did exist. However, it was frequently related to family dysfunction, substance abuse, serious illness, and old age. Poverty did not appear among employed people or their families because full employment guaranteed subsistence, no matter how modest and uniform. According to GUS Central Statistical Office estimates on low income (similar to the social minimum category), it can be assumed that Poland began its transformation period with approximately 20 percent of its population threatened with poverty (Kordos 1992), although this does not necessarily mean deep poverty. 1.18 The World Bank report Understanding Poverty in Poland (a quantitative and very comprehensive study about income poverty in Poland) tried to identify who the people with the lowest incomes were and how their situation evolved during the first years of the transition, as well as how important growth is for reducing poverty, how the labor market enables the poor to lift themselves out of poverty, and to what extent social transfers are targeted at the poor and how these transfers can (given tight budgetary constraints) be restructured to reduce poverty (World Bank 1995). The findings of the study based on HBS individual data gave a somewhat new picture of poverty in Poland. The greatest surprise was that older people are not very poor. The primary group of poor people was the group of unemployed of a relatively young age (more than one-third of the cases), followed by people living in villages, especially peasants and people living in large families (30 percent of the cases). Only 5 percent of the poor were elderly. The second very interesting conclusion from this study was that Polish poverty is "shallow." This means that the average income (in the study on household expenditure) of poor people is only 12- 15 percent less than the given poverty line. At that time (1993), the shallowness also persisted with the use of higher poverty lines. This was an argument proving that there was no identifiable "underclass." 1.19 The second important and comprehensive study on poverty in Poland was carried out by a research institute--the IPiSS (Institute of Labor and Social Studies)--loosely affiliated with the Ministry of Labor and Social Policy. Over the period 1995-96 this institute together with the GUS Central Statistical Office produced data and quantitative research on poverty (Polish Poverty I and II). Those studies have contributed to the systematic monitoring of poverty rates applying different poverty lines and different dimensions (see Table 1.1). 1.20 Many qualitative studies were carried out in parallel. These were often very far-reaching and applied to specific communities (for which it was hypothesized that they were threatened by major poverty). The studies encompassed: · The post­collective farm communities · Small towns with a defunct employer · Major cities · Roma communities · The homeless · Single parents 5 · Women. Table 1.1: Comparison 1. Selected Quantitative Analyses of Poverty Author/Institution Undertaking Subject and Scope of Research Data Source Year of Research Research Quantitative Research World Bank Estimate of poverty incidence and BGD 1990 Milanovic, B. the poverty gap 1994 Topiska, I. 1995 World Bank Dynamics of poverty and research Panel 1993­1996 1993­1996 Okrasa, W. on the effectiveness of the social security net GUS Estimated rate of relative poverty BGD Systematically Szukieloj-Bienkuska, A. and lines established by IPiSS since 1993 (subsistence minimum and social minimum), estimates of subjective poverty Systematically Podgórski, J. since 1990 Panek, T. / Kotowska, I. Evaluating the risk of deficiency Special random February ­ March sample within the 2000 SGH ­ Warsaw School of framework of the Economics 2001 "Diagnoza 2000" (Diagnosis 2000) project IPiSS Evaluating the value of the Records of prices Since 1981 in the (Sajkiewicz, B. / Kurowski, P.) subsistence minimum and the case of Ministry social minimum of Justice and since 1996 in the case of Ministry of Education CASE/ ZeW (Beblo, Intergenerational dynamics of BGD, BAEL and 2000 M./Golinowska, S. /Lauer, Ch. poverty, alcohol abuse data base of social /Pitka, K /Sowa, A.) assistance CASE 2002 POMOST Source: Own comparison 1.21 These studies (see Table 1.2) provided much valuable knowledge. The main result of the research was the specifics of Polish poverty. As maintained by sociologists, candidates for the Polish underclass are mainly people affected by the pauperizing of agriculture, former employees of the state farms, and small-stake farmers who produce not for the market, but exclusively for themselves. 1.22 Polish poverty is not (yet?) ethnic in character and is linked to sex to a significantly lesser degree than in the countries of the West. Compared to other European states, Poland is a very homogeneous country in terms of nationality. It is estimated that no more than 3 percent of the total number of the country's residents represent national minorities which include Ukrainians, Belorussians, Lithuanians, Slovaks, Germans, and Roma. The largest national minority groups are the Ukrainians and Germans. The numbers of those groups can only be estimated (about 3 million for each group). The population census of 2002 results, expected by the end of 2003, will provide better information on the number of national minorities. Ethnic groups in Poland are rather small but are relatively well organized (with the exception of Ukrainians, which is probably due to their displacement to the northern part of Poland in connection with the "Vistula Campaign" following World War II). 6 1.23 The 1990s were characterized by the rapid development of special education opportunities for some national and ethnic groups at the primary and secondary school levels (UNDP/CASE 1998). Some of the schools that cater to national minorities and offer their native language are better endowed than Polish schools (for example, schools for Germans). Only the Roma (25,000 according to the Office for Immigration and Foreigners) do not have their own schools with their native language (and separate classes are provided only rarely) owing to the communist legacy and also to their cultural attitude towards raising their own children. A special assimilation policy for the Roma was introduced in the communist era and from then on they have used Polish as their common language and their children attend Polish schools. However, it may be said that the Roma are to some extent excluded from society because they live in a relatively closed ethnic group. It is thanks to this that they have preserved their customs and national identity. The most recent study concerning Roma in the Central and Eastern European countries showed that the material status of the Polish Roma is quite differentiated compared to that in Hungary or Slovakia. In Poland one can meet very rich Roma as well as poor ones. The qualitative sociological study by Laskowska­Otwinowska (Laskowska­Otwinowska 2002) provides a view of the attitude of the Roma towards work. It is probably true that the Roma face discrimination in the labor market, but at the same time they rarely look for regular jobs. 1.24 Quite interestingly, Polish research does not confirm the oft-voiced premise about the feminization of poverty. However, there is evidence that the phenomenon occurs in some backgrounds. From that point of view, Polish women are in a less favorable situation than women from other countries of the region (Domaski 2002), but this cannot yet be verified on a general scale. 1.25 Single parents raising a child (children)--single mothers for the most part--do not form a group subject to greater material problems than the average in Poland (Rymsza 2001). This conclusion also applies to other countries in Central and Eastern Europe. In general, single and divorced people find themselves in the poverty zone less frequently than others (Domaski 2002). Thus, this poverty profile linked to single parents so characteristic of the Western countries (especially the United States) is not (yet) a factor in countries undergoing transformation. There is no universal, and no convincing, explanation for this phenomenon. In Poland, state social instruments targeted at single parenthood are better developed than those aimed at other social problems. Their genesis harks back to the period of accelerated industrialization as well as mass migration from the country to cities. 1.26 Presently, however, there is increasing evidence relating to violence against women and children in the family, which can be a cause of falling into poverty and an exclusion situation. The statistics on violence are not adequate because women rarely claim violence in the very traditional Polish society with its views on family values, and acceptance of behavior involving such claims is rather weak (including in public institutions). 1.27 Violence in the family is more often than not linked to alcohol abuse, and leads to pathological family life. Alcohol abuse is a cause not only of poverty but of tragic living conditions for all the family members of the alcohol-dependent person. The usual figure for those within this problem circle is 3 to 4 million (Sieroslawski 1998). The cited research shows that after unemployment, alcoholism is the second most troublesome social problem at the local level. The result of this study of public opinion bears witness to the fact that in spite of a significant level of tolerance of excessive drinking, especially in rural areas (corroborated by all studies on the topic to date), excess alcohol consumption is seen as one of the most serious social problems. It should be borne in mind that, while the consequences of excessive drinking are manifold, in the context of poverty the basic problem is the low and falling capacity for employment (Sowa 2002). 7 Table 1.2: Selected Qualitative Research on Poverty in Poland Author/institution Undertaking Subject and Scope of Data Source Year of Research Research Research Tarkowska, E. ­ Zrozumie biednego [To Former employees of the Interviews in the field 1997 understand the poor] state farms (PGR) IFiS PAN 2000 Tarkowska, E. / Korzeniewska Children of the former state Interviews in the field 2000 Instytut Spraw Publicznych farm employees Institute of Public Affairs 2001 Tarkowska, E. /Laskowska ­Roma in Spisz Interviews 2001 Otwinowska, J., Fodor, E. / Domaski, H. Feminization of poverty in Random sample 2000 Poverty, Ethnicity & Gender the countries of Central and analysis IFiS PAN 2002 Eastern Europe Korzeniewska, K / Tarkowska, E. Lata Poverty and farmers (former Residents of two 1999 tluste, lata chude [Fat and lean years], state farm employees and villages: Owczary and IFiS PAN 2002 peasants) Kocielec Szeleny, I. ­ Poverty, Ethnicity and Bulgaria, Poland, Russia, 1999­2000 Gender in Transitional Societies, Center Romania, Slovakia, and of Comparative Research at Yale Hungary University Warzywoda­Kruszyska, W. et al. ­ y iEnclaves of urban poverty in Group interviews 1993 pracowa w enklawach biedy, (y) na Lód with employees of 1994 marginesie wielkiego miasta, social institutions 1997­1999 [To live and work in enclaves of poverty, (to live) on the edges of the big city] Uniwersytet Lódzki/ Lód University 1998 and 1999 Zablocki, G., et al. ­ Ubóstwo na terenach"Poor" gminas Interviews with 1997 wiejskich Pólónocnej Polski [municipalities] in residents, analysis of [Poverty in rural areas of Northern Koszaliskie, Slupskie, official documents, Poland] Pilskie, Elblski, Olsztyskie and interviews with Uniwersytet Toruski i Bank wiatowy/ and Suwalskie voivodships key persons in the Toru University and the World Bank local community 1999 Rossa, J. Workers' neighborhoods of Interviews with 1998 IPiSS i Instytut Polityki Spolecznej UW/ the Silwan and Stilon plants residents of workers' Institute of Social Policy of Warsaw in Gorzów Wielkopolski neighborhoods University Frieske, K. W. /Polawski, P. / Zalewski, Small towns ­ seats of the Expanded interviews 1997 D. gminas [municipalities] with selected families 1998 on welfare Przemeski, A. ­ Bezdomno jako Homeless people Expanded interviews From kwestia spoleczna w Polsce wspólczesnej with homeless people 1992/1993 to [Homelessness as a social problem in 2000 contemporary Poland] ­ Wydawnictwo Akademii Ekonomicznej w Poznaniu 2001 Marmuszewski, S. / Bukowski, A. Beggars Interviews with 1993 (editors) ­ ebracy w Polsce [Beggars in beggars in Wroclaw, Poland], Wydawnictwo Baran i Katowice, Lód, Suszczynski, Cracow 1995 Cracow, and Czstochowa Source: Own comparison 8 1.28 Children are the special victims of adult drinking; they are abandoned by their parents or are taken from their care by the courts. In such cases they most often end up in institutions called "Children's Homes." In Poland these are very poor institutions, in terms of both material resources and their capacity to raise children. Thus, a particularly difficult form of poverty exists in the case of children in institutions. The primary cause for concern is the growing number of children living without their own families. The number of children living outside their own families increased to over 10 percent during the transition period (Kolankiewicz 1999). 1.29 Child welfare institutions are insufficiently financed by the local governments. At the same time, a wide acceptance has developed of a new form of raising abandoned children--the foster families--a form that is strongly supported by the state. In the 1990s, the number of children in foster families increased above the 40 percent mark. In contrast, the number of adoptions is still decreasing (Kolankiewicz 1999). 1.30 Over the transformation period, the homeless and beggars make for a tragic and visible picture of poverty and social exclusion. Today, homeless people and beggars can be seen in railway stations and bazaars, and on the streets of major cities. Several interesting qualitative research studies of these groups have been carried out in Poland, mainly targeted at the homeless (e.g., Przymeski 2001) and at the long-term unemployed under special circumstances (former workers from the former state farms [Tarkowska 2000 and Tarkowska/Korzeniewska 2001], or passivity in the small towns [Frieske 1999]). Those studies identify the main characteristics of homeless people in Poland. First of all, they have no family roots or they have been rejected by their families as a result of alcohol/drug abuse or violence towards family members. Second, they have abandoned such institutions as workers' hostels (special homes dating from the communist era for workers from rural areas), children's homes, psychiatric hospitals, and prisons. The two groups of characteristics very often occur together. C. POVERTY DEVELOPMENT 1.31 The monitoring of poverty requires agreement between the assumptions behind the definition of poverty and the methodological guidelines for collecting, processing, and presenting data. The above-mentioned IPiSS and GUS studies of the mid-1990s provided for the systematic recording of what is known as income poverty. 1.32 Limiting the analysis to the monetary aspects of poverty was subject to criticism as an approach that is too one-sided (see, especially, Frieske 1996). However, the GUS studies were de facto supplemented by multidimensional research into population living conditions and studies based on subjective standards. Applied Poverty Definitions 1.33 Many measures have been applied to analyses of poverty in Poland--both the official measures and those proposed by various experts. This is a result of the relatively rich research tradition in this field. 1.34 In discussing the Polish tradition, it is necessary to begin with the criterion of the social minimum. The Soviet bloc countries applied a special category describing modest living conditions--the minimum of material security (minimuma meterialnoj obespieczennosti). While it is true that this category never defined poverty exactly, it did define the entry or fall into poverty. Those whose income fell below this level were defined as maloobespieczennyje (with small security) (McAuley 1979). 9 1.35 Minimal material security in Poland was defined by Tymowski (1973) and Deniszczuk (1977), and was called the social minimum. In 1981 (the year of the formation of Solidarity--the independent trade union), this category was recognized by the government as an official measure for monitoring living conditions. The social minimum describes the indispensable level of consumption determining social participation and social integration that requires satisfying not only basic needs but also certain needs beyond the basics. It is reflected in the contents of the basket of goods and services indicated as basic for satisfying needs at the social minimum level. The contents of this basket allow participation in social life (work, educating children, family life and socializing, cultural participation) all on a modest level. The basket of goods and services was recognized by specialists (doctors, dietary experts, social workers, statisticians, consumer researchers) as indispensable to the normal functioning of persons in society (with full integration possibilities). The social minimum has been systematically calculated by the Institute of Labor and Social Studies (IPiSS) and published in Polityka Spoleczna (Social Policy) since 1981. Today it can be said that the social minimum is more a measure of social integration than of poverty. 1.36 The social minimum was never used as a criterion for the categories applied in social policy, such as minimum wage, minimum pension, or threshold income in the social assistance scheme. However, at times banks used information relating to the social minimum to assess the credibility of households. Moreover, courts asked for social minimum information to adjudicate alimony to be disbursed by the breadwinner who abandoned the family. 1.37 A new category of poverty was defined in 1995--the so-called subsistence minimum. This category is used to define absolute and deep poverty--misery, in fact. The subsistence minimum was estimated by experts from the IPiSS on the basis of costs incurred by a very poor family's budget. This level of income (expenditure) accounts exclusively for needs whose satisfaction cannot be delayed. Any lower level of consumption leads to biological destruction. The level of the subsistence minimum is more than twice as low as the social minimum. The defining of the minimal existence category was carried out under the impact of the crisis of the initial period of the transformation, with its related far-reaching budgetary restrictions. Such an approach also occurred in other countries of the region. The equivalent category was titled crisis subsistence minimum in Latvia. In Slovakia, as in Poland, two categories are presently functioning: the social minimum and the existence minimum (GVG 2003). 1.38 Both categories--the social minimum and the existence minimum--can be considered aggregate poverty measures. They serve as an information function for policymakers and public opinion regarding the overall effects of social and economic policy in the context of the redistribution of income (Evans 2003). 1.39 The relative poverty line is determined by a value equaling 50 percent of average expenditure calculated for one consumption unit according to the OECD equivalence scale1 (1.0 for the first person, 0.7 for the second and subsequent persons aged over 15, and 0.5 for each child). The premise behind this definition is, acceptance of the subjectivity of poverty (dependency on the average living conditions in the country), while on the other hand, it points to inequalities as the main indicator of poverty. 1.40 The measure of poverty based on subjective evaluations is also defined and estimated. In Poland, what is known as the Leyden method is applied, with estimates by the Central Statistical 1 Recently, Eurostat submitted a proposal for adopting the line at the level of 60 percent of the median of national income, with the adoption of new equivalence scales. 10 Office (GUS) conducted since the beginning of 1990. Evaluations are obtained on the basis of research relating to the family budget: "Taking into consideration the current condition of the household, evaluate the level of income that would enable living on an average, satisfying level, and what income would not enable maintaining the household on even the lowest level" (Podgórski-Dobrowolska 1991). 1.41 For the official poverty line in Poland one ought to adopt the income threshold entitling a person to allowances from social assistance. Initially the threshold was established in relation to the minimum pension: income per person could not be higher than 90 percent of the minimum pension, which was defined as 35 percent of the average wage. This minimum pension was never officially named the poverty line, but in the social policy regulations of that time the minimum pension was used as a screening device: to divide applicants who should be helped from those who should not. This means that in practice the minimum pension determined an entry to the social assistance and housing allowances (Topiska 1997). This approach can be called a poverty proxy. Therefore, the World Bank poverty studies in Poland were made on the basis of the minimum pension as the official poverty line (World Bank 1994 and 1995). 1.42 Since 1996 an income threshold, which entitled one to claim social assistance benefits, was defined explicitly in the law of social assistance. From that time, in defining this income threshold it can be recognized as an official poverty line. The level of this income threshold was given as the value of 35 percent of the net minimum wage as the starting point (1996) and over time this amount (in zlotys) was indexed by price increase. 1.43 The Central Statistical Office on the basis of all these poverty definitions (lines) has produced since 1996 the systematic monitoring of poverty, and since 1998 it has published in the yearly report Warunki ycia ludnoci w, (Living conditions of the population, GUS 1999, 2000, and 2001). D. POVERTY TRENDS 1.44 Two periods of significant poverty increase are visible in the 1990s. The first is a period of a sudden rise marking the collapse of the communist bloc and the beginning of the transition from the command-based economy to the free market system. The GDP fell by approximately 18 percent2 over barely two years (1990-91). 1.45 The second period makes its appearance the end of the 1990s. Starting with 1998, the economic growth of Poland decreased while the unemployment rate became systematically higher. Over time, however, growth in average living costs has been falling because of the falling inflation rate. Food price growth has dropped at the most rapid rate and food weighs most heavily on the budgets of poor households (53-54 percent in households below the relative poverty line and 57 percent in households below the subsistence minimum (GUS 1992). This fact undoubtedly mitigates the growth in poverty, particularly the growth in extreme poverty. Yet the constant growth of living costs in other consumer spending--highest in housing costs (energy and rent) as well as transportation and communication costs--weighs more on households with average incomes. Thus, the increase in the share of people below the social minimum is significant (see Figure 1.1). 2 Estimates concerning the drop in GDP differ even in the GUS yearbooks for subsequent years; generally, the later the yearbook, the smaller the decrease. The 18 percent figure for 1990-91 is the figure most widely used. 11 Figure 1.1: Poverty Rates in 1994 - 2001, According to Different Measures 60% 57,0% 50,4% 52,2% 54,0% 47,9% 46,7% 50,0% 40% 40% 33,0% 34,8% 34,4% 30,8% 30,5% 30,8% 30,8% 32,4% 20% 14,0% 15,8% 16,5% 17,1% 13,5% 12,8% 15,3% 17,0% 12% 14,4% 15,0% 13,6% 13,3% 12,1% 9,5% 6,4% 6,9% 8,1% 5,4% 5,6% 0% 4,3% 1993 1994 1995 1996 1997 1998 1999 2000 2001 Subsistence minimum Relative poverty line Social assistance threshold Subjective line Social minimum Source: IPiSS database/GUS (Central Statistical Office) 2002. 1.46 Education costs (the enormous growth in school textbook prices) and medication costs belong to the group of prices with consistently high growth dynamics. Their impact is particularly high among households with children as well as households with disabled persons and those suffering from chronic diseases. Poverty indicators in these family groups are at least double the average (see Figure 1.2). Figure 1.2: Poverty Rates in 2001 by Family Type Social minimum Relative poverty line Subsistence minimum 0 10 20 30 40 50 60 70 80 90 100 Couple without children Couple with 1 child Couple with 2 children Single parents Couple with 3 children Couple with 4 + children Source: GUS 2002 1.47 The increase in the unemployment rate is a significant factor influencing the rise of poverty at the end of 1990s and at the beginning of the new century. The increase in 12 unemployment is consistently affected by the labor supply (particularly by the increase in new labor resources where education fails to match the needs of an intensively changing economy) and demand factors, where the restructuring of workplaces plays an important role. 1.48 The restructuring processes of the economy intensified over the years 1997 to 2000. A characteristic of these processes is their asymmetry, since the destruction of workplaces dominates their creation considerably (World Bank 2001). Over the years 1998 to 2000 the number of employed decreased by approximately 360,000 (Witkowski 2001). 1.49 Up until the present, new workplaces have been created primarily in the private sector and in services. In recent years both the creation of the private sector and the growth of services have slowed significantly. Over the 1998-2000 period, over 80 percent fewer workplaces were created as compared with 1994-97--a period of dynamic growth. This decrease in the growth of workplaces and in services over the 1998-2000 period exceeded 90 percent. This can be explained by the decrease in employment in social services in the public sector and the lower growth in the private sector as a whole. 1.50 There has been a growth in labor productivity as a result of restructuring. In industry, where the decrease in the number of employed was greatest (over 550,000 during the 1998-2000 period), labor efficiency increased by 14 percent in the year 2000 alone (Witkowski 2001). This growth in labor efficiency in industry is a sign of a desirable modernization of the economy; it is an indication of greater management efficiency throughout the whole of the economy. However, without a corresponding growth in workplaces in other sectors to compensate for the decline in workplaces in industry, this growth has disadvantageous effects on the labor market and has produced a dramatic rise in unemployment and poverty. Poverty Profile 1.51 A synthetic look at research results shows that poverty in Poland has most strongly touched households linked to agriculture, mainly those living in rural areas and in small towns, with poorer qualifications, with younger children, with people who have no family support and/or with problems related to dependences and new social diseases, often resulting in exclusion (Table 1.3). Table 1.3: The Rate of Unemployment Among Groups Vulnerable to Unemployment Groups with special vulnerabilities 1993 1996 1997 1999 2000 2001 XI XI VIII II 1st quarter 1st quarter On average 14.9 11.5 10.7 12.5 16.7 18.2 Young (15-24) 31.6 26.2 23.5 28.5 37.9 41.2 Female 16.9 13.4 13.5 13.5 18.5 19.8 Persons at immobile age (over 45 years) 8.5 6.5 6.5 6.3 10.6 11.9 Poor qualifications (basic and lower 15.0 12.9 12.9 11.7 22.1 22.9 education) Source: GUS (Central Statistical Office); based on the individual data of LFS. 1.52 Poor households in Poland today are not poverty-stricken households in the traditional sense (that is, households experiencing hunger and cold). Access to food, clothing, and even household furnishings is not a dramatic problem. (Over 90 percent of households living below all of the poverty lines own a color television set and refrigerator and also have running water, over 13 80 percent have vacuum cleaners, and 70 percent have flush toilets.) (GUS 2002) This is true even among the poorest households. Problems include limited access to and/or inability to benefit from the advantages of social integration, namely, good above-primary and higher education, high speed communications, transportation, and culture. An emerging problem is access to goods that guarantee an appropriate quality of life for the population. What is important in this case is disease prevention and health promotion as well as the potential for early intervention in the case of diseases that would lower the quality of life in the future. Poverty under such conditions becomes a factor leading to social exclusion--a particular cause for concern when it pertains to the younger population. E. POLICY RESPONSE 1.53 A policy of fighting poverty has not been, and still is not, a priority social program of the governments of Poland and other countries undergoing transformation. As one authority on social protection in candidate countries has described it: "Poverty for many of the transition countries is only a recently held official policy concept. External influences on poverty identification and recognition have been strong--particularly from the World Bank and UNDP and more recently from the European Commission. Social exclusion is little used but is growing in official use--but partly because of European Union accession driven response" (Evans 2003). Institutional reform in the main areas of social security as well as a guaranteed income for old age and disability pensioners have been a priority of social policy. Although true new institutions for the social protection of the unemployed and social welfare have been established, their capacity is limited, especially in Poland. 1.54 Social assistance in Poland constitutes a complement to the social security system in the case of poverty accompanied by grave social problems that could not be solved within the framework of social security or other state social policy programs. These problems were unanimously defined in legislation from 1990. Apart from poverty, this legislation lists ten complementary problems: orphanhood, homelessness, maternity protection, physical or mental impairment, chronic disease and disability, helplessness, addiction, post-incarceration adaptation, disaster, and long-term unemployment. In this vein, Polish social assistance differs from that of other Western countries, where there exists a category of a minimum guaranteed income, and anyone earning less can apply for complementary support. A poor person applying for assistance in Poland can receive such assistance only if his or her income situation can be classified into one of the above listed problem categories. 1.55 Since 1996, social assistance has granted social pensions for the disabled who became disabled in their childhood or adolescence, and thus did not become eligible for allowances from social security. Social pensions are granted on the basis of a medical assessment without the need to submit information regarding the level of income in the family. 1.56 Targeting allowances in social assistance is a commonly accepted solution that does not evoke controversy. However, a problem exists as a result of the arbitrariness of granting temporary cash allowances. Decisions on their granting are dependent on a social worker. In practice, a refusal or a decision against granting a very low benefit is not connected to an actual evaluation of the living conditions of an applying person, but rather to the dearth of funds in many local social assistance centers. For this reason, social workers postulate the legal establishment of obligations with clearly targeted allowances. 1.57 A significant problem facing social assistance as it functions in Poland is helping people of low ability to work whose situation involves a syndrome encompassing poor health, frequent 14 addiction, low qualifications, and bad relations with family members (or the total absence of family members). The strategy for assisting such poor (or even excluded) persons is far more complicated. Recently, programs for social employment, as well as legislation for social employment, for such persons was considered and enacted (April 2003). 1.58 Public support is also required for the following programs: · Combating and treating addiction · Providing advisory services and assistance for the helpless · Providing special and remedial education adjusted to the abilities and potential of intellectually neglected groups with a poor background · Providing psychological and spiritual (religious) assistance · Supporting and promoting nongovernmental organizations that facilitate a more individual treatment of persons in need of assistance (a deeper insight into human life), participating in contacts with employers and participating more broadly in the whole process of reintegration: "Do not speak to us, speak with us." 1.59 An important type of activity in every social assistance system is conditioned assistance. Such assistance is connected with people who are poor "at their own request." In analyzing such self-inflicted poverty, doubts frequently appear as to whether it is really a subject for consideration. Free men can decide their own destiny, and if they do not wish to be members of the mainstream society, it is their choice. The evaluation of this phenomenon is even more complex. Poverty originating from an individually shaped attitude to life is frequent among people who are psychologically troubled, rejected, lost and lonely, and, first and foremost, addicted. These are people who, in a society full of possibilities choices, are not able to choose independently or to their own advantage and are simply unable to participate in the social mainstream. Moreover, free or "loose" people (as they have been called historically) also have children, suffer from diseases, and become impaired and old. It is then that they need support. Before the need for institutional help becomes apparent, instruments of conditioned assistance are needed, such as: allowances conditioned by activity, the provision of useful work, various types of contracts. However, this help should not be impartial. Such help mainly involves allowances for the long-term unemployed and "rotating" unemployed, but it also encompasses the needs of chronic recipients of social assistance. Such instruments are designed with reluctance, partly owing to the pressure of the "sentimental" approach, and partly owing to the fear of the recipients. The problem also involves the exaction of commitment and control. Nobody likes such conditioned assistance, and yet no effective activity can do without it. 1.60 The social assistance system has taken over the implementation of family policies that in the 1990s offered only modest benefits. Family allowances have been subject to systematic reductions over the past decade. Starting with 1995, the family allowance was granted exclusively to less affluent families (a switch from universal benefits to targeted benefits). The value of the allowance was set at PLN 21.00, indexed once a year in line with the price increase index. Because Polish wages have increased dramatically since 1994, price indexation in the case of family benefits has had a negative effect on the benefit-to-wage rate. Within two years, the average value of the benefit had fallen to approximately 6 percent of the average wage. The principles of entitlement to family allowances were again changed in 1998. As in the socialist era, they are again based on the number of children in the family. The relative value of benefits 15 in the years 1999 to 2001 decreased considerably. The family allowance actually reaches about 3 percent of the average wage. The entire family protection scheme in Poland has changed into a narrowly focused form of assistance. Only families with disabled children receive better protection (Golinowska/Topiska 2002). However, the needs of public social services for such children are very modest. This is in contrast to a fairly well-developed movement of non-profit and self-help organizations. 1.61 In summary, Poland is in a situation in which the highest level of poverty is the province of families with children facing low and continuously falling family benefits. The most recent reaction of the government to this problem was an easing of "entry criteria" granting access to the system of social assistance for families with children. This action is based on a proposal to introduce two income thresholds: the social intervention threshold (PIS) for social welfare recipients who do not have children, and family income support (WDR) for families with children. These proposals are contained in legislation concerning family benefits. 1.62 The social protection housing policy, which is also an aspect of social assistance activity, has been transferred to the direct jurisdiction of the social welfare departments of local governments. Housing policy as an anti-poverty policy in the 1990s was provided in different ways over time. At the beginning of the transition it was offered on a relatively generous but only temporary basis through a social assistance institution-operated program of housing allowances introduced for poor families. In 1992 about 1.4 million people took advantage of these benefits. Moreover, the government continued to subsidize housing cooperatives for a few more years in order to maintain the existing apartments and to prevent an excessive rise in rent. From the second half of the 1990s local government became responsible for providing housing allowances. In 1998-99, the legislation on entry criteria for housing allowances was amended in order to provide better access to help. This was connected to the negative social consequences of the law introduced in 1996 relating to the eviction of people defaulting on rent payments. 1.63 At the same time, new private housing construction (focused on the middle class) was supported by the state through significant tax relief. Luxury private housing construction was also developed. Estimates of state support for private housing construction in the form of tax relief as well as subsidies for housing cooperatives amounted to 1.9-1.2 percent of GDP over the period 1994-98 (Ministry of Finance 1998). Later, housing tax relief was reduced and it was ultimately abolished in 2001. F. CHARITY AND NONGOVERNMENTAL ORGANIZATIONS 1.64 Private charity complements the state in its social functions but does not replace it. The range of complementation varies. To bring about its growth, a country needs a strong middle class consisting of people earning a stabilized income and with free time on their hands. In Western countries it was women who developed charity activities. In Poland, this will not become a mass phenomenon for a long time to come. Poland's wealthy men, although ostentatiously visible in the media, constitute a very narrow social group, while Polish women are involved in the labor market and work very hard (with the double obligation of family and paid work). There are few women with a sufficient amount of free time to work as volunteers (e.g., in social assistance homes) although there may be some exceptions. It must be admitted, however, that public social work was not promoted and no motivation was introduced during the transformation period, perhaps because during the period of real socialism the concept of public social work was discredited by some kind of obligation to give unpaid work for public aims with an ideological context (work to build a socialistic society). Political pressure was exerted for its 16 performance. Voluntary social work and mutual help were developed first of all in the private sphere: among families and friends and later also in the political opposition milieus. 1.65 At the end of the 1990s there were emerging calls for the need to regulate the work of volunteers. Social aid volunteers actually served as social case workers without any security and formal recognition. It took time for a greater understanding of the work that is known as volunteer work to evolve. It was not until March 2003 that the Sejm (Parliament) passed an act on governing public benefit organizations and regulating volunteer workers and establishing their rights. 1.66 Obviously, there are many nongovernmental organizations that collect funds for specific, well defined objectives and actually succeed in bringing them to life. (Such organizations were recently described in a systematic way in Hrynkiewicz 2002 and Klon/Jawor 2002. Their activities will not solve mass social problems, but such examples should be indicated as patterns to follow. It may even be worthwhile to create a financially significant instrument as a motivation for the spread of such activity. 1.67 There are not many models of assistance to follow with respect to the most difficult cases of poverty and marginalization. This field is relatively well covered by religious or monastic organizations. As regards to organizations targeted to adults from marginalized backgrounds, especially programs for men leaving penitentiaries and for addicted single mothers, a few nongovernmental organizations conduct such activity in Poland. BARKA (The Barge), a Pozna nongovernmental social organization, is a good example. It runs a special school for marginalized adults based on a Danish model (the Kofed School), inclusive of accommodations and unpaid work for the BARKA community. A very new law on social employment clarifies state obligations with respect to such communities and helps them develop a micro-economy. G. EUROPEAN UNION INTEGRATION AND COMBATING POVERTY 1.68 The process of Poland's integration into the European Union has at least two aspects in terms of poverty. The first is linked to the indirect impact of integration on the improvement of the population's living conditions and a drop in inequality. The important factors here are the potential to accelerate development as a result of the influx of capital, the expansion of sales markets for products made in Poland, and the increased possibilities for developing human capital as a result of the opening of the borders to educational and scientific exchange and to the flow of workers. The appearance of the positive results of these factors may be delayed, as the differences in the level of development between Poland and the EU countries continue to be large. The average per capita GDP in PPP in the EU countries amounts to US$ 22,000 (for the year 2000), while the figure for Poland is US$ 8,900, which is 40 percent of the average level for Europe (Eurostat 2002). This being the case, experience culled from the pre-accession period demonstrates that adaptive processes are extremely difficult and their positive effects do not overshadow the negative ones, and that benefits are not universal. 1.69 The second aspect of the impact on the social sphere of EU integration is coupled with EU policy in this field. The EU has been pursuing two social strategies with respect to poverty for the past few years. One strategy is the European Employment Strategy and the other is the Battle Strategy Against Poverty and Social Exclusion (EU Social Agenda 2000). In both cases, the instrument for the implementation of the strategies is relatively "soft." They are being introduced through what is known as the Open Method of Coordination (OMC). Candidate countries have been invited to learn this strategy and are already preparing national action plans in both areas. A qualification for participation in both strategies is the application of similar 17 social indicators for describing and ultimately monitoring jointly defined targets. Proposals for indicators with respect to both strategies have already been prepared. For the candidate countries this signifies the need to consider the level of the proposed indicators as well as a certain degree of effort in preparing statistical information to make the construction of the indicators possible. With respect to poverty and social exclusion, 18 indicators have been proposed, which differ from those used in Poland to date. Preliminary estimates with respect to several of them (the Laeken indicators) point to a need to significantly develop statistics in this area (see Table 1.4). Table 1.4: Selected Social Exclusion Indicators in the European Union, Average for 15 Member States Indicators EU % Remaining in poverty for at least 3 years, with the poverty line below 60% of the median income 10.0 (1995-97) Remaining at a level of education lower than secondary 19.7 (2000) Long­term unemployment (over 12 months) 3.7 (2000) Population (aged 0­65) in households where nobody is employed 14.0 (2000) Source: Eurostat 2002. 1.70 The European strategy for combating poverty and social exclusion is a logical consequence of the employment strategy. If it is assumed that 70 percent of the population of productive age is to work, and that the improvement of employability and adaptability is the fundamental goal of the strategy, then this must logically lead to the implementation of a program for the introduction (and reintroduction) to the labor market of people and groups not hitherto present for various reasons. Naturally, bringing the problem of social exclusion down to the absence of these people from the labor market is a simplification,3 but the emphasis of EU on integration through work is clearly visible (Esping-Andersen et al. 2001). 1.71 The emphasis on combating poverty and social exclusion through the implementation of a broad program of social integration is a significant change in the approach to social issues. It could be said that a passive approach is being replaced by an active one. It is not enough to provide poor people with cash benefits. They need assistance to help them live with others: in the mainstream of society. The activation of people threatened with exclusion or already excluded is a difficult and costly effort. It often requires the initiation of various types of rehabilitation: social, professional, and medical. The voices addressing the issue of education and employment for people with problems are not always the voices of acceptance. Addressing such issues is expensive, at times even more so than cash transfers (especially when these transfers are low, as in many candidate countries). Moreover, the operation involves establishing what is known as "worse" or "social" market segments. The EU priority needs to be considered carefully. Will all countries be able to afford it to a similar extent? 1.72 Nevertheless, social exclusion is even more expensive. It involves the need for long-term assistance because the people involved are unable to support themselves and have no security for old age. It also contributes to a lack of respect for social standards and leads to political radicalization. The exclusion of young people from the labor market may prove especially costly. On the one hand, it causes a loss of the potential for economic development, and on the other, it endangers the foundations of sustainable and democratic development. 1.73 The EU priorities on social issues are not unequivocal (or, one might say, they have many meanings). On the one hand, their aim is to develop a competitive region of the world through 3 The definition of an excluded person is a person who does not have sufficient resources to fully participate in social life. 18 the development of a modern economy based on knowledge. This requires high quality education, the development of scientific research, and a flexible labor market. On the other hand, there is the desire to introduce those who are less skilled, have problems with adaptation, and are less productive into the labor market. The result is that the European labor market is destined to be segmented and diversified; this appears unavoidable. 1.74 The program for combating poverty and social exclusion is currently at the stage of drafting action plans. The social objectives and indicators at the European level have already been defined. The objectives, adopted in Nice in 2000, are as follows: · To facilitate participation in work for everyone and to provide access to resources, rights, goods, and services · To prevent the risk of exclusion · To help the most vulnerable groups and individuals · To mobilize all authorities and institutions to achieve the above objectives. 1.75 It is vital to note that participation in work is of primary importance. As a result, the program for combating poverty and social exclusion is de facto a part of the European Employment Strategy. It is about providing employment for groups with difficulties or groups threatened by social exclusion. H. CONCLUSIONS 1.76 In general, Polish poverty remains related to the structural backwardness of the economy (relatively high in agriculture) and the deterioration in industry, as well as the high regional disparities. The restructuring process of the economy intensified in the second half of the 1990s and brought many difficulties in living conditions, especially for those groups most strongly affected by the changes. At the same time, a large number of threats potentially affecting young people emerged in the labor market. The high labor supply at the end of the 1990s and the beginning of the 2000s is not being met by a sufficient number of available jobs. This in turn makes young people susceptible to poverty. However, the rate of poverty among children depends, to a large extent, on social policy measures. Social policy in Poland today does not give priority to the well-being and education of children, nor to employment for young people. 1.77 European strategies in the realm of social policy accentuate the problem of the inclusion of people with difficulties in the labor market as well as those suffering exclusion. A better start for young people is also an aspect of the inclusion proposed in both European strategies (although it seems that the accent on various exclusion-related problems is spread uniformly through the EU programs). The analyses of Polish needs show that better education and work for the younger generation require a decided priority whose implementation should be considered as pressing. 1.78 At the same time, people affected by poverty and social exclusion, with poor human capital, living in the countryside and no longer young, are in need of help. The generation of off- farm employment for the work incentive segments of the economy is a key direction for the antipoverty strategy for these groups. 19 REFERENCES Atkinson, T. (2001) ­ Indicators for Social Inclusion in the EU, Report for the Belgian Presidency of the EU, memo. Beblo, M., S. Golinowska, C. Lauer, K. Pitka, and A. Sowa, (2002) ­ Poverty Dynamics in Poland. Selected Quantitative Analyses, CASE Reports, No. 54, Warsaw. Cichomski, B. (1998) ­ "Nierówno szans edukacyjnych w Polsce 1992­1997 w wietle danych PGSS" [Educational inequalities in Poland 1992­1997 on the perspective PPGS research]. Unpublished materials. University of Warsaw, Warsaw. Deniszczuk L. (1977) ­ Wzorzec konsumpcji spolecznie niezbdnej [Pattern of indispensable consumption], Studia i materialy IPiSS, Warsaw. Domaski, H. (1994) ­ Spoleczestwa klasy redniej (The middle class societies), IFiS PAN, Warsaw. Domaski, H. (2002) ­ "Ubóstwo w spoleczestwach postkomunistycznych" [Poverty in post­ communist countries], Instytut Spraw Publicznych, Warsaw. Erhel, C., J. Gautie, B. Gazier, and S. Morel, (1996) ­ "Job Opportunities for the Hard­to­place," in G. Schmid, J. O'Reilly, and K. Schoemann ­ International Handbook of Labor Market Policy and Evaluation, EE, Cheltenham, Brookfield. Esping-Andersen, G., D. Gallie, A. Hemerijck, and J. Myles (2001) - A New Welfare Architecture for Europe? Report submitted to the Belgian Presidency of the EU. Eurostat 2002 ­ Eurostat Statistical Yearbook on Candidate and Southeast European Countries, Luxembourg. Evans, Mark (2003) ­ "Social Exclusion and Poverty," in GVG (2003) ­ Study on the Social Protection Systems in the Thirteen Applicant Countries, European Commission ­ Employment and Social Affairs DG, Brussels. Frieske, K. W. (1996) ­ "Bieda ­ miary i interpretacje" [Poverty ­ Measures and Interpretation], w : Golinowska, S. (ed.) ­ Polska bieda. Kryteria. Oceny. Przeciwdzialanie [Polish poverty. Criteria, estimates, counteractions]. Studia i Materialy IPiSS, Warsaw, pp. 226 ­ 238. Frieske, K. W. (ed.) (1999) ­ "Marginalno i procesy marginalizacji" [Marginality and exclusion process], "Opracowania PBZ" vol. 13, IPiSS, IPS UW, Warsaw. Golinowska, S. (2002) ­ "The European Social Model and Open Coordination of Social Policy," in: Polityka Spoleczna (special edition) December, Warsaw. Golinowska, S. (ed.) (1996 and 1997) ­ Polska bieda. Kryteria. Oceny. Przeciwdzialanie [Polish poverty. Criteria, estimates, counteractions]. Studia i Materialy IPiSS. Warsaw. 20 Golinowska, S., and I. Topiska (2002) ­ Pomoc spoleczna ­ zmiany i warunki skutecznego dzialania [Social assistance ­ changes and conditions for effective actions], CASE, Warsaw. Górecki, B., and M. Winiewski (1998) ­ "Zrónicowanie dochodów gospodarstw domowych w Polsce w latach 1987­1995," in Podzial dochodu i nierównoci dochodowe. Fakty, tendencje, porownania [Income division and income inequality. Facts, tendencies, comparisons]. Council of Ministers Social and Economic Strategy Council, pp. 133­170. GUS (1997) ­ Wskaniki ubóstwa w swietle bada budzetów gospodarstw domowych [Poverty indicators in light of household budget surveys], Warsaw. GUS (1999)­ Warunki ycia ludnoci w 1999 [Living conditions of the population in 1998], GUS, Warsaw. GUS (2001) ­ Warunki ycia ludnoci w 2000 [Living conditions of the population in 2000], GUS, Warsaw. GUS 2002 ­ Warunki ycia ludnoci w 2001 [Living conditions of the population in 2001], GUS, Warsaw. GVG (2003) ­ "Study on the Social Protection Systems in the Thirteen Applicant Countries," European Commission ­ Employment and Social Affairs DG, Brussels. Hrynkiewicz, J. (2002) ­ Przeciw ubóstwu i bezrobociu: lokalne inicjatywy obywatelskie (Against poverty and unemployment: local civil initiatives), Instytut Spraw Publicznych, Warsaw. Innocenti Research Center of UNICEF (1995) ­ "Poverty, Children and Policy: Responses for a Brighter Futur," Regional Monitoring Report No. 3. Florence. Innocenti Research Center of UNICEF (2001) ­ "A Decade of Transition, The Monee Project CEE/CIS/ Baltics," Regional Monitoring Report No. 8, Florence. Klon/Jawor (2002) - Podstawowe fakty o organizacjach pozarzdowych ­ raport z badania 2002, Warsaw. Klugman, J., J. Micklewright, and G. Redmond (2002) ­ "Poverty in the Transition: Social Expenditures and the Working­Age­Poor." Innocenti Working Papers, No. 91, Innocenti Research Center of UNICEF, Florence. Kolankiewicz, M. (1999) ­ "Zagroone dziecistwo. Adopcje, rodziny zastpcze, instytucje opieki w okresie zmian spolecznych" [Menaced childhood. Adoption, foster families, institutional care in times of change], in Balcerzak­Paradowska, B. ­ Sytuacja dzieci w Polsce w okresie przemian [The situation of children in Poland in a period of change], IPiSS and SOS Foundation, Warsaw, pp. 192­211. Kordos, J. (1992) ­ Poverty Measurement in Poland, in PTS/GUS ­ Poverty Measurement for Economic in Transition in Eastern European Countries, PTS/ GUS, Warsaw. 21 Laskowska­Otwinowska, J. (2002a) ­ "Stereotyp biednego Roma" [Stereotype of a poor Rom], in Tarkowska, E. (ed.) Lata tluste lata chude [Fat years, lean years], IFiS PAN, Warsaw, pp. 239­258. Laskowska­Otwinowska, J. (2002b) ­ "Elementy kultury romskiej jako ródlo marginalizacji i bogacenia si Romów polskich" [Aspects of the culture of Roma as a source of exclusion and growing wealth among Polish Roma), in Tarkowska, E. (ed.) [Fat years, lean years], IFiS PAN, Warsaw, pp. 211­238. Laskowska­Otwinowska, J. (2002c) ­ "Community Study Report ­ Czarna Gora," in Poverty, Ethnicity, Gender (international project) www.yale.edu.ccr/laskowska.pdf. Leisering, L., and S. Leibfried (1999) ­ Time and Poverty in Western Welfare States. United Germany Perspective, Cambridge University Press, Cambridge. Lewitas, A., S. Golinowska, and J. Herczyski, (2001) ­ "Improving Rural Education in Poland," report prepared for the Warsaw Delegation of the European Commission, CASE Foundation, Warsaw. Lipton, M. (1995) ­ "Growing Points in Poverty Research: Labor Issues," in Rodgers, G. (ed.) New Approaches to Poverty. Analysis and Policy ­ I. The Poverty Agenda and the ILO. Issues for Research and Action, ILO, Geneva. McAuley, A.(1979) ­ Economic Welfare in the Soviet Union: Poverty, Living Standard and Inequality. University of Wisconsin Press, Madison, Wisconsin. Milanovic, B. (1997) ­ Income, Inequality and Poverty during the Transition from Planned to Market Economy, The World Bank, Washington, D.C. Mingione, E. (1996) ­ "Urban Poverty in the Advanced Industrial World: Concepts, Analysis and Debates," in Mingione, E. (ed.) Urban Poverty and the Underclass, Basil Blackwell, Oxford. Ministerstwo Finansów (Ministry of Finance) (1998) ­ Biala ksiga podatków (White paper on taxes), Warsaw. Nowicka, E., and S. Lodziski (2001) ­ U progu otwartego wiata. Poczucie polskoci i nastawienia Polaków wobec cudzoziemców w latach 1988­1998 [On the threshold of an open world. Perceptions of being Polish and attitudes towards foreigners, 1988­1999], Nomos, Cracow. Podgorski, J., and B. Dobrowolska (1991) - Subiektywne granice ubóstwa a dochody gospodarstw domowych [Subjective poverty lines versus household income], GUS Warsaw. Przymeski, A. (2001) ­ Bezdomno jako kwestia spoleczna w Polsce wspólczesnej [Homelessness as a social problem in contemporary Poland], Wydawnictwo Akademii Ekonomicznej w Poznaniu, Pozna. Rymsza, M. (2001) ­ Samotne macierzystwo i polityka spoleczna (Single mothers and social policy), ISP, Warsaw. 22 Sieroslawski, J. (1998) ­ "Problemy alkoholowe w ocenie mieszkaców" [Alcohol­related problems from the citizens' perspective]. Report on nationwide polls conducted in 1998 r., www.parpa.pl. Sowa, A. (2002) ­ "Alcohol Abuse and Poverty," in: Beblo, M., S. Golinowska, C. Lauer, K. Pitka, and A. Sowa ­ Poverty Dynamics in Poland. Selected Quantitative Analyses, CASE Reports, No. 54, Warsaw, pp. 67-88. Szulc, A (1996) ­ "Skale ekwiawalentnoci w pomiarach ubóstwa i dobrobytu" [Equivalency scales in poverty and wealth measurement], in Golinowska, S. (ed.) Polska bieda. Kryteria. Oceny. Przeciwdzialanie [Polish poverty. Criteria, estimates, counteractions]. IPiSS studies and materials. IPiSS, Warsaw, pp. 204­219. Tarkowska, E. (1998) ­ "Ubóstwo w bylych PGR­ach. W poszukiwaniu dawnych ródel nowej biedy" [Poverty in former collective farms. In search of old sources of new poverty]. Kultura i Spoleczestwo, No. 2. Tarkowska, E. (2000) ­ "Popegerowska bieda n" [Post­collective farm poverty], in E. Tarkowska (ed.), Zrozumie biednego. O dawnej i nowej biedzie w Polsce [Understanding the poor man. About the old and the new poverty in Poland], Typografika, Warsaw. Tarkowska, E., and K. Korzeniewska (2001) ­ Mlodzi z dawnych PGR­ów [Youth from the old state farms], Instytut Spraw Publicznych, ISP, Warsaw. Topinska, I. (1997) ­ "Ubóstwo w Polsce: 1993­1996" [Poverty in Poland: 1993­1996], in S. Golinowska (ed.) Polska bieda. Kryteria. Oceny. Przeciwdzialanie [Polish poverty. Criteria, estimates, counteractions]. SIPiSS studies and materials. IPiSS, Warsaw. Tymowski, A. (1973) ­ Minimum socjalne. Metodyka i próba okrelenia. [Social minimum. Methodology and trial of estimation], PWE IHW, Warsaw. UNDP 2000 ­ Rural Development. National Development Report Poland 2000, Warsaw. UNDP/CASE (1998) ­ "Access to Education," Human Development Report Poland '98, UNDP, Warsaw. Winiewski, M. (2002) ­ "Nierównoci dochodowe i majtkowe oraz zakres i formy pastwowej redystrybucji dochodów" [Income and propoerty inequalities and scope of state income redistribution], in Rada Strategii Spoleczno ­ Gospodarczej przy Radzie Ministrów ­ Nierównoci Dochodowe i Majtkowe. Zakres i Formy Redystrybucji Dochodów Pastwa, Report No. 49, Warsaw, pp.32-40. Witkowski, J. (2001) ­ "Transformacja rynku pracy w Polsce" [The transformation of labor market in Poland], in Kontrola Pastwowa [State control], No. 3, pp. 40­57. World Bank (1990) ­ Poverty World Development Report, 1990, Oxford University Press. World Bank (1994) ­ Poverty in Poland, Washington, D.C. World Bank (1995) ­ Understanding Poverty in Poland, Washington, D.C. 23 World Bank (2000) ­ Making Transition Work for Everyone. Poverty and Inequality in Europe and Central Asia, Washington, D.C. World Bank (2001) ­ Poland's Labor Market. The Challenge of Job Creation, Washington, D.C. 24 2. MACROECONOMIC DEVELOPMENTS OVER THE LAST DECADE: ECONOMIC GROWTH, INCOME DISTRIBUTION AND POVERTY Pierella Paci, Marcin J. Sasin and Jos Verbeek 2.1 In 1989 the first partially free elections in Poland's post-War history took place, and Poland became the first country in Central and Eastern Europe to reestablish democracy and embark on an economic and social transition to a market economy. A "shock therapy" program implemented during the early 1990s enabled the country to transform its economy into one of the most dynamic and robust in Central and Eastern Europe. Today, it's economy is by far the largest among the EU-acceding countries, as it accounts for over 50 percent of the cumulative population of the 10 candidate countries4 and also for about 50 percent of their cumulative GDP. 2.2 The beginning of Poland's transition in 1990 was marked by exceptionally difficult macroeconomic conditions, which included high inflation, a large legacy of external debt, and a high black market foreign exchange premium. Saddled with a large part of the enterprise sector that was considered "value subtracting," Polish policymakers took huge risks by making the zloty convertible, fixing the exchange rate, and lowering import barriers. With privatization lagging behind at that time, many predicted a crisis based on the notion that enterprises would not be able to cope with market conditions, which would lead to politically unacceptable mass bankruptcy and social upheaval. This did not happen. On the contrary, Poland turned out to be unique among the large front-running European transition countries in having an unbroken growth record once growth resumed after the initial output collapse. A. A SUCCESFUL MACROECONOMIC PERFORMANCE 2.3 However economic growth was not uniform over the last decade. After five years of growth rates of 5 percent or above, growth declined substantially after 1998 to reach a meger 1 percent in 2001, only to increase again after then with good prospects for the future (see Fig 2.1 and table 2.1). The slow down in growth was accompanied by an increase in poverty as it is evident from Fig 2.1. 4The ten acceding countries are the Czech Republic, Cyprus, Estonia, Hungary, Latvia, Lithuania, Malta, Slovakia, Slovenia, and Poland. 25 Figure 2.1: Growth, Employment and Poverty Poverty change (%points) Employment (LHS, mn) 6.6 7% GDPgrow th (%y/y) No. of poor (RHS, mn) 16.0 6% .nm, 6.4 5% 6.2 15.5 entm nm, 4% oypl poor 6.0 3% 15.0 mE of.oN 5.8 2% 5.6 1% Poverty increase 5.4 0% 14.5 -1% 5.2 Poverty decline -2% 14.0 5.0 1995 1996 1997 1998 1999 2000 2001 2002 1994 1995 1996 1997 1998 1999 2000 2001 2002 Source: Staff calculations based on CSO's data. Table 2.1: Selected Economic Indicators, 1997-2003 Main macroeconomic indicators 1997 1998 1999 2000 2001 2002 2003 GDP growth, % y/y 6,8 4,8 4,1 4,0 1,0 1,3 3,5 Private consumption growth, % y/y 5,2 2,7 2,0 3,3 3,3 Gross fixed investment growth, % y/y 21,7 14,2 6,8 2,7 -9,8 -7,2 0,0 Unemployment rate, e.o.p., % 10,3 10,4 13,1 15,1 17,4 18,1 17,7 Poverty headcount, % of population 14,7 13,1 14,3 14,8 15,6 16,6 . . General gov't expenditures, % of GDP 43,5 41,9 42,7 41,1 43,9 44,5 46,9 General gov't budget balance, % of GDP -2,8 -2,4 -3,1 -3,0 -5,1 -6,0 -6,3 Consumer price inflation, % y/y 13,2 8,6 7,3 10,1 5,5 1,9 0,9 NBP refinancing rate, % 28,0 21,0 21,5 24,0 16,5 9,75 6,0 Current account deficit, % of GDP -2,9 -4,2 -7,2 -6,1 -3,9 -3,5 -3,0 External debt, % of GDP 35,5 36,0 42,4 40,3 38,1 40,7 . . e.o.p ­ end of period, 2003 ­ forecast or latest available Source: CSO's data. 2.4 The remainder of this chapter attempts to analyze the relationship among growth, income distribution, and poverty and to discuss the influence of the macroeconomic policy stance on the poor in Poland. Policy recommendations are given in the final part. The remainder of this section provides a brief discussion of recent economic developments in Poland. The Polish Central Statistical Office's annual Household Budget Survey (HBS) micro-data sets (1994-2002), as well as publicly available real sector and monetary macro data, are used to analyze linkages between the macroeconomic environment on one hand and poverty, inequality, and consumption distribution trends on the other. The factors behind the trends in growth 2.5 While many factors can account for this growth record - including consistent economic policies despite frequent changes of government - the main economic factor was that Poland managed the macro-micro policy linkages well. At its core, this meant maintaining a combination of hard budget constraints for enterprises, a competitive real exchange rate, and a post-privatization governance structure that allowed business, in particular small and medium- size enterprises, to flourish. By reducing fiscal deficits and placing public debt on a stable and sustainable trajectory (including making debt reduction agreements with the Paris Club), Poland 26 achieved a macroeconomic environment that was conducive to growth and that allowed a gradual decline in inflation (see Table 2.1). However, it is important to note that during the early years of transition the restructuring process was aided by relatively easy access to social safety net programs for redundant workers and that the dismantling of the overly generous social assistance network of the communist era has not been tackled with the same vigor as the market-oriented reforms. 2.6 The loss of export markets in the East as a consequence of the Russian crisis amounted to around 3 percentage points of GDP and triggered a new round of enterprise restructuring to curtail falling profitability. This time, the resulting improvements in productivity were brought about in large part by reducing employment. This reduction, together with the increased numbers of newcomers to the labor market owing to the baby boom of the early 1980s, led to significant increases in unemployment. As of April 2003, over 3.2 million people were unemployed (or 18.4 percent of the labor force). 2.7 These developments coincided with a shift in monetary policy in 1999 leading to the creation of an independent Monetary Policy Council (MPC). Inflation targeting became the main anchor of monetary policy and later became one of the means of facilitating accession to the EMU. When inflation rebounded during 1999 and 2000 into double digits, the MPC progressively tightened monetary policy by raising its rediscount rate to a peak of 21.5 percent in August 2000. As a result, ex-post real lending interest rates increased to over 13 percent during 2001, up from around 10 percent during 1999-2000. In combination with possible over- investment during the high growth years of 1996-99 this led to domestic demand, and in particular investment, declining significantly as from the second half of 2000. Inflation declined impressively to less than 1% in 2003. Figure 2.2: GDP Growth, Export, and Industrial Performance GDPDevelopments GDP growth 8 Real growth y/y, % initial distr. 7 newdistr. 6 0.4 onlymeanshift Real growth y/y, % 5 povertyline 4 h 0.3 ction 3 grpowt produ 2 0.2 port density 1 Ex strial 0 0.1 Indu 791 792 793 794 891 892 893 894 991 992 993 994 001 002 003 004 101 102 103 104 201 202 203 204 301 302 QT QT QT QT QT QT QT QT QT QT QT QT QT QT QT QT QT QT QT QT QT QT QT QT QT QT equivalentconsumption GDPreal growth y-o-y 4 Qtrs Moving Average 0 Source: Staff calculations based on CSO's data. 2.8 The macroeconomic environment deteriorated progressively at the end of the last decade. While monetary policy tightened during the second half of 2000, fiscal policy eased considerably during 2001. Not only did the General Government Deficit widen from .1 percent for 2000 to 6.3 percent for 2002, but the structural deficit (i.e., corrected for business cycle effects) worsened by over 2 percent to over 5 percent of GDP (see Figure 2.3). This "tight monetary and loose fiscal" 27 policy mix has crowded out investment and moderated the potentially expansionary effect of fiscal policy. Figure 2.3: Fiscal and Monetary Developments, 1996-2003 1 Structural De ficit (% of GDP) M one tary Conditions Inde x 2.5 0 2.0 -1 1.5 -2 1.0 -3 0.5 Monetary tightening -4 0.0 -5 -0.5 -6 -1.0 Monetary easing -7 -1.5 1996 -2.0 19971998 S tru c tu ra l 1999 -2.5 2000 2001 96M1 96M8 97M3 97M10 98M5 98M12 99M7 00M2 00M9 01M4 01M11 02M6 03M1 H e a d lin e 2002 2003 Headline Cyclical Structural B. DID POVERTY RESPOND TO GROWTH? 2.9 It is clear that the high growth rates of the mid-1990s to 1998 facilitated a decline in poverty in Poland suggesting high responsiveness of poverty to economic growth. However, from 1999 onward this responsiveness appeared to be reduced as poverty begun to increase even with growth remaining at 4 percent for 1999 and 2000 and positive there after. A Statistical Artifact? 2.10 However, in fully evaluate the role of growth in poverty reduction it is important to note that household consumption - as captured by the Household Budget Surveys (HBS) ­performed much worse than the national accounts (NAS) suggest with average household consumption declining in the low ­ but positive - growth years post 1998. This difference persists even when the figures for changes in aggregate private consumption are considered rather than GDP growth. This is not a phenomenon unique to Poland and may be explained by the unusually high investment accumulation experienced in the Polish case over the 1990s and by the relatively high export rate.5 Reconciling these two series is beyond the scope of this report but it is important to note that the HBS shows declining values of average per capita consumption for the period 1999 ­ 2002 while the NAS data show moderate positive growth. Hence the estimates of poverty, which are based on the HBS, showed increases in poverty in these years despite the positive consumption growth rates shown by the national accounts data. 2.11 When the responsiveness of poverty to growth in average consumption is considered ­ rather than to GDP growth ­ the analysis shows a very strong link between growth and poverty reduction. Over the last decade the average estimated elasticity of poverty with respect to growth in average consumption from the Household Budget Survey (HBS) has been 3.6 with a peak of 4.11 in 1999.6 Given Poland's GDP level, his value compares very well with the estimates for other transition countries and with those typically found in countries outside the region [Bruno, 5Richard (2003) 6These estimates are in line with the value of 3.5 given in World Bank (2002). 28 Ravallion, and Squire (1998); Ravallion and Chen (1997)].7 To sum up: 1 percent growth in mean consumption brings about a 3 percent reduction in the poverty rate (headcount) which is equal to about 0.5 percent of the total population or (multiplying by 38.6 million) 185,000 people. Given the relatively high size of the elasticity and the optimistic predictions for growth in 2004, the prospects for poverty reduction in Poland are good. Assuming inequality to increase at the average rate experienced over the last five years, this should reduce poverty by 6.4%, lowering it by 1.5 percentage points to 15.1% in 2004. But Inequality Matters Too. 2.12 However, inequality has also risen steadily Figure 2.4: Inequality, 1994-2002 through the second half of the 1990s (see Figure 2.4). In 2002 the Gini coefficient for consumption 0.29 Gini coeff, LHS 0.13 inequality was 0.28, which was moderately high 0.28 Mean log deviation, compared to other CEEC countries (see, for 0.28 RHS 0.12 example, World Bank 2000 and 2002). Given the 0.27 formerly centrally planned economic systems in 0.27 CEEC countries and their subsequent economic 0.11 0.26 development, it is not surprising that growth be 0.26 associated with rising inequality. In general, 0.10 however, ECA countries constitute an outlier in the 0.25 overall picture ­ any simple universal link between 0.25 growth and inequality has been flatly rejected by 0.24 1994 1995 1996 1997 1998 1999 2000 2001 2002 0.09 recent studies (see for example, Ravallion and Chen 1997). Source: Staff calculations based on GUS's data. 2.13 Table 2.2 decomposes the overall inequality ­ measured by the Theil's mean log deviation index into two components: (i) differences between groups and (ii) within group inequalities. The focus on variations within (and between) (i) regional groups (96 groups = 16 voivodships crossed with 6 town sizes), (ii) educational groups (3 groups: higher, secondary, other) and (iii) age categories. If, for example, mean consumption for all regions was the same ­ ie, across region variation was 0, then the numbers in the respective row in the table would be 0 and all inequality could be entirely attributed to inequalitties within each region. Analogously, if everyone within a given region had equal consumption ­ ie, within regional variation was zero - but regions differed in mean consumption, the "between" inequality component would be 100 percent. The table shows that inequality caused by variation within groups, as could be expected, is the main component of the total inequality for all groups analysed. However, differences between different educational groups are a substantial and sharply incresing component of the overall inequality and so are the differenes between different regions. On the other hand differences across different age groups were and remain minimal. 7 Just to give few examples, the growth elasticity of poverty for Romania was estimated at 3 in 1996 and 1997 but has declined since to below 2 (World Bank, 2003). An average elasticity of 3.7 was also estimated for Croatia (World Bank (2002). However the limited information available suggests much lower responsiveness in other countries. Estimates of around 2 were derived for Russia (World Bank, 2002) and less than 1 for Armenia, Azerbaijan and the Kyrgyz Republic (World Bank 2003 29 Table 2.2: Decomposition of Theil Inequality Index (Inter-Group Inequality), 1994-2001 1994 1995 1996 1997 1998 1999 2000 2001 Theil index 0.100 0.095 0.097 0.107 0.110 0.114 0.120 0.120 Decomposition Regional Within groups 0.092 0.087 0.089 0.097 0.097 0.100 0.104 0.106 Between groups 0.008 0.008 0.008 0.010 0.013 0.014 0.016 0.014 Share of between groups 7.9% 8.3% 8.1% 9.4% 11.5% 12.5% 13.0% 11.6% Education Within groups 0.087 0.083 0.084 0.092 0.093 0.095 0.097 0.097 Between groups 0.013 0.012 0.130 0.015 0.017 0.020 0.023 0.023 Share of between groups 13.0% 12.9% 13.4% 13.8% 15.9% 16.5% 19.5% 19.1% Age Within groups 0.098 0.093 0.095 0.105 0.107 0.111 0.118 0.118 Between groups 0.002 0.002 0.002 0.002 0.003 0.003 0.002 0.002 Share of between groups 1.7% 2.2% 2.2% 2.0% 2.5% 2.4% 1.5% 1.7% Source: Authors' calculations based on HBS. 2.14 Thus that regional inequalities clearly increased during 1994-2002 not only in absolute terms (see Chapter 3 on regional poverty and prices) but also in relative terms. Or, equivalently, regional inequalities rose even more than total inequalityand the role of regional inequalitise in expleining overall inequality increased. The increase in education premia is also an important component of the raising inequality. 2.15 Further since the disparities in regional prices grew as well (see Chapter 3 on prices), a straightforward question is whether and to what extent these effects interact. Do relative price changes alleviate increasing regional inequality? The right panel of Table 2.3 tries to answer this question. In both approaches inequalities between regions during 1994-2001 grew at a similar pace, so it is not possible to detect or establish such an intuitive link. 2.16 When inequality increases (if other things are kept constant), poverty normally increases as well, thus dampening the beneficial effect of growth on poverty. Fortunately, changes in the shape of the consumption distribution are much slower than the change in the mean. Therefore, the positive growth impact should outweigh the negative effect of a widening of the distribution itself. Decomposing Trends in Poverty 2.17 Having described the trends in growth and in the inequality, it is important now to analyze the effects of these trends on poverty. Table 2.3: Decomposition of Changes in Poverty, 1995-2002 (percentage points) Poverty change 1995 1996 1997 1998 1999 2000 2001 2002 Total -0.4% -1.5% -0.3% -1.6% 1.3% 0.5% 0.8% 1.0% Due to growth 0.4% -2.1% -1.1% -2.0% 0.5% 0.4% 0.6% 0.6% Due to inequality -0.9% 0.3% 0.8% 0.5% 0.8% 0.0% 0.2% 0.5% Source: Staff calculations based on GUS's data. 2.18 The results of such a decomposition for Poland are presented in Table 2.4 and Figure 2.6.8 It is evident that economic growth was a main driving force in poverty reduction. Two 8For a description of the methodology used see Bourguignon (2002). 30 periods are readily distinguishable. During Poland's high-growth years, 1994-98, the economic growth effect (5 percent) far outweighed the negative effect of the increase in inequality (-1 percent), resulting in a cumulative decline in poverty of about 4 percentage points. During the subsequent period, 1998-2002, and because of a decline in mean consumption, lack of growth and increased inequality jointly resulted in an increase in poverty of about 3 percentage points. Figure 2.5: Decomposition of Changes in Poverty,1994-2002 3% 2% 1% 0% -1% -2% -3% -4% -5% total poverty change due to grow th -6% due to inequality 1994-1998 1998-2002 1994-2002 Note: The 2002/1994 change is approx. a sum of 1998 ­2002 and 1994-1998. Source: Staff calculations based on GUS's data.. 2.19 For additional illustrative purposes, one can look at the consumption level and its dynamics at different parts of the consumption distribution. It is interesting to compare mean incomes/consumption across the ranked population. Inverting the cumulative distribution of consumption at a p-th percentile, one can infer consumption levels at this percentile. This line ­ depicted in the left panel of Figure 2.6 ­ is called a quintile function, or, sometimes, a Pen's parade (Pen 1971). For example, individuals at the seventy-third percentile had in 2002 a consumption level of almost PLN750. Figure 2.6: Consumption & Growth in Mean Consumption by Percentiles/Quintiles, 1995-2002 Consumption growth for selected 1339 The Pen's parade: 7.1 t consumption by percentile quintiles 1130 6.9 1 adul 1994 5% 953 1998 2 6.7 eq./ 4% 5 804 2002 6.5 3% 679 month/ 2% 6.3 572 NLP ` 1% 483 6.1 0% 407 5.9 -1% 344 5.7 percentile -2% 290 5.5 -3% 1 10 19 28 37 46 55 64 73 82 91 100 1995 1996 1997 1998 1999 2000 2001 2002 Source: Staff calculations based on GUS's data. 31 2.20 The left panel of Figure 2.6 presents three curves for three different years (1994, 1998, and 2002). Between 1994 and 1998 the line shifted almost parallel upward implying a rise in consumption across the entire population. However, between 1998 and 2002 the line generally shifted downward, (except for the higher percentiles). The magnitude of this decline was clearly greater for the lower percentiles. The consumption level for the poor returned to, or even fell below, its 1994 level. The right panel in Figure 2.6 presents the same story but from a dynamic perspective. The period 1995-98 was generally favorable for all the consumption groups,9 but in 1999 an across-the-board deterioration emerged, with only the higher quintiles maintaining growing consumption. 2.21 It is straightforward to extend the idea of Pen's parade to show the change in mean consumption (rather than level) for various percentiles of consumption distribution during the period of interest. This approach, which uses the so-called growth incidence curve ( gic ), is meant to show how gains from aggregate economic growth are distributed among the population.10 Figure 2.7 shows growth incidence by percentile. Figure 2.7: Growth Incidence Curves 2.22 It should be noted that there was almost no change in the consumption level at the tenth percentile between 1994 and 2002. ) 15% However, the consumption level at the ngeahC ninetieth percentile and higher grew by over 10% 15 percent over this period. These results are %( 5% consistent with what has already been 0% identified. The curves are, for example, ption -5% (almost) strictly increasing, meaning that um inequality has risen, at all levels of -10% onsC distribution. -15% 1 10 19 28 37 46 55 64 73 82 91 100 2.23 During 1994-98 the curve was above zero everywhere, indicating that poverty had 1994/199 1994/2002 1998/2002 fallen, whatever the poverty line and 8 whatever (reasonable) poverty measure is adopted. Although, technically speaking, the Note: The 2002/1994 change is approx. a sum of 2002/1998 and 1998/1994; Source: Staff calculations based on GUS's data. growth could be described as moderately pro-rich (higher growth on the higher quintiles), it brought significant benefits to the poor as well. During 1998-2002, the situation changed radically: only the highest decile managed to maintain its consumption level and the reduction was the hardest for the poorest percentiles. 9 There could be two explanations for the 1997 weakperformance of the lower quintiles: a tightening of social benefits eligibility and a massive flooding in the summer. 10Warning: The curves should not be interpreted as a growth in consumption of an average p-percentile Mr. X during that period. The cross-sectional character of the exercise implies that the graph rather depicts the difference of mean consumptions of two groups of individuals happening to be in the same percentile at both ends of the period of interest. We are comparing the poor in year A with the poor in year B, not asking whether they are the same people, thus (leaving aside mobility within the distribution) we underestimate the improvement for the particular poor and overestimate a growth of the particular rich. Having said that, we are sticking to our curves because we are interested in the poor in general, whatever their history (i.e., unconditionally). 32 2.24 In addition to the discussed indicators, Ravallion and Chen (2003) derive an intuitive, straightforward index to measure pro-poor growth: it is equal to a mean growth (change in (h) consumption) for the poor:11 = gic(p)dp (h) p where (h) is a percentile where 0 the poverty line is drawn (equal to the poverty headcount). 2.25 This index can then be compared with an average growth to infer whether the growth is pro-poor. Table 2.4 presents the comparison. Table 2.4: Pro-poor Growth Growth 1995 1996 1997 1998 19999 2000 2001 2002 Average -0.7% 3.5% 2.1% 4.2% -0.9% -0.9% -1.0% -1.0% Pro-poor 0.3% 3.2% -0.9% 3.2% -2.1% -2.8% -2.3% -2.1% Source: Staff calculations based on GUS's data. C. WHAT IS BEHIND THE INCREASE IN CONSUMPTION INEQUALITY? 2.26 Changes in inequalities of consumption can be the result of a number of factors of which four important ones are changes in (i) the distribution of original income; (ii) the progressivity of the tax and benefit system; (iii) the distribution of savings and (iv) the realtive prices of the basket of goods bought by household at different points in the income distribution. 2.27 Unfortunately the Polish HBS has very limited information on household savings. Thus the analysis below focuses on the other factors highlighted above. Changes in the distribution of original income and its components. 2.28 Original income refers to the income the household has before paying taxes and receiving benefits (other than old-age pensions). As shown in Table 2.5, over the period 1994-2002 inequality in original income increased by around 7 percent from 0.38 in 1994 to 0.41 in 2001. The increase was also surprisingly moderate considering the extent of transformation the country went through. It was particularly so considering that it occurred entirely after 1998 with inequality in original income fluctuating around a stable value in the high growth years 1994-98. 2.29 Table 2.5 also shows the main sources of inequality and the main factors behind the increase. Inequalities in labor income were and remain, a major source of total income inequality. In addition their relative role in the overall inequality has increased over time until 2001 (from 70.5 percent tpo 72.3 percent) only to decline slightly in 2002. The combination of incresed inequality and the growing share of this inequality that is due to labor income dispersion resulted in a increase in the Gini coefficient for this component of income of 8 percent, higher than for overall income and for any other component. 11This index has an additional desirable property of being based on an "axiomatically appropriate" poverty measure, (h ) namely the Watts index: W = ln( H / c( p )) dp . 0 33 Table 2.5: Income Inequality and Decomposition by Income Sources, 1994-2002 Gini Coefficients: 1994 1995 1996 1997 1998 1999 2000 2001 2002 Original Income* 0.379 0.377 0.386 0.395 0.3657 0.377 0.395 0.395 0.407 of which: Labor income 70.5% 68.9% 70.8% 71.6% 72.1% 73.7% 74.1% 72.3% 69.9% Old age pension 15.1% 16.5% 15.9% 16.0% 16.4% 16.4% 15.3% 17.8% 18.8% Income from farm 10.1% 11.6% 9.3% 8.9% 7.6% 6.4% 6.2% 5.7% 6.6% Other income 4.3% 3.0% 4.1% 3.5% 3.9% 3.5% 4.4% 4.3% 4.7% Disposable Income** 0.308 0.304 0.312 0.323 0.293 0.298 0.311 0.308 0.317 Concentration Coefficients***: Labor Income 0.389 0.382 0.396 0.402 0.377 0.398 0.420 0.414 0.421 Old age pension 0.394 0.408 0.404 0.410 0.380 0.385 0.372 0.407 0.423 Income from farm 0.332 0.358 0.326 0.371 0.290 0.273 0.292 0.287 0.338 Other income 0.307 0.254 0.320 0.289 0.293 0.265 0.304 0.280 0.295 Method: decomposition of the Gini coefficient into components, ranking by total income, see Chapter 2 in Volume 2 for mathematical representation * Before taxes and benefits. ** After taxes and benefits *** Household ranked by original income. Source: Authors' calculations based on HBS. 2.30 To conclude, an important source of the incraeses in inequality of consumption over the 1994-2002 period that is extensively described above can be found in the growth in inequality in original income. The main force behind this increase has been the growth in ineqaulity in labor income with is reflected in both a higher Gini coefficint for this type of income and the incraesed role played by this component in the overall measure of inequality. The role of taxes and social benefits 2.31 While the increased inequality in original income is clearly an important component of the dispersion of consumption. It clearly is not the full explanarion. At 7 percent, the growth in the Gini coeffient for original income is just over half of the increase in Gini coefficient for consumption (13 percent). Thus other factors must be at play. Amongst them of potential importance are changes in the tax and benefit system which ­ for any given distribution of original income - may affect the distribution of income households have to spend ­ ie, the disposable income -. 2.32 The top part of Table 2.6 shows that the concentration index for taxes has increased by 13 percent in the period 1994-2002 suggesting that the tax burden has become more concentrated. At the same time the social benefits system has changed from one favoring the rich ­ although very slightly ­ to one that favors the poor even more slightly. This would suggest that as from the mid- 1990s the changes in the tax and benefit system have counteracted the increased inequality in original income leading to a much smaller growth of inequality in disposable income (only 3 percent from 0.308 to .316). 34 Table 2.6: The role of taxes and social benefits in inequality trends 1994-2002 Concentration Coefficients*: 1994 1995 1996 1997 1998 1999 2000 2001 2002 Taxes 0.305 0.303 0.312 0.321 0.313 0.322 0.308 0.328 0.344 Benefits 0.020 0.003 -0.009 -0.006 -0.023 -0.019 0.028 -0.018 -0.013 Disposable income 0.308 0.304 0.312 0.323 0.293 0.298 0.311 0.308 0.317 of which: Original income 113% 114% 115% 113% 114% 114% 111% 110% 110% Taxes income 1% 0% 0% 0% -1% -1% -1% -1% -1% Benefits -14% -15% -15% -13% -13% -13% -10% -9% -9% Progressivity**: 1994 1995 1996 1997 1998 1999 2000 2001 2002 Taxes -0.070 -0.072 -0.071 -0.069 -0.045 -0.012 -0.010 0.002 0.004 Benefits 0.650 0.695 0.707 0.71 0.72 0.746 0.767 0.762 0.771 Method: * Household ranked by original income ** Kakwani index. Source: Authors' calculations based on HBS. 2.33 An alternative way of looking at the effect of fiscal policy on poverty is to consider its redistributive effect. This can be visualized by comparing the Lorenz curve for original income with the concentration curves for taxes and benefits. Fig 2.8 compares the redistributional impact of the system in 1998 and 2002. The comparison reveals some interesting conclusions. The first surprising result is that in 1998 the concentration curve for taxes dominated the Lorenz curve for original income ­ ie, was above the Lorenz curve for al points in the income distribution -. This suggested that the overall tax system was regressive ­ ie, the poor paid more than the better-offs as a proportion of their income -. However, since the distance between the two curves was very small, the magnitude of the redistribution involved was very limited. In 2002 the situation for the poorest individuals had not changed. The share of taxes they paid exceed their share of income , suggesting a redistribution away from the poor. However, in 2002 the concentration curve for taxes crossed the Lorenz curve making imposisble to derive firm conclusions about the overall level of redistribution. Fig 2.8: Progressivity of tax and benefit system in 1998 1 1 45 line 1998 45 line 0.9 0.9 Original income 2002 Original income 0.8 Taxes 0.8 Taxes 0.7 Benefits 0.7 Benefits 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0 0 1 1 1 1 1 1 1 1 1 1 1 1 79 7 5 3 1 9 7 5 3 1 9 7 91 18 27 36 45 54 63 72 81 90 99 15 23 31 39 46 54 62 70 78 85 93 35 2.34 It is in cases like this that it becomes important to use an index of progressivity such as the Kakwani index - described in Annex ­ which measures (twice) the area between the Lorenz curve for pre-tax income and the concentration curve for tax liabilities and the concentration curve for tax liabilities(benefits receipts).12 Table 2.6 above gives the values of the Kakwani indexes for taxes and benefits over the period 1994-2002. These values confirm the intuitive conclusions derived from Fig 2.8. In 1994 the Polish direct tax system was slightly regressive but the degree of regressiviness declined over time until it became very slightly propressive in 2001. The benefit system , on the other hand, was always higly pro-poor and grew increasingly so over the period under consideration (the Kakwani index increased by nearly 2 percent). However, much of the change increase occurred between 1994 and 1998 (an increase of 14 percent). 2.35 The latter is a most interesting result, as households are increasingly dependent on social transfers to finance their consumption expenditures. Although work income is the most important source of financing expenditures, it clearly varies across the expenditure distribution (see Figure 2.9). For 2002, its share varies from 41 percent for the first decile to 68 percent for the tenth decile. Even for the bottom quintiles, income from work remains the main source of financing consumption expenditures (see Figure 2.9). Figure 2.9: Comparison of Income Composition: by Decile in 2002 (left) and over 1994-2002 for the Poor (right) Income Composition by Decile, 2002 Income Sources of the Poor 100% 350 PLN / month/ eq. adult 90% 300 80% other 70% 250 agric 60% 200 50% o.a.pension 150 40% own food 30% 100 social transf 20% 50 10% work 0 0% 1994 1995 1996 1997 1998 1999 2000 2001 2002 d1 d2 d3 d4 d5 d6 d7 d8d9 d1 consumption deciles Source: Staff calculations based on GUS's data. 2.36 However Fig 2.10 shows that the number of people relying on social transfers to close their poverty gap increased steadily after 1997 to approach 20 percent in 2002. This means that theeadcount of poverty would have been 20 percentage points higher if social assistance had not been available as a safety net. 12Kakwani (1977). 36 Figure 2.10: Poverty and Social Transfers, 1994-2002 Social transfers per capita % of population relying on transfers 104 23% to close their poverty gap 102 22% 100 21% 98 20% 96 19% 94 18% 92 90 17% 1994 1995 1996 1997 1998 1999 2000 2001 2002 19 19 19 19 19 19 20 20 20 94 95 96 97 98 99 00 01 02 Source: Staff calculations based on GUS's data. 2.37 Even though social transfers (excluding old age pensions) have a direct poverty reducing effect, the system is extremely costly and inefficient. Calculations based on household budget surveys estimate the total amount of social transfers in 2002 at 6.3 percent of GDP, which is close to what the government statistics report. However, 56 percent of this amount (or 3.6 percent of GDP) is not reaching the poor: 26 percent of this amount is in excess of the amount necessary to bring household consumption to the poverty line level (called "overflow"), while 30 percent (an equivalent of 2 percent of GDP) goes directly to non-poor households (see left-hand panel of Figure 2.11). Although benefit tightening has been tried on several occasions, the amounts leaking were of a similar order of magnitude throughout the analyzed period. Figure 2.11: Targeting of Social Transfers (Excluding Old Age Pensions) Who receives social transfers ? Social transfers per capita by consumption deciles % of resources going to: poor non-poor overflow d1 d10 d9 d8 d7 26% 44% esl d6 deci no d5 d4 30% pti um d3 d2 cons PLN / d1 month / 0 eq. adult 50 100 150 Source: Authors' calculations based on HBS 2.38 The right-hand panel of Figure 2.11 illustrates the incidence of social transfers (excluding old age pensions) from an ex post rather than ex ante perspective and reveals a disturbing pattern: 37 there is very little difference in the actual nominal per capita amount of social transfers across the expenditure distribution. Whatever quintile is evaluated, whether rich or poor, each (equivalent) individual receives on average 120-140 PLN per month. However, if these resources were given with perfect targeting, rather than being spread more or less uniformly over the distribution, they would be able to eradicate poverty completely, twice over. 2.39 The conclusion, therefore, has to be that Poland has a costly and inefficient system for distributing social transfers, only a relatively small portion of which reaches the poor. A major part of these resources is not going to the most needy. However, it might be the case that poverty reduction is not the only objective of the distribution. The other objective ­ given the even spread of these resources ­ could be the improvement of the well-being of the "median voter" in order, for example, to win support for government policies. Inflation and Exchange Rate Movements and Expenditure Composition. 2.40 It should be borne in mind that inflation is a regressive tax, the burden of which is typically borne disproportionately by the poor. For example, the poor tend to hold most, if any, of their financial assets in the form of cash. In addition, when inflation reaches a certain threshold it negatively affects investment and output growth and therefore affects employment opportunities. The composition of inflation is frequently regressive, and it is well known that food prices often tend to rise faster than the overall CPI ­ a situation that particularly affects poorer households. In addition, large exchange rate depreciations tend to hurt the poor more than others, as, again, their consumption basket tends to contain a larger component of tradables than that of richer households, and such depreciations normally negatively affect inflation. 2.41 The composition of food and non-food expenditures changed significantly between 1998 and 2002 (see left panel of Fig 2.12) with every income group spending less of its consumption on food in 2002 compared to 1998. During this period the food price index changed 15 percentage points less than the overall CPI (right panel of Fig 2.12), indicating that less of one's income needed to be used to provide for one's daily bread and butter. Apparently, the income effect outweighs the substitution effect. The divergence in average and food price indexes means that food prices are indeed important in poverty measurement (for further discussion of this issue, see Chapter 3. 38 Figure 2.12: Composition of Food and Non-food Expenditures and CPI and Food Price Developments, 1998-2002 Food as % of consumption 16 CPI and Food Price developments 1998-2001 135 70 14 130 60 12 1998 2002 10 125 50 8 120 40 6 115 30 4 2 110 20 0 105 10 -2 100 0 -4 Annual CPI e.o.p. Annual Food e.o.p 1 2 3 4 5 6 7 8 9 10 -6 CPI-Index (RHS) Food-Index(RHS) 95 Consumption deciles July: 1998 1999 2000 2001 2002 2.42 The same situation applies to developments in the real effective exchange rate and the composition of tradables versus non-tradables in each decile's expenditure pattern (see Figure 2.13). The appreciation of the zloty during the latter part of the period induced all expenditure groups to spend less on tradables and more on non-tradables, indicating once again that the income effect dominates the substitution effect. Figure 2.13: Composition of the Expenditures of Tradables versus Non-tradables and REER Development, 1998-2001 35 Non-tradables as %of consumption Real effective exchange rate 25% 135 developments 130 30 1998 2001 20% 1998-2001 125 15% 25 120 10% 115 20 110 5% 105 15 0% 100 95 10 -5% 90 5 -10% De Mar-98 Jun-98 Sep-98 De Mar-99 Jun-99 Sep-99 De Mar-00 Jun-00 Sep-00 De Mar-01 Jun-01 Sep-01 De 85 c-9 c-9 c-9 c-0 c-0 0 7 8 9 0 1 1 2 3 4 5 6 7 8 9 10 Consumption deciles Annual Change (e.o.p) (LHS) REER Index (RHS) C. POLICY RECOMMENDATIONS 2.43 In the first section we have argued that macroeconomic stability is a key component of a growth-promoting environment and is indirectly, therefore, a foundation for any successful poverty reduction strategy. The link is that macroeconomic stability encourages investment and 39 promotes productivity growth and employment creation. In this sense, macroeconomic stability is a public good. In recent years Poland's macroeconomic environment has been relatively volatile and, unfortunately, trapped in an adverse loose fiscal/tight monetary policy mix. 2.44 A labor-shedding restructuring of firms that was not matched with an environment conducive to employment creation by the private sector (particularly by small and medium enterprises) on the demand side, as well as the unchecked access to social safety nets and the recent baby boom on the supply side, have contributed to a dramatic worsening of labor market conditions, a consequent rise in unemployment, and the resulting poverty increase. 2.45 Hence, a return to robust growth (as the main factor in poverty reduction) requires a stable macroeconomic environment with a balanced fiscal-monetary policy mix. To prevent the re-emergence of economic imbalances, a fiscal consolidation and the reduction of overall deficits are needed to complement the recent monetary easing. A parallel reduction in the size of the government budget would provide greater space for private sector development. Together with an additional and necessary recomposition of government expenditures from consumption to investment, these policies would significantly strengthen Poland's growth prospects. 2.46 In order for growth to be poverty-reducing, the link between economic development and labor market improvement must be strengthened significantly (or perhaps even re-established). In other words, the growth environment must be made more labor friendly. In the short to medium term this would involve a necessary reduction in the tax wedge (payroll taxes) as well as an increase in the flexibility of the labor market. In the longer run, policies are needed that will close the mismatch of skills between labor demand and supply, which means promoting investment in human capital and education. Recent World Bank studies, particularly Public Expenditure and Institutional Review (World Bank, 2003) and Labor Market Study (World Bank, 2001), have explored the agenda and prospects for policy reforms in these areas. Their recommendations are still applicable. 2.47 Price stability is an essential component of a poverty alleviation policy. Inflation predominantly hurts the poor, who have little opportunity to protect their assets. For this reason low inflation in general and particularly low food price inflation should be perceived as a genuine achievement of recent years. In addition, the decline of nominal and real interest rates brought about by the monetary easing should work to the advantage of the poor by increasing their capacity to smooth consumption. 2.48 Fiscal policy works best if it is pro-poor. Since the poor have a higher propensity to spend, leaving more resources with them during hard times provides is an effective tool against cyclical downturns of the economy, contributes to the well-being of the vulnerable, and feeds back positively to overall macroeconomic stability. 2.49 The tax component of fiscal policy appears to perform its anticipated stabilization role. Any changes to the system to improve its pro-poor impact should be considered carefully so that they do not further increase the distortions that taxes bring to the economy, particularly so that they do not increase the overall tax burden. The elimination of exemptions (tax expenditures are directed mainly toward the rich) complemented by an overall reduction in payroll taxes in at least a budget neutral way is a plausible recommendation. 2.50 Poland has a costly and inefficient system of social transfers which is only partially poor- oriented and which does not appear to play a significant stabilization role. The major part of resources is not going to the most needy. If these funds were targeted properly, rather than being 40 spread more or less uniformly over the entire consumption distribution, they could eradicate poverty completely ­ twice over. Therefore, improving the targeting of social transfers is an immediate necessity. Poor targeting not only leaves many of the vulnerable without the help of the state, but it also skews incentives to work and poses a significant burden on government finances. 2.51 As has been demonstrated, Polish society is still changing. It has become wealthier but at the same time more polarized. Therefore, there is a strong case for good poverty/social monitoring that involves going beyond aggregations and applying more complex analytical tools, such as various decompositions, growth incidence, benefit incidence, etc. Such monitoring can provide both the public and policymakers with a better knowledge of whether and how the society is changing, and with the ability to identify the sources of such changes. It can also make it possible to assess winners and losers, and, in particular, it can become a basis for judging whether these inequalities are acceptable or not. Moreover, an ex ante assessment of the impact of social programs on income distribution and poverty could become a valuable tradition. 2.52 Despite a still positive GDP growth of about 1 percent in recent years, poverty has been increasing. In 2003 Poland stands at the threshold of EU accession with all its related challenges and opportunities, while at the same time signs of economic recovery and growth are becoming visible. Therefore, actions are required to improve the pro-poor impact of economic growth and to make this growth work for the society in general and for the poor in particular. 41 REFERENCES: Adams, R. (2003), Economic Growth, Inequality and Poverty: Findings from a New Data Set, World Bank Policy Research Working Paper 2972, Washington, D.C. Bourguignon, F. (2002), "The Growth Elasticity to Poverty Reduction: Explaining Heterogeneity across Countries and Time Periods," in T. Eicher and S. Turnovsky (eds.), Growth and Inequality, Cambridge, Mass: MIT Press. Bruno, M., M. Ravallion and L. Squire (1998), "Equity and Growth in Developing Countries: Old and New Perspectives on the Policy Issues," in Vito Tani and Ke-Young Chu (eds.), Income Distribution and High Growth, Cambridge, Mass: MIT Press. Pen, J. (1971), Income Distribution, New York: Praeger Publishers. Ravallion, M. (1995), "Growth and Poverty: Evidence for Developing Countries in the 1990s," Economics Letters, 48. Ravaillon M. (2001), Measuring Aggregate Welfare in Developing Countries: How Well Do National Accounts and Surveys Agree?, World Bank Policy Research Working Paper 2665, Washington, D.C. Ravaillon, M., and S. Chen (1997), "What Can New Survey Data Tell Us about Recent Changes in Distribution and Poverty?", World Bank Economic Review, 11(2). Ravaillon M., and S. Chen (2003), "Measuring Pro-poor Growth," Economic Letters, 78. Shorrocks, A. (1982), "Inequality Decomposition by Factor Components," Econometrica, 50 (1). Theil, H. (1967), Economics and Information Theory, North Holland, Amsterdam. World Bank (2000), Making Transition Work for Everyone: Poverty and Inequality in Europe and Central Asia, World Bank, Washington D.C. World Bank (2001), Poland's Labor Market: The Challenge of Job Creation, World Bank, Washington D.C. World Bank (2002), Transition: The First Ten Years, World Bank, Washington D.C. World Bank (2003), Poland: Towards a Fiscal Framework for Growth: A Public Expenditure and Institutional Review, World Bank, Washington D.C. 42 Annex This annex discusses three important issues concerning to the analysis presented in this chapter. First, it discusses statistical issues related to consumption data as reported in the National Accounts and the consumption data aggregated using Household Budget Surveys. Second, it points out issues related to the distribution of consumption and its normality. Third, it provides the reader with formulas for the inequality measures. SNA versus HBS The need to discuss the comparison of National Accounts (SNA) with Household Budget Surveys (HBS) is a direct implication of the principal question underpinning this study, namely, "What are the linkages between growth and poverty trends?" (generally), and "Why is poverty increasing when there is growth in the economy?" (particularly). Do the poor benefit from growth? The straightforward positive answer is obviously true ­ but on the average and in the long run. For a given country (such as Poland) a question is, "Is maximizing growth a good policy recommendation for poverty reduction?" There exists, however, a top-down view that whatever disproportionate growth there is, will ultimately reach the poor as a spillover from the rich ­ but it is no longer obvious how this would happen and what links are there to assure that this process works. Hence, a legitimate empirical question exists as to whether there is a link between growth and poverty reduction that all poverty studies should address. Figure A.2.1 leaves no doubt as to Figure A.2.1: Private Consumption Growth, whether this question is relevant for Discrepancies between NSA and HBS, 1995-2001 Poland. Despite a healthy growth of 4-5 percent per annum in 1999-2000 and a Private consumption (NSA) 9% Mean consumption per capita (HBS) slower but still positive growth of 1-1.5 8% percent per annum for 2001-02, the poverty 7% rate rose during 1999-2002. The growth 6% does not seem to have benefited the poor 5% during that period, and a conclusion that 4% (perhaps prematurely) can be drawn based 3% solely on this graph is that a massive 2% growth of some ca. 6-7 percent per annum 1% is needed to at least slightly alleviate 0% poverty. Such growth is not expected in -1% the near future, and therefore other factors -2% must be analyzed and policies must be put 1995 1996 1997 1998 1999 2000 2001 2002 in place to address the problem. Source: Staff calculations based on GUS's data. The first straightforward suggestion is to check the "accounting". For example, differences in the inflation of food and non-food prices recorded recently in Poland, were important in sustaining the purchasing power of the poor, who usually have a larger share of food in their consumption baskets. Failing to incorporate this effect overstates the increase in poverty. (This issue is discussed in a separate chapter of this study.) In addition, for a clearer view, it is important to compare si factors ­ not GDP and poverty rate, but rather private consumption from SNA and household consumption from HBS, as in Figure A.2.1. The figure indicates that SNA private consumption and HBS consumption do 43 not appear strongly correlated. Moreover, the mean household consumption dynamic was negative during 1999-2002, therefore, so it should be intuited that poverty has increased. There are multiple reasons why SNA and HBS aggregates are different. First, the consumption registered in HBS does not generally include non-market value added generated and consumed within households, one important component of which is housing. Households that own their apartment are accounted for as renting the apartment to themselves for an implicit rent, which constitutes a part of GDP. Second, consumption aggregates for welfare analysis do not usually include expenditures on durables, which leaves out about 18 percent of total goods and services consumption. Third, there is a difference in coverage ­ for example, HBS excludes group lodging facilities (hotels, student houses). Moreover, private SNA consumption is sometimes13 a residual item in SNA, thereby picking up any error in any other component of GDP. Last, but not least, there is a potentially large measurement error in both aggregates. For comparison, SNA private consumption in 2002 was PLN504 billion, while the estimate of total consumption from HBS yields only PLN290 billion. These explanations are sufficient when we are concerned with levels, but there is little reason why the dynamic should be different if the composition of these aggregates remains roughly unchanged (which is a plausible assumption). Possible explanations include an inflation difference (CPI versus SNA private consumption deflator) and ­ again ­ a measurement error. This is neither a new nor a country specific problem. These discrepancies have not yet been reconciled. Adams (2003) shows that SNA and HBS derived growth figures (for developing countries) move in the opposite direction in one out of three cases. Adams also points out that economic growth, as measured by GDP per capita, was on average much stronger, a situation that is consistent in the case of Poland. On the other hand, Ravallion (2001) cannot reject the hypothesis that in the long run GDP growth is an unbiased (although a very volatile) estimator of household consumption for the whole sample, but, ironically, he rejects this hypothesis for ECA countries. We decided to measure growth by a change in survey mean, what is justified by the fact that if growth and household consumption are derived from one source then any measurement error should be positively correlated between and within years and should cancel out to a great extent. The use of two different sources would make the error additive and would obscure the meaning of the results. Therefore, having posed the question of why poverty rises if there is growth, it is tempting to blame statistical discrepancies and to claim that there is no poverty reduction because there is actually no growth (in mean HBS consumption). However, this would be, a very premature conclusion. Distribution of Consumption It is key to realize that growth is not just a change in mean but is a complex process of distributional dynamics, including change in shape (growth, spread) as well as intra-distributional shifts (mobility).14. Figure A.2.1 presents Poland's consumption (and log consumption) distribution for 2002 across this population. Consumption is defined as a per equivalent adult real (CPI-deflated) consumption based on appropriate consumption aggregation. The vertical line 13Although not in the case of Poland. 14However, the unavailability of a good, long and comparable panel confines this study to distribution changes only, thus leaving out income mobility. 44 indicates the poverty line set at about PLN344 per month (2001 prices).15 Table A.2.1 shows the basic characteristics of these distributions. Figure A.2.2: Distribution of Consumption and Log Consumption, 2002 yti 0.0018 0.9 yti dens 0.0016 0.8 dens 0.0014 0.7 2002 2002 0.0012 poverty line 0.6 poverty line 0.001 0.5 0.0008 0.4 ` 0.0006 0.3 consumption, 0.0004 0.2 PLN / month / eq. adult 0.0002 0.1 log-consumption 0 0 0 215 430 645 860 1075 1290 1505 1720 1935 2150 2365 4. 4. 5. 5. 5. 6. 6. 7. 7. 7. 8. 8. 5 86 22 58 94 3 66 02 38 74 1 46 Source: Staff calculations based on GUS's data. Table A.2.1: Distributional Characteristics, 1994-2002 1994 1995 1996 1997 1998 1999 2000 2001 2002 Mean consumption 578 571 593 611 639 636 634 627 625 Mean log consumption 6.26 6.25 6.29 6.31 6.35 6.34 6.33 6.32 6.31 Std log consumption 0.44 0.43 0.44 0.46 0.46 0.47 0.48 0.48 0.50 Source: Staff calculations based on GUS's data. A first glance at the graphs reveals that (as one could expect) consumption is almost ideally log normally distributed, and (hence) a log of consumption is normally distributed. Closer examination reveals that cumulative differences from normality do not exceed 1 percent for the entire part of the distribution and for most years (1994-2002) in the sample period. There seems to be a tendency, while addressing growth-poverty issues, to talk about a clustering of the poor around the poverty line. This may be another way of expressing the supposition that poverty is shallow. The clustering theory gives the impression that only a little growth is needed to pull this cluster above the poverty line and to almost completely eradicate poverty. However, since, if there are no major distortions, consumption is by nature (log) normally distributed (not only in Poland), there is no particular clustering beyond what is normal. Almost by default, poverty is not particularly more shallow than in any other typical country at a similar stage of development and similar social equity. Therefore, "a little growth" would not pull this cluster above the poverty line. 15Equal to the adjusted social assistance eligibility threshold, average for 2001. 45 Inequality Measures The most common inequality index used is the Gini coefficient: G = 2 n i - µ n2 i=1 n +1ci , where i is a rank index ascending in consumption ci and µ 2 is mean consumption (Shorrocks, 1982). The Gini index ranges from 0 (perfect equality) to 1 (perfect inequality). The Gini index is decomposable into components. Denoting ci = c k k n , i ; µk = c i k k 2 n a contribution of component k equals to Sk = µk G , where is G = i - k µ µk n2 i=1 n +1ci 2 a concentration index of component k, called also a pseudo-Gini (because ranking i is still along consumption). Other commonly used measures are members of the so-called Generalized Entropy class: -1. GE() = 1 n ci n( -1) i=1 µ Substituting, for example, equal to 0 and 1, we obtain two Theil's measures of inequality (Theil, 1967): a mean log deviation and a Theil index, respectively. These measures differ in sensitivity to various parts of the distribution, the lower/higher the , the more the lower/upper part of the distribution is emphasized. In the paper a mean log deviation is used. GE(0) = 1 n . n lnc i=1 µ i 46 3. THE IMPORTANCE OF PRICES IN MEASURING POVERTY, THE CASE OF POLAND, 1994-2001 16 Pierella Paci and Marcin J. Sasin A. INTRODUCTION 3.1 One of the puzzles of Poland's current socioeconomic situation is the fact that an unbroken record of growth since the end of the transition recession, and particularly in the second half of the 1990s, has not transformed itself into a distinct decline in poverty ­ at least as measured by governmental (as well as nongovernmental) statistics. This situation is illustrated in Figure 3.1. Figure 3.1: GDP Growth and Poverty Rates, 1994-2001 GDP growth (left Relative poverty rate (right Below subsistence Poverty rate "legal" (right 8% 18% 7% 16% 6% 14% 5% 12% 4% 10% 3% 8% 2% 1% 6% 0% 4% 1994 1995 1996 1997 1998 1999 2000 2001 Source: Staff calculations based on GUS's data. 16This is a background paper for Poland's Living Standard Assessment, 2002/2003, prepared initially for presentation at the annual World Bank DEC Economist Forum, April 10, 2003, Washington, D.C. 47 Poland's Macroeconomic Situation 3.2 The start of Poland's transition in 1990 was marked by exceptionally difficult macroeconomic conditions, which included high inflation, a large legacy of external debt and a burden of insolvent state-owned enterprises. By taking the risks of making the currency convertible, lowering import barriers, and imposing hard budget constraints on economic enterprises as well as through consistent economic policies, Poland became a leader among European transition countries, with growth averaging 5 percent per annum in the mid-1990s. 3.3 However, after 1998, Poland saw a new round of enterprise restructuring as a response to the loss of FSU export markets and to progressively tightened monetary conditions (with real interest rates reaching 13 percent in 2001). The resulting improvements in productivity were brought about for the most part by reducing employment. This, together with the increased numbers of newcomers in the labor market, owing to the baby boom of the early 1980s, has led to unacceptable increases in unemployment (to 19.7 percent of the labor force -- ILO definition -- as of end-2002). A global slowdown and a collapse in domestic investments only contributed to the deceleration of economic activity, with growth rates falling to below 1 percent annually in early 2002. Poverty in Poland 3.4 These unfavorable developments have brought poverty issues back to the front line of attention. Depending on the measure used, Polish poverty in 2001 varied between 9 and 17 percent of the population. The lower line in Figure 3.1 depicts the poverty rate according to the subsistence minimum criterion ­ consumption below this level leads to physical deprivation. The threshold is computed each year by an independent research institute, based on the requirements for calorie intake. The middle line ("legal") is a Polish social assistance eligibility criterion and is set administratively. The upper line measures poverty in a relative way ­ the threshold is 50 percent of an average consumption.17 3.5 According to these methods, we do not see a decline in poverty in the second half of the 1990s, nor do we see any direct link between growth and an improvement in living standards as measured by consumption. 3.6 For relative poverty rates this trend might be to some extent explicable (provided we assume that rising inequality is an inherent characteristic of a transition process), since relative poverty is more or less directly related to measures of inequality. On the other hand, as the initial transformation is over, these inequalities (and hence the relative poverty) should be overridden by a growth effect and by increased participation in the market economy. Poverty rates obtained by applying the two absolute measures (meant to present an objective picture) leave much room for doubt as to whether Polish growth is a pro-poor growth. Direction of Research 3.7 There may be many explanations for this phenomenon: jobless growth, skewed incentives to participate in the labor market, a vicious cycle owing to the inter-generational heritage of poverty and deprivation, regional decline, peripheralization, etc. 17If not stated otherwise, consumption here means equivalent consumption per capita. 48 3.8 This discussion takes a path as yet unexplored (in Poland): it attempts to investigate the importance of regional price differences for poverty measurement (and hence for poverty reduction policies). Taking into consideration the fact that growth in Poland is unevenly distributed, with growing disparities between regions, it could well be that differences in prices compensate partly for differences in nominal income levels. Given that the poorer regions should be the regions that are growing slowly, we should find more poor people in the areas in which prices are lower, and hence applying the national average would overstate the level of poverty and perhaps skew the poverty profile. If price differences are widening with growth they should also be able to partly offset the supposed growth of poverty. Another channel in which prices can influence poverty dynamics is the effect of a different composition of consumption baskets between the poor and the non-poor, particularly regarding the share of food. Since the poor are known to have a higher share of food in their consumption basket, the relative well-being of the poor versus the non-poor changes when the food price index and the consumer price index follow divergent trajectories. 3.9 This chapter analyzes both the spatial and the time dimensions of prices in an attempt to improve poverty measurement. The rest of the chapter is organized as follows. Section B explains the data and methodology used. Section C explores the consequences of applying regional (and household specific) price deflators to the poverty profile (spatial dimension), while Section D analyzes the implications of the evolution over time of prices (time dimension). Section E provides a summary and conclusions. B. DATA AND METHODOLOGY Household Budget Surveys and Price Statistics 3.10 This analysis is undertaken on the 1993-2001 rounds of the Household Budget Surveys (HBS) conducted by the GUS on a rotating yearly sample of about 32,000 households, each of which was surveyed for a period of one month. The data are presented in a breakdown into 16 administrative entities (called voivodships) and 6 types of "place of residence" (rural/urban and 5 classes of towns, by size); there is, therefore, a set of 6 x 16 = 96 theoretically possible combinations of these two variables (each of them, for the purpose of this study, we will call a "region" or a "locality"). Since not all voivodships have cities of all sizes in their territory, the actual number of possible voivodship/city-size combinations is 73. 3.11 The data are representative for the whole country and for the rural/urban/city-size dimension. The representativeness at the voivodship level is limited. Although the data are not representative at the disaggregation of the 73 regions, this should not prevent us from going ahead with the analysis as long as we refrain from judgments on a given, particular locality. 3.12 We use two sources of data for prices. First, HBS collects, where possible, the information on quantities consumed, thus enabling us to compute unit values. We treat the median unit value within a region as a regional price and divide it by the median unit value for the whole sample to obtain a relative price. Where not enough observations (i.e., less than five) are available within a region for a given category, we substitute the overall regional relative price. To avoid adding nails to hammers, this procedure is applied only to consumption categories that are sufficiently homogenous -- that is, mainly to food, which, however, constitutes about 50 percent of the whole consumption basket. Relative food prices are derived from HBS for each year under analysis. 49 3.13 Second, prices of other categories are derived from officially published statistics. Although GUS does not compute voivodship-specific relative prices, it publishes (since 2000) absolute prices of selected goods by voivodship and by a rural/urban/city-size breakdown. By matching these prices to the COICOP classification used in HBS we were able to derive relative prices for non-food items for 2000 and 2001. Unfortunately, this was not possible for all years and we had to assume that the structure of non-food prices remained constant through the 1990s. Finally, there is a group of goods and services ­ usually delivered by (natural) monopolies ­ for which prices are equal everywhere. This includes, for example, postal services, pipeline gas, telecommunications, etc., and constitutes about 10 percent of the consumption basket. Consumption Aggregate 3.14 This discussion applies an "income" (or "monetary") concept of poverty, and so we treat (price adjusted) household consumption expenditure as a measure of well-being. There are known caveats to calculating a proper consumption aggregate (see, for example, Deaton and Zaidi, 2002, for an excellent overview) with which we have dealt in the standard way. 3.15 The first caveat relates to rental charges. Since some inhabitants own their apartments while others have to pay rent, both groups derive similar utility from the lodging but one appears richer than the other. Data collected by HBS do not allow us to impute rents, and consequently, for the purposes of this study, we decided to leave all rent payments out. 3.16 The second limitation is related to the treatment of durables. It is usually not the possession of a car or a television set per se that makes a person better off, but rather the flow of utility associated with owning and/or using the said durable (using a car, watching television). The first best solution would be to calculate the "use"-consumption derived from durables (as proxied by depreciation). As with the previous caveats, this is not possible using the information contained in HBS: therefore, following the advice of Deaton and Zaidi, we have left all durables out in our calculations. 3.17 To control for household size we apply the standard OECD equivalence scale (first adult: 1; other adults: 0.7; children: 0.5). One insignificant exception is made in order to conform with Polish social assistance practices: one-person households were assigned an additional 0.1 consumption units. 3.18 Figure 3.2 presents the cumulative distribution of per capita equivalent consumption. For example, point 490 (40%) means that 40 percent of Poland's population has a consumption lower than 490 PLN per month. Poverty Line 3.19 We use the Polish social assistance eligibility threshold as our benchmark poverty line. The advantages to using this measure are that it is simple (exogenous and published in nominal figures) and appealing ("owned" by policymakers and better understood by the public). On the disadvantage side, it is set administratively and is not always kept constant in real terms. 3.20 There is a need to adjust the poverty line to bring it in line with the method of construction of the consumption aggregate. Initially, the poverty line is expressed in terms of monthly income, which for the poor should closely correspond to monthly consumption. But, as we leave durables and rents out, the line should be adjusted accordingly. Since the poverty rate in Poland is about 15 percent, we choose the scaling factor to equal roughly the ratio of total to 50 adjusted consumption at the fifteenth percentile (i.e., 88.5 percent). In this way we do not change the poverty headcount by changing the consumption aggregation method. 3.21 To sum up, the poverty line for 2001 is equal to 344 PLN per consumption unit per month, which is equivalent to about 80 current US dollars or about 180 PPP adjusted constant 1998 international dollars. The line is depicted in Figure 3.2, according to which Poland had a poverty rate (headcount) of 15.6 percent in 2001. Figure 3.2: Cumulative Distribution of Consumption 100% 90% 80% noi 70% 60% opulatpfo 50% % 40% 30% 20% 10% PLN 0% 80 220 360 500 640 780 9201060 1200 1340 1480 Source: Staff calculations based on GUS's data. C. ACCOUNTING FOR PRICES: SPATIAL DIMENSION Empirical Evidence of Price Differences 3.22 There are solid grounds, both theoretical and empirical, for including prices in the living standard measurement. 3.23 Figure 3.3 presents the results of a computation of relative regional price indices in the breakdown into 16 voivodships (left panel) and 6 residence classes/sizes (right panel); the capital district and its residence class have been highlighted. We notice a moderate differentiation between voivodships (+/-3.5 percent) and a more significant one along urban/rural-large/small- town dimensions (from ­4 percent to +9 percent). The limited success of the inclusion of regional prices in poverty measurement so far (see, for example, World Bank, 1995) may be a consequence of relying only on inter-voivodship differences. Moreover, if we include the combination of both (see Figure 3.4) the variation in prices would range from ­5 percent to +11 percent. Figure 3.4 also supports the supposition that regional prices matter for poverty analysis, as they apparently correlate with regional poverty rates. 51 Figure 3.3: Relative Prices, by Voivodships (left) and by Town Size (right) 4.0% 10% 3.0% 8% 2.0% 6% 1.0% 4% 0.0% 2% -1.0% ` 0% -2.0% -2% -3.0% -4% -4.0% -6% k l j l i d e o h s k m u Pd Pd m Ku Lb 500T+ 200T+ 100T+ 20T+ 20T- Rural Wr Sw Lo Wi Op Zc Dl Sl Mal Po Lb Maz Source: Staff calculations based on GUS's data. Figure 3.4: Poverty Rates versus Regional Price, 2001 35% 30% y = -1.4356x + 0.1523 R 2 = 0.4897 etarytr 25% veoP 20% 15% 10% 5% 0% Deviation from national average -6% -4% -2% 0% 2% 4% 6% 8% 10% 12% Source: Staff calculations based on GUS's data. Theoretical Background 3.24 We try to measure welfare as a money metric utility, given by: Um = h p 0 h kqk (1) k which means that we value quantities of goods k consumed by the household h (qhk) by comparable (reference) prices (p0k) ­ in our case national averages. However, since quantities are not always reported we have to circumvent this by applying household specific price deflators; namely, we use the Paasche-type index which in this case is defined as: 52 p h h k qk Ppaasche = h k p 0 h k qk k (2) which uses prices actually faced by the household (phk). The term in the nominator equals the total household expenditures on consumption. Combining (1) and (2) we obtain: Um = h Exph Ppaasche h (3) We can alternatively express the household specific price deflator as: Ppaasche = h Exph = 1 = 1 p 0 h 0 0 kqk h k pk pk qk h h w pk k pk h k k pkh (4) Exph where whk mean a share of good k in the household's consumption basket. In order to apply (4) we no longer need quantities consumed by a household; however, we need the (relative) prices it faces (phk). These can be approximated by relative regional prices (i.e., by the ratio of median unit values). 3.25 We are aware that unit values may not precisely reflect prices (see, for example, Deaton, 1997) and that by this application we assume the average quality of goods between localities to be the same. This, of course, might not necessarily be the case, but we believe that the quality shading effect should have a lower magnitude than differences in local prices. 3.26 It is important to acknowledge that expression (4) not only corrects for relative price differences but also corrects for different consumption patterns between households. Weights whk ensure that the higher the share of good k (say, food) in the household's basket is, the higher is the importance of food prices for this household. The index is computed for each household individually, hence, rather than a "regional price" it should be called a "household specific deflator." Figure 3.5 presents its empirical distribution. 3.27 We see that the index takes values between 0.8 and 1.2, implying fairly significant variation. At the same time the distribution is skewed to the left, meaning that more people face slightly cheaper prices while some people have more expensive consumption. Results 3.28 We proceed to analyze how applying these adjustments affects consumption and poverty. Table 3.1 presents results for the basic statistic: the poverty rate. It turns out that accounting for household specific prices corrects the poverty rate from 15.6 percent to 14.6 percent (i.e., by 1 percentage point or, equivalently, by about 7 percent of the total poor). This is not surprising, as we have already established that poorer regions tend to have lower prices. The difference is significant at the usual level (standard errors in parentheses). 53 Figure 3.5: Distribution of a Household Specific Price Index .1 .08 .06 Fraction .04 .02 0 .8 .85 .9 .95 1 1.05 1.1 1.15 1.2 Source: Staff calculations based on GUS's data. Table 3.1: Results for the Poverty Rate Initial profile Price adjusted Significance Poverty rate 1=0.1558 2=0.1460 1= 2 (0.0020) (0.0019) prob=0.0000013 Source: Staff calculations based on GUS's data. 3.29 Figure 3.6 plots (a part of) the initial and the price adjusted density estimate of consumption, scaled for better inference by nominal value of the poverty line (344 PLN). We see that the area below the distribution left of the line shrank, meaning that the actual poverty rate is lower than it seems as the distribution is located more to the right (towards higher values of consumption). The area between the profiles (and left of the poverty line) is exactly equal to 15.6 percent -- 14.6 percent = 1 percent (i.e., to the correction we made to the poverty count). Figure 3.6: Impact of Regional Prices on the Distribution of Consumption 0.6 0.5 noitca Fr Regional price corrected 0.4 Initial poverty line 0.3 0.2 0.1 Multiples of poverty line, PLN 0.0 11 17 22 28 33 38 44 49 55 6 2 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 66 71 76 82 87 93 98 04 09 14 0. 0. 0. 0. 0. 0. 0. 1. 1. 1. 1. 25 31 36 41 47 1. 1. 1. 1. 1. Source: Staff calculations based on GUS's data. 54 3.30 The modification that we introduced into the picture affects not only the poverty rates but also the measure of poverty depth. The poverty gap, defined here as the amount of money that, if transferred to the poor assuming perfect targeting, would pull them out from below the line (thus eradicating poverty completely), is also lower by about 8 percent. 3.31 Figure 3.7 shows this correction more illustratively. It presents the inverted cumulative distribution of consumption. For example, point (11, 310) means that 11 percent of the total population has a monthly consumption below 310 PLN. The thick horizontal line depicts the poverty line (=343 PLN) while the two vertical lines indicate the two calculated poverty rates: initial 15.6 percent and adjusted 14.6 percent. The area between the poverty line and the distribution reflects the poverty gap. The shift upward means that the poverty gap is actually narrower and the area between the profiles (below the poverty line) is exactly equal to the correction we made to the poverty depth. 3.32 More important, however, than simple headcounts and gaps is whether and how the adjustment alters the picture of the poverty profile. It would be straightforward to expect that people living in rural areas and small towns should turn out to be somewhat better-off after the correction. 3.33 We approach this by calculating the so-called relative poverty risks (see, for example, Luttmer, 2000). The relative poverty risk for a given group is the ratio of a poverty rate within the group to the poverty rate for the whole population, and indicates how many times more probable it is for a member of one group (knowing nothing else) to be in poverty compared to the average individual. Figure 3.7: Regional Prices Correction to Poverty Depth 400 300 NLP,noi umpt Initial distribution Regional price corrected 200 onsC Ranking of individuals by consumptions (percentiles) 100 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Source: Staff calculations based on GUS's data. 3.34 Table 3.2 presents very rough, univariate and unrefined estimates of the relative risks for various groups, demonstrating that poverty in Poland has typical correlates: low education, rural residence, unemployed household head, etc. 3.35 As would be expected, the application of regional prices adjusted downwards lowers the poverty risk associated with living in rural areas. The most obvious group for which the poverty rate seems to have been overestimated is the farmers. The correction brought their poverty risk from 1.3. to 1.1 (i.e., significantly closer to the unconditional mean). 55 Table 3.2: Estimates of Relative Poverty Risks for Various Groups Characteristics Initial profile Price adjusted Difference Rural 1.43 1.27 0.16 Urban 0.72 0.72 0.00 Hh head unemployed 3.47 3.56 -0.09 Higher education 0.11 0.10 0.01 Secondary level educ. 0.50 0.48 0.01 Lower or no educ. 1.42 1.33 0.10 Worker 0.84 0.87 -0.03 Worker and farmer 1.22 1.16 0.06 Farmer 1.31 1.11 0.20 Self-employed 0.58 0.56 0.02 Old age pension 0.61 0.61 0.01 Disability pension 1.40 1.41 -0.01 Other (social assistance) 2.64 2.72 -0.08 Source: Staff calculations based on GUS's data. D. ACCOUNTING FOR PRICES: THE TIME DIMENSION Regional Prices through Time 3.36 The analysis of regional prices could be augmented by a time dimension. If relative price levels between localities are not constant through time and if they move in the opposite direction from relative regional growth rates, this situation could reduce inequalities and perhaps also alleviate the problem of poverty in less growing regions. 3.37 In this discussion we limit ourselves to merely sketching out this issue. Figure 3.8 (left panel) plots relative prices in 2001 (horizontal axis) against similar relative prices seven years earlier in 1994 (vertical axis). There is a 45° line, aligning along which would mean no change in relative price levels. Although most regions tend to cluster in a cloud around the center, the average direction, as indicated by a thick line fitted between the points, suggests that there are apparent divergences in price levels. The right panel of Figure 3.8 complements the picture by tracking the standard deviation of regional prices. Figure 3.8: Regional Prices, 1994 versus 2001 (left) and the Standard Deviation (right) 15% y = 0.7662x + 0.0029 4.5% R2 = 0.7154 10% 4.3% 5% 1994 ni 4.1% 0% onsi ativeD 3.9% -5% -10% Deviations in 2001 3.7% -10% -5% 0% 5% 10% 15% 1993 1994 1995 1996 1997 1998 1999 2000 2001 Source: Staff calculations based on GUS's data. 56 Accounting for Food Price Inflation 3.38 In this discussion, we will focus on yet another channel through which prices could affect the poor, namely, on the price divergence across food and non-food items (shares of these components being unequal between the poor and the non-poor). 3.39 Figure 3.9 presents the known fact ­ called the Engel law ­ that the share of food in the budget declines together with rising income. According to the HBS (our consumption aggregate), in Poland the first 15 percent percentiles of the population devote two-thirds (69 percent in 1993 and 64 percent in 2001) of their budgets to food, while for the whole population this ratio is about 45 percent. Because weights in the CPI basket are more closely aligned with the consumption basket of an average household, the CPI might not be appropriate to deflate the consumption in order to compare the relative well-being of the poor. Figure 3.9: Food Share in Consumption by Half-decile 1 0 0 % 9 0 % 8 0 % 7 0 % 6 0 % 5 0 % 4 0 % 3 0 % 2 0 %F o o d H o u s in g O t h e r 1 0 % 0 % 1 3 5 7 9 1 1 1 3 1 5 1 7 1 9 Source: Staff calculations based on GUS's data. 3.40 In Figure 3.10, the left panel shows the development of prices in Poland over the recent decade. We see that the country has gone through a gradual but successful period of disinflation. It is interesting and very relevant to note that from 1997-98 food and overall inflation started to diverge, then for a brief period they returned to a parallel path, and then food prices nosedived. For almost all of 2002, food price deflation continued. 3.41 We guess that during high inflation periods price differences matter less, since they are wiped out more rapidly and it is only when inflation is brought down to around one digit that price levels and their development gain in importance and make a difference. 3.42 In parallel to the CPI we have hence constructed a "Poor Price Index" ("Poor PI") using the weights relevant to poor people ­ that is, with greater emphasis on food. The right panel of Figure 3.10 graphs and compares the two indexes in both levels and differences. Their trajectories start to obviously diverge from 1998, which is reflected in a steep increase in their ratio. Between 1998 and 2002, that is, in the time-span of five years, the relative purchasing power of the poor improved by 5.5 percent. 57 Figure 3.10: "Poor PI" versus CPI (left) and Food Price Inflation versus CPI Inflation (right) 40% CPI (lef t), av erage 1995=1 PoorPI (lef t), av erage 1995=1 35% 5% 2.5 ratio (right) 30% Cons umer pric es Food pric es 4% 25% 2 3% 20% 1.5 2% 15% 10% 1% 1 5% 0% 0% 0.5 -1% 94 95 96 97 98 99 00 01 02 93 94 95 96 97 98 99 00 01 02 -5% 19 19 19 19 19 19 20 20 20 19 19 19 19 19 19 19 20 20 20 Source: Staff calculations based on GUS's data. 3.43 Our benchmark in the previous section was the Polish social assistance eligibility threshold. Poverty figures obtained by its application are commonly accepted as "poverty in Poland." It would, however, be appropriate to verify whether this threshold is constant in relevant real terms and to make sure that we apply the same yardstick. Figure 3.11 compares the path of an absolute legal poverty line with the path it would have resembled had it been adjusted by the appropriate poor-specific price index ("Poor PI"). We see that from its introduction in September 1996 the legal threshold has been rather "generously" updated, which cumulated in about a 14 percent real difference at the end of 2001. 3.44 Figure 3.12 recapitulates the discussion. It depicts the simplest measure ­ the poverty headcount ratio ­ with and without adjustments developed above. Disparities appear to be quite meaningful. According to the corrected figures, there was actually a rather significant decline in poverty rates during the high growth years of 1995, 1996 and 1997 (compare, for example, with Figure 3.1). During 1998-2001, when growth gradually stagnated the (adjusted) poverty rate increased, however at a slower pace than it initially seemed ­ helped by food price stability. Figure 3.11: Properly Deflated Social Figure 3.12: Poverty in Poland (adjusted) Assistance Threshold versus Actual s tandard (c pi, national av er v age) 18% 1.6 proper ly def lated ( ac c ount. f or f ood) 17% fur ther , regional dif f- adjus ted 1.5 Actual line 16% PoorPIdeflated 1.4 15% 1.3 14% 13% 1.2 12% 1.1 11% 1 10% 1996 1997 1998 1999 2000 2001 94 95 69 97 98 99 00 01 19 19 19 19 19 19 20 20 Source: GUS, Ministry of Economy, Labor and Social Source: Staff calculations based on GUS's data. Policy and Staff calculations. 58 E. CONCLUSIONS 3.45 This discussion argues that accounting for price divergences, both regional and through time, is important for measuring and understanding poverty and its link to growth. Although it is less important whether the actual poverty was 16 percent or 15 percent in 2001 ­ it is important that, in general, the headcount, profile and dynamics need a slightly more than moderate correction. 3.46 The application of this conclusion to the case of Poland reveals that poverty rates seem to have been overstated by about 1 percentage point, or, equivalently, by 7 percent of the total poor, when using national averages. Also, lower relative regional prices, observed mostly in rural areas and small localities, are correlated with regional poverty rates and, because of that, the problem of poverty in rural areas has been relatively exaggerated, particularly among farmers. Moreover, poverty dynamics need to be corrected for a divergence in inflation rates between food and non- food items, which in recent years worked to the advantage of the poor. The analysis presented here stresses the importance of inflation (and food price inflation in particular) in poverty reduction. 3.47 There are also policy implications to the analysis. For example, a calculation of (changes in) the purchasing power parity of the poor could help in the proper adjustment of social assistance eligibility thresholds. Second, price disparities between regions mean that the replacement ratios between work income and social transfers are different in different parts of Poland, which can have important implications for labor market policies. 59 REFERENCES Deaton, A. (1997), The Analysis of Household Surveys, Johns Hopkins University Press, Baltimore, MD. Deaton, A., and S. Zaidi (2002), Guidelines for Constructing Consumption Aggregates for Welfare Analysis, LSMS WP 135, World Bank, Washington D.C. Living Conditions 2001, Central Statistical Office, Warsaw. Luttmer, E. (2000), "Poverty and Inequality in Croatia," in Croatia: Economic Vulnerability and Welfare Study, World Bank, Washington D.C. Understanding Poverty in Poland (1995), World Bank, Washington D.C. 60 4. POVERTY IN POLAND: PROFILE, 2001 AND CHANGES, 1994-2001 Irena Topiska and Karol Kuhl 4.1 Poverty studies are numerous in Poland. These studies provide a variety of results, using both quantitative and qualitative methods of investigation. To date, however, certain topics have been neglected, including important methodological concepts and some empirical aspects of poverty assessment. Two of these topics are considered in the poverty analysis presented in this chapter. 4.2 First, a new approach to the measurement of the standard of living is explored. The "total expenditure" concept of the evaluation of the standard of living is replaced by the "consumption flow" concept. While this approach has been adopted for poverty assessment in various countries, it has not yet been applied in Poland. 4.3 Second, regional and urban-rural differences in the standard of living are taken into account. More precisely, the impact of regional price differentiations on poverty has been considered. This issue has not received adequate attention in studies on poverty in Poland despite the fact that various questions regarding human and economic well-being in the regions have often been discussed. 4.4 These methodological concerns have influenced the investigation and the content of this chapter. For 2001, as many as four measures of the living standard have been used for poverty evaluation. The results concerning all of them are reported and discussed at length, to provide better insight into the poverty profile and to allow for studying the impact of the methodological approach adopted. However, in the investigation of poverty over the eight-year period 1994- 2001, new concepts have not been fully explored. Because of data constraints, only the regional aspects have been fully included in this temporal analysis. The consumption flow concept has in fact been only partially approached. 4.5 The chapter starts with a methodological section, followed by three main parts that discuss the key results of the investigation. The first of these parts presents the poverty profile in 2001. The second (section C) sketches the changes in 1994-2001. The third (section D) contains the results of a longitudinal analysis conducted on two panel sub-samples: 1994-1996 and 1997- 2000. The text is summarized by major findings and is supplemented by a methodological annex (as well as the Statistical Appendix to the report), which provide some details with regard to the methodology as well as tables showing poverty figures that are not fully covered by the text. 61 A. DATA SOURCE AND METHODS Data Source 4.6 Household Budgets. The investigation is based on household budget surveys (HBSs) conducted by Glówny Urzad Statystyczny (GUS). Since 1993, the annual HBS sample contains over 31,000 households, or about 100,000 individuals. The sample's size is one of its positive features. It is large enough to provide various profiles which reflect social, regional or demographic characteristics of individuals. It also allows for making multi-layer classifications of households. 4.7 It should be pointed out that in the Polish HBS, monthly rotation of households is effective. In general, each household is surveyed during the period of one single month: thus, the annual sample consists of 12 monthly sub-samples of fairly equal size (over 2,500 each).18 4.8 Survey Representativeness. In principle, HBS is representative for virtually the whole population, excluding those who live with foreigners or outside of "standard" households (in dormitories, for instance). The non-response rate of the Polish HBS is quite high (49 percent in 2001) and refusals reach 25 percent. As a result, HBS loses its (assumed) representativeness. In order to regain it, the sample weights are established. In this research, GUS weights adjusted to the sample size have been thoroughly applied. It should be remembered, however, that the Polish HBS sampleas most similar surveyslooks underrepresented with respect to specific sub- populations, such as excluded or marginalized groups, and, on the other hand, the most affluent and powerful. And the HBS reveals a certain bias toward the upper end of the standard of living distribution. This may be seen when sample indicators of, for example, unemployment or housing equipment are compared with indicators for the whole population. Such indicators are usually slightly better for HBS. 4.9 Data Content. The Polish HBSs provide data on household income (from various sources), expenditures (over 200 items), type of residence, housing conditions, durables (assume 30), land area (if any), etc. Fairly detailed information is also given for each household member: age, sex, education, relation to the household head, and individual source of income (for working persons: type of employment, sector and branch of the economy; for social beneficiaries: benefit type and its level; for the unemployed: also unemployment duration). At the household level, various categorizations submitted by the GUS are also available. They include the so-called socioeconomic category of a household (7 categories according to the main source of income), family type (over 10 categories), place of residence (6 groups), etc. 4.10 Two comments that are important for the analysis which follows should be made at this point: (i) the money value of consumption in kind (food and goods received free and/or produced by the household) is estimated by households with GUS assistance and is included in the total expenditure; (ii) data on durables are very limited. The HBSs provide information only on whether the household has a particular type of durable, and how many items--but do not provide information on their quality (estimated value, date of purchase, brand name, etc.). 4.11 Monthly Rotation Problems. Monthly rotation of households in the HBS produces at least three problems that must be dealt with: seasonality, inflation, and coverage of infrequent expenditures. Seasonality has not been handled in this research. This would be crucial for the 18However, a given fraction of households re-enters the sample in the same month for several years (up to four times in 1993-2000, up to two times since 2000). This reflects a rule of partial rotation of the sample. 62 investigation based on incomes, which may fluctuate significantly, but it seems less important if the total consumption flow approach is used. 4.12 The inflation question has been given more attention. Although recently inflation has been rather moderate (3.6 percent in 2001 measured with CPI, December to December) and its rate is declining, it was fairly high in the past (29.4 percent in 1994, and 21.6 percent in 1995, December to December CPIs). In order to avoid the impact of inflation, all items registered in HBSs in current zlotys of a given year have been expressed in June prices of that year. Separate monthly CPI deflators have been used for 13 main expenditure categories (such as food, clothes, health) and the overall CPI has been applied if specific deflators have not been available. 4.13 Infrequent expenditures are handled in part by the GUS. For about 30 items, including selected durables, health treatment, and holiday trips, data on expenditures during a quarter (instead of a month) are collected. With regard to these items, the HBSs report one-third of quarterly expenditures. Some problems, however, have remained unsolved. These pertain to expenditures on a variety of vital goods and services which are bought with a frequency lower than a month, but which are not covered by a quarterly questionnaire. The most important example includes spending on energy (coal, gas, electricity).19 As a result, quite a number of households report no charges while other households seem to spend a lot.20 However, this problem could not be solved given the data available.21 Measures of the Living Standard 4.14 In this research, three questions have been approached with regard to the measurement of the living standard (LS): (i) the general concept; (ii) regional differentiation; and (iii) the role of household composition. 4.15 General LS Concept. Until now, poverty studies in Poland have used total expenditure as the main measure of the LS. This refers both to official institutions, such as the GUS, and to individual researchers. Occasionally, an income approach has been applied, and in some cases a multidimensional analysis of poverty has been conducted.22 In this study, the flow of consumption has been considered the main measure of the household living standard. The total flow has been split into three parts and each part has been evaluated separately. 4.16 First, the consumption of non-durables and services has been calculated. It has been set as equal to the household expenditures plus the money value of the consumption in kind. The latter includes self-produced food and non-durables received free, all evaluated by surveyed households. The consumption of non-durables and services includes: food, beverages and tobacco, clothing and footwear, health and hygiene, education (including support for children living outside of the household), energy and house maintenance, transportation, communication, recreation and culture, plus miscellanea. It should be noted that this consumption aggregate includes expenditures on energy but does not include current rental charges and that it is in fact a part of the total expenditure registered by the GUS and makes up about 80 percent of the total. 19Energy bills are paid every two or three months in Poland (depending on the region/district). 20In 2001, over 30 percent of households reported zero expenditure on electricity, while all of them used electricity. 21The imputation of energy charges seems to be a proper solution in this case. However, total electricity consumption estimated from the HBS was already higher than the officially published figures by about 10 percent (in 2001). In such a case, imputation would add additional bias to the sample estimates, and thus it has been abandoned. 22See, for instance, GUS (various years), Living Conditions of the Population, GUS (2000a), T. Panek (1996) and (2001c). 63 4.17 Second, the flow of consumption from the use of a house or an apartment has been evaluated. This has been carried out through the rent imputation. Imputed rent has been estimated from the regression of the (log) current rental charge on a number of regressors, such as apartment quality measured with its equipment and access to various facilities, its space, type of building, location, etc. Location has been measured with the following variables: region (voivodship), place of residence (rural or urban, plus the size of the city), and rate of unemployment in the region (as a proxy of the general quality of the neighborhood). The regression itself has been run on selected households ­ given their house/apartment ownership and registered housing expenditures.23 The rent imputation procedure has been applied only for 2001, and it has not been repeated for 1994 ­ 2000. The main reason for this is the lack of all necessary housing information for the whole period. Thus, it would not be possible to generate the same regressions for all of the years under investigation. It should also be remembered that the regression results for 2001, although reasonable and acceptable, are not fully satisfactory (R2 less than 8 percent). More housing information would be needed to improve them. In fact, they should be treated as "experimental." 4.18 Third, the use of durables has been imputed. Monthly consumption has been evaluated for over 25 durables registered in the HBS. Given the scarcity of information in the HBS, the following imputation scheme has been adopted.24 It has been assumed that a durable life is equal to about 10 years. A few exceptions include PCs and printers (shorter life period) and sewing machines (longer life period). For the depreciation, the assumptions appear as follows: the rate is set to 15 percent for the first year and to 9 percent for all other years with exceptions as listed above). This distinction has been possible with the HBS information on current expenditures. Next, it has been assumed that the total value of a durable is equal to its current price. This total value has been estimated after the examination of current expenditures from the HBS and current market prices published by the GUS (data collected for computing CPIs.) The published price has been applied only if the HBS information has not been reliable because of an insufficient number of cases registered in HBS or the unclear classification of current expenditures. In most cases, HBS information has been used. The price has been set as equal to the overall mean expenditure on a durable, or to the overall median expenditure, or to the mean expenditure in a quartile group of a household (quartiles based on ex ante consumption, before any imputation). The last solution, which is intended to differentiate values of durables according to the household material status, has been given a priority. Mean or median has been used only if expenditures by quartile groups have appeared unreliable. The results of this imputation procedure, based on a number of quite restrictive assumptions and arbitrary decisions, should be treated cautiously. In fact, as in the case of rents, they should be treated as "experimental." Therefore, the imputation of the use of durables has been implemented only for 2001. 4.19 Regional Deflators. Regional differences concerning voivodships and rural-urban areas are quite visible in Poland. They concern various dimensions of economic and social life, as shown by a number of recent regional studies.25 Regional price differences are important for poverty assessment but they have not been given any special attention to date.26 In this research, regional price deflators have been used in order to account for differences in household standards of living across voivodships as well as rural and urban areas. These deflators have been computed on the basis of the HBS alone. Both price ratios and individual weights have been 23See the first section of the Methodological Appendix for details. 24See the second section of the Methodological Appendix for details. 25On the development of regions (general issues) see, for instance: Czyewski, Góralczyk­Modzelewska, and (2001), Gorzelak [report for the current project] and (1999), and Niemczyk (2001). 26World Bank (1995) has partially approached this question. This seems rather exceptional. 64 calculated on the basis of HBS information on household expenditures. The procedure has been repeated for the entire eight year period under investigation. 4.20 Regional deflators point to a considerable price differentiation (Table A1, Statistical Appendix). In 2001, for example, these deflators remained in the range of 0.7516 to 1.2364, reaching the highest average level in Mazowieckie (1.0433), Pomorskie (1.0268), Malopolskie (1.0241) and lskie (1.0135) voivodships and the lowest levels in Podkarpackie (0.9542), Podlaskie (0.9560), Lubelskie (0.9611) and Kujawsko-Pomorskie (0.9698). It should be noted that the first group of voivodships contains those that are usually the richest in Poland (especially Mazowieckie and lskie, with the highest per capita GDP) (see Map 4.1), while the second contains those with the lowest GDP (except for Kujawsko-Pomorskie, which is in the middle range). Price differences between rural and urban areas are even more pronounced. In 2001, deflators started at 0.9629 in the countryside and increased uniformly up to 1.0824 in the large cities. As a result, price deflators for main rural populations, farmers and mixed households are all below 1, while those for main urban residents (workers and employees and their families) are mostly above 1. 4.21 This indicates how the use of regional price deflators would affect the standard of living measures of various population groups, and in turn, how it would affect the poverty profile. With the use of regional price deflators, the standard of living of the rural population, especially of farmer families in eastern voivodships (except for Warmisko-Mazurskie), would be found to be relatively higher, and the standard of living of the urban population, mainly residents of the large cities of Mazowieckie (Warsaw) and Pomorskie (Gdask) would be found to be relatively lower. Quite possibly, the poverty profile would change accordingly: lower poverty would be estimated for the rural areas of the east than for the urban areas of the center and north. It should be borne in mind, however, that the use of regional measures of living standard is reasonable as long as it is assumes that there is no migration across regions. Allowing for a high level of mobility (such as living in one place and working elsewhere) weakens the arguments for the overall use of regional price deflators. 4.22 Household Composition. In order to account for household composition, equivalence scales have been applied. These scales appear as follows: 1.1 for a single person, 1.0 for a head of a multi-member household, 0.7 for other adults, and 0.5 for children under 15. The reason for selecting these scales was twofold. First, they are (implicitly) used in the Polish Social Assistance (SA) system and so, in a sense, they have an official status. Second, they are very close to the OECD scales, widely used in many studies.27 The only difference between the Polish SA and the OECD scales is in the coefficient for a single person. According to the Polish SA, this coefficient is equal to 1.1 (OECD sets it at 1.0). This seems quite reasonable, since an individual living alone (in order to equalize the household material status with other non-single households) needs a considerably higher level of consumption. It should be noted that the new OECD scales (1.0/05/0.3) appear quite different from the scales appropriate in the Polish case28 and therefore they have not been considered in this study. A per capita approach has been touched upon, but the discussion of the results (see the tables in the last section of the Statistical Appendix) has been passed over. 4.23 LS Measures Adopted. The use of a new methodological approach has required paying more attention to the impact of LS measures on poverty indices. Therefore, four different living 27In fact, most poverty studies in Poland, including those undertaken and published by the Central Statistical Office, use OECD 1.0/0.7/0.5 scales. Exceptions include subjective poverty analyses, or some others covering especially the issue of equivalence scales (see, for instance, Panek [2001b], and Szulc [1996, 2000 and 2001]). 28For evidence, see the references quoted in the previous footnote. 65 standards measures (consumption aggregates) have been explored for studying the 2001 poverty profile. Two of these measures follow the GUS expenditure concept, and two others reflect the World Bank (WB) flow of consumption concept. They are specified in Table 4.1. 4.24 All aggregates have been expressed in June prices. GUS consumption has been deflated with the use of a general CPI. In the case of WB consumption, the following approach has been adopted: (i) expenditures on non-durables and services have been deflated with the use of specific CPIs for sub-aggregates (all together, 13 of them); (ii) imputed rents have been deflated implicitly, through actual rental charges, with the use of a proper CPI deflator; and (iii) the imputed consumption flow of durables has been deflated with the use of a general CPI. With the exception of the GUS aggregate, all aggregates have been deflated by regional deflators. Outliers have been excluded only for "eat-out" expenditures in the WB concepts. Finally, in order to account for the household composition, Polish social assistance equivalence scales have been used. Table 4.1: Consumption Aggregates Concept/Name Description All households' expenditures (cash and in kind) on consumer goods and services GUS plus "other expenditures" which include support for other households and so called non-current taxes. GUS reg Regionally deflated GUS. All households' expenditures (cash and in kind) on selected goods and services, WB0 except: support for other households, non-current taxes, durables and rents (sewerage/water and energy expenditures were included); regionally deflated. WB2 WB0 with two imputed flows of consumption: rents and durables. Note: Changes in poverty in 1994-2001 have been studied only with the use of GUS, GUS reg and WB0 aggregates. 4.25 LS Measures: Selected Results. Various characteristics of consumption aggregates used for 2001 poverty assessment are displayed in Table 4.2 in this chapter and in Table A2 in the Statistical Appendix. Table A17 in the Statistical Appendix presents aggregates for the period 1994-2001. Table 4.2: Imputed Consumption by Quartile Groups and Place of Residence, 2001 Household residence/imputed Ex post consumption quartile* group consumption Total 1 2 3 4 WB0 aggregate as percent of GUS reg All households 83.4 81.8 80.1 76.9 79.8 Urban households 81.4 79.9 79.2 75.8 78.1 Rural households 85.6 84.9 82.1 80.4 83.1 Imputed consumption as percent of WB0 aggregate All households Imputed rents 10.8 9.1 7.9 6.0 7.9 Imputed flow of durables 6.1 6.5 7.0 7.2 6.8 WB2 aggregate 116.9 115.5 115.0 113.2 114.8 Urban households Imputed rents 11.0 9.0 7.9 5.9 7.7 Imputed flow of durables 5.7 6.2 6.8 7.1 6.7 WB2 aggregate 116.7 115.2 114.7 113.0 114.3 Rural households Imputed rents 10.5 9.1 8.0 6.4 8.4 66 Imputed flow of durables 6.6 7.0 7.5 7.4 7.1 WB2 aggregate 117.1 116.0 115.5 113.9 115.5 Note: * Quartiles are for households, set for equivalent WB2. Source: HBS 2001, own computation. See also Tables A2 and A17 in the Statistical Appendix. 4.26 First, it should be noted that on the average the WB0 aggregate was lower than the GUS aggregates by about 20 percent in 2001, but this difference was smaller in the past. In 1994 it accounted for only 15 percent, and it has increased continuously thereafter. This indicates that expenditures on durables have become more important in household budgets over the last years. It should also be noted that the difference between WB0 and GUS consumption aggregates is larger for urban than for rural households, and that it has increased with the household quartile group. This provides evidence of a lower overall differentiation in the living standard according to WB0 as compared with GUS. Also, since the LS of rural households is generally below that of urban households, it suggests a lower discrepancy between rural and urban populations if measured with WB0. 4.27 Second, on the average, imputed rents add almost 8 percent to the "basic" WB0 consumption, and the imputed consumption flow of durables adds almost 7 percent. These proportions, in view of the results for other countries,29 seem reasonable. What is most important is, however, the way in which imputed consumptions are distributed across various categories of households. On the average, rural households have "received" slightly more than urban households. Certainly, this would diminish the difference between rural and urban households measured with WB2, compared to WB0 or GUS aggregates. In relative terms, imputed rents are less equally distributed across households from various quartile groups than is the imputed consumption of durables. Moreover, while imputed rents decrease with the quartile group (from almost 11 percent of WB0 in the first group down to 6 percent in the fourth), the imputed flow of durables actually increases slightly (from 6 percent up to over 7 percent). This means that LS measured with imputed rents alone would be more equal than when measured with WB0 (i.e., without the imputation). On the other hand, the equalizing impact of the imputed consumption flow of durables would be less pronounced. 4.28 Third, Gini coefficients confirm the findings already listed, namely, that the distribution of LS measured with total expenditures (GUS aggregates) is less equal than when measured with WB aggregates, and that the application of regional deflators also has an equalizing impact. In 2001, the overall Gini reached 0.333 for GUS reg, and it was slightly lower than for GUS (0.344) (Table A17 in the Statistical Appendix). But the Ginis of both WB aggregates were considerably lower. They reached only 0.290 for WB0 (before imputation) and 0.283 for WB2 (after imputing rents and consumption of durables). This would certainly have an impact on the assessment of poverty depth. It should be added that the poverty profile would be modified as well, given the change in the position of rural versus urban households. 4.29 Finally, it should be noted that the total consumption in kind, including food, imputed rents, and consumption of durables, is much higher in poor households than in better-off households, especially in rural areas. Rural households in the first quartile group consume in kind as much as 27 percent of the total WB2, while urban households in the fourth quartile consume only 12.4 percent (Table A2 in the Statistical Appendix). This means that, in terms of cash liquidity, poorer rural households are certainly deprived. However, this aspect of the living standard has not been handled. 29Kosovo or Croatia, for example. 67 Measures of Poverty 4.30 Poverty Lines. Two main poverty lines have been used for each consumption aggregate. The intention was to depict and investigate both "medium" and "hard" poverty. Both lines stem from Polish standards; they are often found in various country reports, so they can be easily interpreted by Polish analysts. Both were chosen for 2001, and were kept constant in real terms (with the use of annual CPIs) for the analysis of 1994-2001.30 4.31 The first line, depicting medium poverty, reflects the rules of the Polish social assistance system. It has been set equal to the annual average of the SA threshold for the household head (PLN 389, or constant 1998 PPP$ 190). In practice, this threshold applies to the household's equivalent income, or ­ implicitly - to its total equivalent expenditure. Therefore, this line has been used with GUS aggregates, since they are actually equal to the total expenditure. This line makes up almost 60 percent of the median or 50 percent of the mean of GUS aggregates in 2001, and so it reaches the level quite often used in poverty studies (when relative poverty is examined). Poverty lines for WB aggregates have been adjusted using the ratio of the first quartiles of the appropriate consumption aggregates. It has made it possible to keep this line approximately "constant" at the lower ends of all consumption aggregates. 4.32 The second line, depicting hard poverty, reflects the subsistence minimum basket. This is a very narrow basket, and it includes only basic goods and services. Evaluated in money terms, it gives an amount of a minimum which barely allows covering the essential needs of everyday life. Therefore, it seems to be a proper threshold for severe poverty. The subsistence minimum is evaluated regularly by the Institute of Labor and Social Affairs (Instytut Pracy i Spraw Socjalnych, IPiSS) for several types of worker and retiree households. In this research, the 2001 annual average of the subsistence minimum for a single worker has been used as a hard poverty line for GUS aggregates. It is equal to PLN 336 per month (constant 1998 $PPP 165), or slightly over 50 percent of the GUS median. Hard poverty lines for WB aggregates have been adjusted using the ratio of the third vintiles of the appropriate consumption aggregates. As in the case of medium poverty, the intention was keeping the chosen line approximately "constant" at the lowest ends of consumption aggregates. 4.33 For 2001, an additional poverty line has been added, namely, $PPP 4.30 a day (in constant 1998 US$). This line is often used in international studies of poverty. It corresponds to PLN 263 per month and is lower by over 20 percent than a subsistence minimum. This amount seems to be too low for investigating poverty in Poland. Therefore, it has only been used to indicate overall "extreme poverty." The line 263 PLN has been applied for GUS aggregates; for WBs it has been adjusted with the ratios of the first deciles of the appropriate consumption aggregates. 4.34 Poverty Indices. The headcount for persons is the main poverty index used and explored in this research. However, the poverty gap has also been calculated, as well as the poverty depth index (or FGT1) and the poverty deficit index ­ all of them at the country level exclusively. The latter allows for the proper evaluation of the minimum amount necessary to close the poverty gap in case the gap is calculated with the use of equivalent units. Formulas of all indices are displayed in Table 4.3. 30All poverty lines discussed below are displayed in most tables in the Statistical Appendix (see, for example, Statistical Appendix Tables A3 and A16). 68 B. POVERTY IN 2001 Overall Picture of Poverty 4.35 In 2001, over 14 percent of the total population ­ or more than 10 percent of households lived in "medium" poverty in Poland.31 This means that poverty affected over 5.5 million people (see Table 4.4 and, for more detail, Table A3 in the Statistical Appendix). 4.36 The poverty estimate depends on the way the living standard has been measured. The total expenditure or GUS concept gives a slightly higher poverty rate than the WB flow of consumption concept (15.2 percent, or almost 5.9 million poor). It may also be noted that the use of regional deflators alone decreases the poverty headcount. But, in general, estimates of poverty are fairly similar. This might have been expected, given the way of setting poverty lines. Table 4.3: Poverty Measures (indices) Headcount Q Q - number of poor individuals H = N N ­ total population z - poverty line Poverty depth 1 Q P1= ( z - yi) yi - equivalent consumption of the poor individual "i" N z i=1 1 Q Poverty gap PG = ( z - yi) z - poverty line Q z i=1 yi - equivalent consumption of the poor individual "i" wi - the sum of equivalent units in a household of Poverty deficit* 1 Q i -yic PD = individual "i" divided by the size of this household N wzz yci - per capita consumption of the poor individual "i" i=1 The PD index may be interpreted as a minimum amount (expressed per person and relative to the poverty line) that one should give to the poor in order to take all of them out of poverty. PD is used only with the equivalent unit approach, with per capita PD = P1, and the above stated interpretation holds for P1. It should be noted also that, unless stated otherwise, all indices are evaluated for persons (household members). Source: HBS 2001, own computation. 4.37 Estimates of poverty depth differ depending on the LS concepts. The poverty gap index for medium poverty is in the range of 19 to 20 percent if one relies on the flow of consumption approach, but it is higher by 2 to 3 percentage points if one refers to the total expenditure approach. This is not only the well known effect (resulting from the formulas) that higher headcounts are often accompanied by higher gaps. In this case, the expenditure approach (GUS) indicates a higher poverty depth than the flow of consumption approach (WB), even for equal headcounts. This may be seen when comparing GUS reg and WB2. For both LS concepts, the headcounts are identical (14.2 percent), but the GUS reg gap is considerably higher than the WB2 gap (21.7 percent compared to 19.1 percent). This is the result of the imputation procedures which have tended to equalize the distribution of the standard of living. 31This estimate is similar to the one published by GUS for the comparable concept (see Statistical Appendix Table A16, upper panel). 69 Table 4.4: Poverty According to Various LS* Concepts: Selected Indices Poverty index [%] LS* concept GUS GUS reg WB0 WB2 Medium Poverty Headcount 15.2 14.2 14.8 14.2 Poverty gap 22.1 21.7 20.4 19.1 Hard Poverty Headcount 9.6 9.0 8.9 8.6 Poverty gap 20.2 19.6 18.7 17.9 Extreme Poverty Headcount 3.8 3.4 3.6 3.2 Poverty gap 17.6 17.3 15.6 14.9 Note: * LS = Living Standard. For methodology - see Section 1 and Methodological Annex. Source: HBS 2001, own computation. See also Table A3 in the Statistical Appendix. 4.38 Poverty depth may be interpreted in terms of the total amount necessary to close the poverty gap. This amount (evaluated for medium poverty, and adjusted for the equivalence scales) is fairly low in Poland. In 2001 it accounted for less than 0.5 percent of GDP (WB aggregates ­ Table A6, Statistical Appendix), but certainly no more than 0.6 percent of GDP (GUS aggregates). 4.39 The extent of hard poverty is quite noticeable. Headcounts of 2001 vary from 8.6 percent (WB aggregates) to 9.6 percent (GUS aggregate), and this means that about 3.5 million people live below the subsistence minimum line (or the equivalent). In this case, the total gap accounts for 0.24 to 0.29 percent of GDP, meaning that approximately two time less would be needed to pull all the poor out of hard poverty (Table 4.4 and Table A3, Statistical Appendix). As expected, poverty depth measured with the GUS consumption aggregate appears larger than that measured with WB aggregates, no matter what index is used, P1, PG or PD. 4.40 Finally, one may examine the extent and depth of extreme poverty delimited with an equivalent of $PPP 4.30 per day. Although the line is really very low, over 1.2 million people are still found below the line (or even more, close to 1.5 million, if the LS concept of GUS is used). Certainly, the total gap cannot be large in this case, and in fact it is far below 0.1 percent of GDP. Unfortunately, the extreme poverty headcount of 3.2 to 3.8 percent does not allow for investigating the poverty profile of the extreme poor. Therefore, further analysis would be restricted to hard and medium poverty exclusively. Table 4.5: Voivodship Ranking by Poverty Rate Medium Poverty Voivodship GUS GUS reg WB0 WB2 Warmisko-Mazurskie 2 1 1 1 Pomorskie 5 2 2 2 Zachodniopomorskie 6 6 3 3 Dolnolskie 7 5 4 4 Lubelskie 1 4 5 5 witokrzyskie 4 3 6 6 Kujawsko-Pomorskie 8 8 8 7 Podkarpackie 3 7 7 8 Lubuskie 11 10 9 9 Malopolskie 12 9 12 10 Wielkopolskie 10 11 10 11 70 Podlaskie 9 12 11 12 Lódzkie 13 14 16 13 lskie 16 16 13 14 Opolskie 14 15 14 15 Mazowieckie 15 13 15 16 Hard Poverty Voivodship GUS GUS reg WB0 WB2 Pomorskie 4 2 1 1 Warmisko-Mazurskie 3 1 2 2 Dolnolskie 6 4 3 3 Zachodniopomorskie 7 6 4 4 Lubelskie 2 5 5 5 witokrzyskie 1 3 6 6 Kujawsko-Pomorskie 8 9 7 7 Podkarpackie 5 7 8 8 Wielkopolskie 10 8 9 9 Lubuskie 12 10 10 10 Podlaskie 9 13 12 11 Malopolskie 13 11 14 12 Lódzkie 14 16 15 13 lskie 15 14 11 14 Opolskie 11 12 13 15 Mazowieckie 16 15 16 16 Note: Voivodships are ranked according to WB2 H. The lowest rank = the highest poverty rate. Source: HBS 2001, own computation. See also Table A4 in the Statistical Appendix. Where Do the Poor Live? 4.41 Voivodships. As a beginning, let us examine poverty at the level of voivodships. The figures in this chapter in Table 4.5, as well as Maps 4.1 ­ 4.4, tell the main story. (For this purpose, see also Statistical Appendix Tables A3 ­ A5, and A15. Appendix maps A1 and A2 may also be useful.) 4.42 First, it should be noted that poverty rates differ significantly across the voivodships (see maps and Statistical Appendix Table A4). In 2001, medium poverty headcounts varied between 10.6 percent and 19-21 percent (depending on the LS concept), or, in other words, they varied between 0.7 and 1.4 of the country average. Hard poverty rates were even more diverse. They remained in the range of 6 to 14 percent, but this means that they made up 0.7 to 1.6 of the country average. In the case of hard poverty, quite surprisingly, the differentiation measured with WB aggregates was higher than that measured with GUS aggregates. 71 Map 4.1: Medium Poverty 2001 (CSO) POMORSKIE WARMINSKO-MAZURSKIE ZACHODNIOPOMORSKIE PODLASKIE KUJAWSKO-POMORSKIE MAZOWIECKIE WIELKOPOLSKIE LUBUSKIE LODZKIE LUBELSKIE DOLNOSLASKIE SWIETOKRZYSKIE OPOLSKIE SLASKIE Poverty rate [%] PODKARPACKIE CSO MALOPOLSKIE 20.2 - 21.1 (3) 17.6 - 20.2 (3) 14.7 - 17.6 (2) 13.4 - 14.7 (3) 10.5 - 13.4 (5) Source: HBS 2001, own computation. 4.43 The most interesting finding, however, is the ranking of voivodships with respect to poverty extent. This ranking is virtually the same for GUS reg, and all WB aggregates. But if one uses the GUS total expenditure approach without regional deflators, the ranking is different. Nevertheless, there are voivodships which are always very close to the top of the list, revealing the highest headcounts, and there are those that are always close to the bottom ­ no matter how poverty is measured (see Table 4.5). 72 Map 4.2: Medium Poverty 2001 (CSO reg) POMORSKIE WARMINSKO-MAZURSKIE ZACHODNIOPOMORSKIE PODLASKIE KUJAWSKO-POMORSKIE MAZOWIECKIE WIELKOPOLSKIE LUBUSKIE LODZKIE LUBELSKIE DOLNOSLASKIE SWIETOKRZYSKIE OPOLSKIE SLASKIE PODKARPACKIE MALOPOLSKIE Poverty rate [%] CSO reg 18.0 - 18.94 (3) 17.29 - 18.0 (2) 13.95 - 17.29 (3) 11.93 - 13.95 (3) 10.62 - 11.93 (5) Source: HBS 2001, own computation. 73 Map 4.3: Medium Poverty 2001 (WB0) POMORSKIE WARMINSKO-MAZURSKIE ZACHODNIOPOMORSKIE PODLASKIE KUJAWSKO-POMORSKIE MAZOWIECKIE WIELKOPOLSKIE LUBUSKIE LODZKIE LUBELSKIE DOLNOSLASKIE SWIETOKRZYSKIE OPOLSKIE SLASKIE PODKARPACKIE MALOPOLSKIE Poverty rate [%] WB0 18.98 - 20.28 (2) 16.65 - 18.98 (3) 15.23 - 16.65 (3) 13.83 - 15.23 (3) 10.66 - 13.83 (5) Source: HBS 2001, own computation. 74 Map 4.4: Medium Poverty 2001 (WB2) POMORSKIE WARMINSKO-MAZURSKIE ZACHODNIOPOMORSKIE PODLASKIE KUJAWSKO-POMORSKIE MAZOWIECKIE WIELKOPOLSKIE LUBUSKIE LODZKIE LUBELSKIE DOLNOSLASKIE SWIETOKRZYSKIE OPOLSKIE SLASKIE PODKARPACKIE MALOPOLSKIE Poverty rate [%] WB2 18.06 - 19.42 (3) 16.74 - 18.06 (2) 15.42 - 16.74 (3) 12.21 - 15.42 (3) 10.55 - 12.21 (5) Source: HBS 2001, own computation. 4.44 The list of voivodships with the highest poverty rates appears to be as follows. Warmisko-Mazurskie and Pomorskie have the highest medium and hard poverty headcounts for GUS reg and both WB aggregates, and they are at the top of the list for the GUS aggregate as well. Among the poorest, one can find: Zachodniopomorskie, Lubelskie (the highest GUS medium poverty), Dolnolskie and witokrzyskie (the highest GUS hard poverty). Kujawsko- Pomorskie and Podkarpackie are also rather poor, although certainly not extremely poor. 4.45 At the other end of the ranking are voivodships with the lowest poverty headcounts. This group includes: Mazowieckie, lskie and Lódzkie. Opolskie and Malopolskie may also be added to this group. 4.46 The maps shown here give an additional insight into the regional aspects of poverty and make the investigation based on voivodships indices easier. They show that various LS concepts allocate the poor in a different way, but differences are not very pronounced. For example, GUS aggregates indicate the southeast as the poorest region, while according to WB aggregates this region is the second poorest (WBs allocate the highest poverty to the north). 4.47 Therefore, a common finding for all aggregates may be formulated. Voivodships with relatively high poverty rates are located in the north of Poland (Warmisko-Mazurskie, Pomorskie, Zachodnio-Pomorskie). Next, poverty spreads into the southwest, and forms an isolated island in the southeast. Voivodships in the belt that begins in the northeast, crosses the center, and ends in the south are much less affected by poverty. 75 4.48 If we look for correlates of regional poverty, we may start with economic indicators such as GDP or unemployment (Maps A1 and A2 in the Statistical Appendix). It would seem that the voivodship poverty rates reflect their GDP levels rather vaguely. It is only for the GUS concept that both rankings appear similar. For example, they indicate very low GDPs and at the same time very high poverty headcounts for three "Eastern Wall" voivodships,32 namely, Warmisko- Mazurskie, Lubelskie, and Podkarpackie. The same pattern holds for witokrzyskie. At the other end, lskie and Mazowieckie are among the richest, according to both poverty rates and GDP. 4.49 For the GUS reg and WB concepts, poverty and unemployment rankings seem closer. This may be seen for Mazowieckie, Podlaskie, and Wielkopolskie (low poverty and low unemployment), and ­ on the other hand ­ for Warmisko-Mazurskie, Pomorskie, and Zachodniopomorkie (high unemployment and high poverty). Nevertheless, there are voivodships such as Lubuskie and Opolskie which do not reveal any clear poverty-unemployment links. However, some other correlates of voivodship poverty may be found. 4.50 Former State Farms. It is not difficult to see that most of the voivodships with the highest poverty rates belong to the so-called "Megaregion 2." This is one of the three megaregions which group voivodships with a similar type of agriculture. In the past, the agriculture of six voivodships constituting "Megaregion 2" (Warmisko-Mazurskie, Pomorskie, Zachodniopomorskie, Lubuskie, Dolnolskie and Opolskie) was dominated by large state farms (Pastwowe Gospodarstwa Rolne, or PGRs). PGRs were restructured at the beginning of the 1990s, and most PGR laborers ­ usually uneducated and helpless ­ lost their jobs. At present, many of them are unemployed, and it is only occasionally that they have the status of tenants or part-time workers. This social group is especially vulnerable. While their standard of living has been studied by many researchers,33 the poverty rates have not been estimated to date. 4.51 Such an estimate may be made on the following basis. It may reasonably be assumed that former PGR laborers live in rural areas, in the voivodships of Megaregion 2. Moreover, they live in apartment projects or in row houses rather than in family houses. Table 4.6 displays the appropriate figures. Table 4.6: Poverty of Post-PGR Population WB2 headcounts [%] for rural population Medium poverty Hard poverty Percent of Place of residence/ population in Type of building Type of building voivodship apartment/ raw Apartment / Family house Apartment / raw Family house raw house house house (PGR) (Other) (PGR) (Other) Ddolnolskie 44.5 32.3 19.7 23.4 13.4 Lubuskie 35.2 28.7 18.7 15.8 12.5 Pomorskie 47.9 34.8 23.0 29.5 16.5 Warmisko-Mazurskie 53.1 30.3 17.8 21.9 11.8 Zachodniopomorskie 65.2 27.3 26.1 19.0 17.8 TOTAL (all voivodships) 20.6 26.9 16.7 18.7 9.9 Note: Opolskie in Megaregion 2 is skipped (the share of the rural population living in apartments is too small). Source: HBS 2001, own computation. 32"Eastern Wall" is the name given to four poor (in terms of GDP) and predominantly rural voivodships of the East of Poland, namely Warmisko-Mazurskie, Podlaskie, Lubelskie and Podkarpackie. 33See, for example: Kawczyska-Butrym (2001a) and (2001b), Tarkowska (ed.) (2000), UNDP (2000), and Zablocki et al. (1999). 76 4.52 The rural population of the voivodships in Megaregion 2 is often housed in apartment projects (the exception is Opolskie) -- from one-third in Lubuskie to two-thirds in Zachodniopomorskie. According to previous arguments, this population may be identified as post-PGRs. The Poverty rate for the post-PGR population is very high. In Pomorskie, for example, the medium poverty headcount reaches almost 35 percent, in Dolnolskie it is 32 percent, and in other voivodships it is close to 30 percent, while the country average is only 14.2 percent (WB2). Moreover, rates for the post-PGR population are 30 to 80 percent higher than for other rural residents. This refers to both medium and hard poverty. There are, however, exceptions. In Zachodniopomorskie, for example, poverty rates are quite evenly distributed among the whole rural population. At this point, however, it is difficult to answer why this is the case. 4.53 Rural versus Urban Areas. The regional analysis of poverty already presented draws attention to rural areas. In fact, poverty in rural areas is usually much greater than in the towns and cities. Table 4.7 in this chapter and Tables A5 and A15 in the Statistical Appendix display main indices comparing rural and urban poverty for various LS concepts. 4.54 In 2001, rural poverty was about 2.5 times higher than urban poverty, according to the total expenditure LS concept. GUS headcounts indicated that, on the average, 10 percent of the urban population and 23 percent of the rural population lived in medium poverty. At the same time, GUS hard poverty headcounts reached 6 percent and 15 percent, respectively. According to these figures, 61 percent of the medium poor and 63 percent of the hard poor lived in the countryside, while the proportion of rural residents in the total was lower than 40 percent. Table 4.7: Poverty in Rural and Urban Areas Headcounts [%] Population in poverty [%] Place of residence GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Medium Poverty Urban 9.9 10.0 11.6 11.2 39.1 42.3 47.1 47.5 Rural 23.2 20.7 19.8 18.8 60.9 57.7 52.9 52.5 rural/urban 2.4 2.1 1.7 1.7 Hard Poverty Urban 5.9 6.1 6.7 6.5 37.3 40.6 45.6 45.7 Rural 15.2 13.5 12.1 11.8 62.7 59.4 54.4 54.3 rural/urban 2.6 2.2 1.8 1.8 Note: For methodology, see Methodological Annex. Source: HBS 2001, own computation. See also Statistical Appendix Tables A5 and A15. 4.55 The picture differs slightly if regional price deflators are used for adjusting living standard measures. Since the average price level is usually lower in rural than in urban areas, the difference between rural and urban headcounts diminishes. The GUS reg headcount shows relatively lower poverty in rural areas than the GUS headcount. But it still shows that the rural population is more than two times poorer than the urban population. 4.56 The use of WB concepts additionally reduces the difference between rural and urban headcounts, because the imputed consumption of rural households is relatively higher. As a result, a less sharp poverty picture is received. In 2001, according to the flow of consumption LS concept, medium poverty reached 11 percent of those in rural areas, and 19 percent of those in urban areas, and hard poverty reached 6.5 percent and 12 percent, respectively. This gives 53 to 77 54 percent of the poor living in rural areas considerably less than was estimated with the use of GUS concepts. 4.57 It may also be noted that the urban-rural pattern of poverty holds in all voivodships (Statistical Appendix Table A6). The largest rural-urban ratio of headcounts is found in Mazowieckie and in Pomorskie. In the Mazowieckie case, it is due to the very low urban headcount (only 7 percent medium poverty measured with WB2) which possibly reflects the Warsaw position. In the case of Pomorskie, it is due to the extremely high rural poverty (WB0 medium poverty headcount equal to 31 percent, WB2 to 29 percent), most likely affecting the post-PGR population. 4.58 On the other hand, the smallest rural-urban differences are in lskie (large cities, small rural sector) and witokrzyskie (poor in general, in urban and in rural areas). In lskie, rural and urban WB2 headcounts are almost equal, with both among the lowest in the country. 4.59 Small Towns versus Large Cities. Poverty in the cities clearly shows that the larger the city, is the smaller the poverty count is. This may be seen in Figures 4.1 and 4.2 and in Statistical Appendix Table A5. 4.60 In the largest cities, medium poverty accounts for 4 to 6 percent (depending on the LS concept used), while at the same time it reaches about 16 percent in the smallest towns. For hard poverty, the respective figures are: 2 to 3 percent for large cities, and almost 10 percent for small towns. Medium cities are in between. One may notice, however, that the scope of poverty is larger in cities of 100,000 ­ 200,000 inhabitants than in towns of 20,000 ­ 100,000, but this difference appears negligible. One may even conclude that poverty in medium cities of 20,000 ­ 100,000 inhabitants is quite similar. 4.61 As in the case of rural-urban poverty, the use of regional price deflators and the move from total expenditure toward the flow of consumption concept renders poverty differentiations less sharp. This may be explained on similar grounds as previously. Living in small towns is cheaper than living in large cities, and imputation procedures add more in small towns. Therefore, if deflators and imputations are applied, poverty differences between cities will shrink. 78 Figure 4.1: Medium Poverty by Place of Residence, Poland 2001 25 20 [%] 15 10 Headcount 5 0 Rural 20 000 - 20 - 100 000 100 - 200 000 200 - 500 000 500 000+ Place of residence CSO CSO reg WB2 Source: HBS 2001, own computation. Figure 4.2: Hard Poverty by Place of Residence, Poland 2001 16 14 12 [%] 10 8 6 Headcount 4 2 0 Rural 20 000 - 20 - 100 000 100 - 200 000 200 - 500 000 500 000+ Place of residence CSO CSO reg WB2 Source: HBS 2001, own computation. What Drives People into Poverty? 4.62 Three groups of poverty determinants -- called here economic, demographic, and social -- are investigated in this section. The first group includes work related factors, such as employment status and source of income. The second group concentrates on individual age and family status, and the third group covers education and gender. At the beginning, they are 79 discussed separately. Their common impact is described with a logit model of poverty which is presented at the end of this section. It should also be pointed out that the discussion focuses on poverty measured with the WB2 aggregate. All other poverty rates are displayed in the Statistical Appendix, and they are only mentioned occasionally. Economic Factors: Work Status and Main Source of Income 4.63 Poverty is certainly directly related to the work status of household members. Headcounts in Table 4.8 and in Tables A7 ­ A11, and A15 in the Statistical Appendix clearly show that it is in fact one of the major determinants of poverty. 4.64 Main Source of Income. Let us compare first households which have been put into seven so-called socioeconomic categories, according to main source of income. This classification allows us to distinguish those who live primarily on their work from those who rely mostly on social transfers. 4.65 In general, people living on social transfers are in a worse position than the other groups (with the exception of families of retirees). Poverty headcounts for those living on social incomes (i.e., on social assistance, unemployment benefits, alimony, etc.) are far above all others. In 2001, the medium poverty headcount for recipients of social incomes reached over 40 percent and it was 2.9 times higher than the average. This group accounted for 14 percent of the (medium) poor, although its share was only 5 percent of the total population (Table 4.8 and Statistical Appendix Table A15). At the same time, it accounted for 17 percent of the hard poor, and the hard poverty rate for this group was equal to 29 percent, or as much as 3.4 times the average. Quite possibly this reflects the impact of various adverse conditions, such as many dependents, sickness, and low education, which are concentrated in this group. 4.66 People living on disability or survivor pensions constitute the second group that is most affected by poverty. But their 2001 poverty rates were considerably lower than the rates of other social beneficiaries, reaching 20 percent for medium poverty and 12 percent for hard poverty (in both cases, 1.4 times the country average). 4.67 The situation of retirees and their dependents is quite different. This is the only group living on social transfers which does not suffer considerable poverty. On the contrary, the poverty extent of retirees remains among the lowest: 8.3 percent for medium poverty, and only 4.8 percent for hard poverty in 2001. Each time, it was much less than the overall country headcount (by about. 40 percent). This is the result of both household characteristics (very few dependents) and the features of pension schemes (the relatively high replacement rate of the retirement pensions). Table 4.8: Work Status and Poverty WB2 headcounts [%] Composition of the Poor [%] Population category Hard Medium Hard Medium poverty poverty poverty poverty Socioeconomic category* of a hhld 100 100 Worker/Employee 7.5 12.5 38.2 38.5 Worker-Farmer 9.1 15.9 12.1 12.9 Farmer 7.6 14.7 4.9 5.8 Self-employed 4.3 8.0 3.8 4.3 Retiree 4.8 8.3 8.9 9.3 80 Disabled/Survivor 12.2 20.0 15.0 14.9 Living on social income 29.0 40.7 17.0 14.4 Employment status of a hhld head 100 100 Unemployed 36.6 48.3 9.8 7.8 Employed** 6.7 11.8 51.7 55.4 Other 10.7 16.9 38.5 36.7 Unemployment of hhld members 100 100 0 unemployed 5.7 10.3 53.3 58.4 1 or more unemployed, of which: 20.8 30.4 47.1 41.6 1 unemployed 16.5 25.1 29.8 27.4 2 or more unemployed 38.0 51.2 17.4 14.2 Employment of hhld members 100 100 0 employed** 12.5 18.7 35.8 32.4 1 or more employed**, of which: 8.4 12.8 73.7 67.6 1 employed 9.5 15.9 36.2 36.7 2 or more employed 5.7 10.4 28.0 30.9 Note: *Socioeconomic category reflects the main source of income. **Employed ­ having a permanent job (including farmers, self-employed and professionals). Source: HBS 2001, own computation. See also Statistical Appendix Tables A7 ­ A11, and A15 in the Statistical Appendix. 4.68 The poverty rates of retirees' households are only slightly higher than the headcounts for the self-employed, by 0.3 to 0.5 percentage points. Moreover, some LS measures indicate a slightly lower poverty rate for retirees than for the self-employed (GUS reg for medium poverty and WB0 for hard poverty) (Statistical Appendix Table A7). This result is rather unexpected and it appears too optimistic from the point of view of retirees. This requires a comment. It would seem that the poverty of the self-employed might be overestimated because of the distribution of the refusal rate in the HBS (higher refusal among the wealthy). On the other hand, the poverty of retirees might be underestimated because of the LS concepts adopted. These LS concepts, for example, do not account for health conditions, which are often considered an important factor in the standard of living34 (possibly they are worse among retirees than among the self-employed). 4.69 Farmer versus Worker Households. The comparison of farmer and worker populations needs special attention, for both economic and political reasons. As has been noted, poverty in rural areas is more pronounced, and rural areas are generally perceived as poor. This does not necessarily mean that farmers and their families, who make up less than 50 percent of the rural population,35 are among the poorest. 4.70 In 2001, the medium poverty rate was equal to 12.5 percent for members of worker households and it was somewhat higher for members of farmer households (almost 15 percent). Hard poverty rates were 7.5 percent for workers and 7.6 percent for farmers. Thus, surprisingly, farmer households were found to be only slightly poorer than worker households. At the same time, members of mixed (i.e., farmer-worker) households suffered higher poverty than farmer households (medium poverty, 16 percent, and hard poverty, 9 percent). This finding seems 34Certainly, health is not the only factor neglected in the LS concepts adopted (mainly for technical reason of data scarcity). For the discussion of the impact of health conditions on LS and poverty, see, for example, Sen (1992). 35According to Census 2002, only 16.3 percent of rural households declared agriculture as their main source of income (www.stat.gov.pl, July 5 2003, NSP 2002 table 29). Given the average household size, the share of farmer population is rather higher. HBS 2001 shows 22.2 percent of the farmer population in rural areas (see Table A15 in the Statistical Appendix). 81 reasonable. It may be that rural poverty is "located" not necessarily in farms but in households which do not live only by farming. The high poverty level of the former PGR population, discussed in the previous sub-section, confirms this hypothesis. 4.71 Nevertheless, it should be pointed out that the poverty rates just quoted are based on LS concept WB2, which includes imputed rents and consumption of durables. Headcounts evaluated for GUS aggregates show much larger differences between farmer and worker poverty headcounts: 19 percent and 11.4 percent, respectively, for GUS reg medium poverty, and 23.5 percent (farmer) and 11.8 percent (worker) for GUS medium poverty (see Statistical Appendix Table A7). 4.72 Farm Size. If we look more closely at the farmer situation, we may investigate poverty according to farm size, measured with land area (Table A8 and Figures A1 and A2 in the Statistical Appendix). As might have been expected, the smaller the farm is, the higher is the poverty rate. Exact figures, however, are interesting. Headcounts for people living in farms up to 5 hectares are really high: 20 percent for medium poverty, and 11.0-11.5 percent for hard poverty (WB2). These rates are over one-third higher than the country average. On the other hand, poverty headcounts for people in farms larger than 10 hectares may be called moderate, and for those living in farms of over 15 hectares, poverty headcounts shrink to the "acceptable" level (below the rural average). One may conclude that, on the average, people living in farms larger than 15 hectares are not especially poor. In fact, poverty affects mostly small farmers. But it should be noted that the share of people living on large farms is modest: 13 percent of all farming on land areas over 1 hectare (see Statistical Appendix Table A15). And, once again, one should remember that poverty figures based on total expenditure LS concepts are higher for all farmers. 4.73 Unemployment. While farming drives some people into poverty, it is certainly not the major poverty factor. The impact of unemployment is much more noticeable, even in the countryside ­ as has been shown in the discussion on post-PGRs. And indirect evidence of the role of unemployment has been given in the form of headcounts for people living on social incomes. In many cases, these headcounts refer to the family members of the unemployed. Direct evidence of the predominant role of unemployment derives from the comparison of two groups of households -- with and without unemployed -- or from the investigation of households headed by the unemployed. 4.74 Figures in the middle panels of Table 4.8 (see also Statistical Appendix Table A4) show that the medium poverty rates in the households headed by the unemployed reaches 48 percent, and so it is over four times higher that in the households headed by an employed person. For hard poverty, this relation is even higher. As a result, this group of the population, which accounts for only 2.3 percent of the total (Statistical Appendix Table A15), makes up almost 8 percent of the medium poor and 10 percent of the hard poor. 4.75 The position of households with unemployed persons is important because this is quite a large group, amounting to almost 20 percent of the total population. Headcounts for this group are very high as well: 30 percent for medium poverty, as compared to 10 percent for households without unemployed persons. The difference between hard poverty rates is even more pronounced. If there are no unemployed in the household, the headcount is less than 6 percent; if there is an unemployed person, it is almost 21 percent. It may also be noted that if more than one household member is unemployed, poverty rates soar up to 51 percent (medium) and 38 percent (hard). Fortunately, this population is rather small (Statistical Appendix Table A15). It should be noted that these results are almost identical for all LS measures. 82 4.76 Poverty is also related to unemployment duration, as is seen in Table 4.9. Table 4.9: Poverty by Unemployment Duration WB2 Headcounts [%] Unemployment duration (weeks) Poverty Hard Medium 0 weeks 6.2 10.8 4 weeks or less 15.6 21.9 5 - 13 16.4 25.9 14 - 26 18.0 26.7 27 - 52 17.5 26.3 53 - 72 19.6 30.9 73 ­ 144 20.7 29.7 145 weeks or more 25.9 37.3 Total population 15+ 7.5 12.5 Source: HBS 2001, own computation. 4.77 The figures in the table refer to the population aged 15 and over. They show that the first month of unemployment pushes up the poverty rate significantly: from 11 percent up to 22 percent (medium), and from 6 percent up to 16 percent (hard). These rates increase slightly in the second month and then remain more or less stable until the end of the first year of unemployment. But the most significant increase in poverty occurs in the long-term unemployed. In the third year of unemployment, the above medium poverty headcount reaches 37 percent, and the hard poverty headcount reaches as much as one-fourth. Of course, these rates are far above the average. In view of the poverty results just discussed, one may ask how rewarding is it to have work. The answer is not that clear, for there are usually several interrelated factors which determine the ultimate poverty status. Figures in the lower panel of Table 4.8 and in Statistical Appendix Tables A9 and A11 permit us to investigate some aspect of this issue. The poverty rates displayed there relate to two groups of households. The first group contains households with zero working members, and so they are living mainly on social transfers or on part-time jobs, while the second group contains those with at least one person working on a permanent basis (including farmers and the self-employed). Both may include unemployed members. It is evident that a permanent job alone rewards a household considerably, diminishing poverty rates by about one-third. But it is also evident that one working person cannot reduce the poverty rate below the average. Moreover, comparing the headcounts in the respective panels of Table 4.8 allows us to conclude that unemployment has a stronger impact on the poverty rate than permanent work. Demographic Factors: Health, Family Status and Age 4.78 The profile of poverty with respect to selected demographic factors can be seen in Table 4.10 and Figure 4.3 in this chapter, together with Statistical Appendix Tables A12 ­ A13 and Figures A3 ­A4. 83 Table 4.10: Health, Age, Family Status, and Poverty WB2 headcounts [%] Composition of the Poor [%] Population category Hard Medium Hard Medium poverty poverty poverty poverty Disability of hhld members 100 100 0 disabled 8.2 13.2 65.4 63.9 1 or more disabled, of which: 9.6 16.5 34.6 36.1 1 disabled 9.6 16.4 26.1 27.0 2 or more disabled 9.4 16.7 8.5 9.1 Individual age 100 100 0 ­ 14 13.0 21.0 30.9 30.1 15 ­ 24 11.1 17.7 22.4 21.6 25 ­ 44 8.6 14.3 26.3 26.4 45 ­ 64 5.4 9.4 15.7 16.5 65 + 3.7 7.1 4.7 5.5 Family type (selected) 100 100 Single person 2.3 4.5 1.4 1.7 Couple + 0 children 1.4 2.8 1.8 2.2 Couple + 1 child 3.7 6.2 5.0 5.0 Couple + 2 children 5.6 11.0 11.8 13.9 Couple + 3 children 11.7 19.6 10.9 11.1 Couple + 4+ children 32.7 44.9 17.8 14.8 Single mother with children 13.8 20.7 4.4 4.0 Source: HBS 2001, own computation. See also Statistical Appendix Tables A12-A13, and A15. 4.79 Disability. Although health conditions may heavily influence poverty, they cannot be studied at length given the data available. The only proxy with regard to health which may be found in the Polish HBS is the official disability status of individuals, followed by information about disability benefits. 4.80 As has been mentioned, poverty is extensive among people living on disability/survivor pensions (in terms of the share of people affected). However, the very presence of a disabled person in a household is not that harmful. The headcount ratio for the population which lives with disabled family members is higher than the headcount for the others by 3 percentage points in the case of medium poverty, and by 1.4 points in the case of hard poverty. This is not a tremendous impact, especially if compared with the influence of the unemployment or work status on poverty (Table 4.8). This is an effect of the social protection system. If a person holds an official disability status, quite possibly she/he receives benefits (pensions, nursing allowance, and the like). This prevents the poverty rate from making a sharp increase. On the other hand, nothing can be said at this moment about the poverty of those who suffer medical problems but are not legally disabled. 84 Figure 4.3: Medium and Hard Poverty (WB2) by Age, Poland 2001 24.0 22.0 20.0 18.0 [%] 16.0 14.0 12.0 10.0 8.0 Headcounts 6.0 4.0 2.0 - 0 - 4 5 - 9 10 -14 15 - 19 20 - 24 25 - 29 30 - 34 35 - 39 40 - 44 45 - 49 50 - 54 55 - 59 60 - 65 65 - 69 70 - 74 75 + Age brackets Medium poverty Hard poverty Source: HBS 2001, own computation. 4.81 Dependent Children. Regarding family type, it is easy to see that couples with no children have the lowest poverty rate. In 2001, this rate was below 3 percent for medium poverty, and only 1.4 percent for hard poverty. Headcounts for people living alone were higher; however, they were still far below the country average. Having children increases the poverty rate, but this increase is not uniform: it is rather moderate up to two children, and becomes intense starting with the fourth child. Headcounts for families with one or two children are below the average; they are above the average for couples with three children, reaching almost 20 percent (medium poverty) and 12 percent (hard poverty). It may be noted that headcount ratios for single mothers are quite similar. 4.82 The scope of poverty is alarming for families with four or more children. In 2001, as much as 45 percent of members of these families lived in medium poverty, and almost one-third lived in hard poverty. As a result, 15 percent of the medium poor and 18 percent of the hard poor lived in 4 plus children families, while the average share of this population was below 5 percent (Table 4.10 and Statistical Appendix Table A15). It should be noted that the poverty level for multi-children families remains very high despite the special treatment of these families by the social protection system (higher rates of family allowance for the second and following children). 4.83 Age. Poverty headcounts show a clear pattern with age. They decrease significantly with the age bracket. For medium poverty, this is a decrease from 21 percent for children under 14, down to 7 percent for the elderly over 65, and for hard poverty, from 13 percent to 4 percent, respectively. Thus, in general, children are much more affected by poverty than are adults and the elderly (two times more than adults aged 45-64, and three times more than the elderly aged 65 and over). Observing narrower age brackets allows for more detailed description of the age- poverty pattern. Figure 4.3 shows that in fact poverty declines steadily with age for the youngest population, up to 30-34, but then it stabilizes somewhat with a slight tendency to increase until the age of 44. In the middle age group (45 and over), the next "round" of the poverty decline begins, but it ends at the age of 60-65. For those aged 65 and over the poverty rate increases once again, and at the age of 70-75 it is higher than at the age of 55-60. At age 75 and over some stabilization is visible, possibly due to the nursing allowance. In short, one may add to the overall 85 age-poverty pattern the following finding: first, the lower the age of a child, the higher the poverty rate, and, second, the higher the age of the elderly, the higher the poverty rate, at least up to 75. These changes in poverty headcounts reflect both the family life cycle and the labor market or wage cycle. Therefore, it may be expected that they would be fairly stable. Social Factors: Gender and Education 4.84 The impact of gender on poverty is not very clear; it depends on age, family size and education. 4.85 Gender. Figures A3 and A4 (Statistical Appendix) show that, in fact, there are no major differences in poverty rates between men and women. Poverty rates are almost equal up to age 30-39; men are apparently more affected by poverty at age 40-65, and then women became more affected. Higher poverty rates for the oldest women result from the pension schemes (lower pensions for women), while higher rates for the middle aged men may be possibly attributed to the higher differentiation of their material status. The indices in Table 4.11 provide additional evidence of the impact of gender on poverty. Table 4.11: Gender, Family Size, and Poverty WB2 Headcounts [%] Hard poverty Medium poverty Household size (no. of hhld members) Hhld headed by Hhld headed by man woman man woman 1 4.2 1.6 8.0 3.3 2 1.7 3.2 3.3 5.7 3 3.3 5.9 5.7 10.6 4 5.5 8.6 10.4 14.9 5 10.2 14.9 18.2 23.0 6+ 21.9 30.5 33.2 44.4 Total 8.2 9.4 13.8 15.1 Note: Headcounts refer to the population. Source: HBS 2001, own computation. 4.86 In 2001, poverty headcounts for the population of female headed households were, on the average, somewhat higher than those for male headed households, but the difference was fairly small. In the case of medium poverty, it was equal to 1.3 percentage points (13.8 percent for the first group, 15.1 percent for the second), and it was only 0.8 percentage points in the case of hard poverty. This pattern holds for all household sizes, except households of those living alone. For the singles, the situation looks quite different. Headcounts are actually significantly higher for men than for women, despite the fact that, on the average, the male LS is higher than the female LS. This is the result of the much higher living standard differentiation between men and women (the 2001 WB2 Ginis were equal to 0.279 for single men, but to only 0.223 for single women). 4.87 Education. Education is the last factor to be considered here in the analysis of the poverty profile. But it is one of the most important poverty factors. It appears that a university degree almost prevents people form falling into poverty, whether medium or hard. In 2001, the medium poverty (WB2) headcount for those who depended on the household head holding a university degree was equal to 1.3 percent, and thus it was over ten times lower than the country average (Table 4.12 and Statistical Appendix Table A14). At the same time, primary education 86 pushed the medium poverty headcount up to the level of 25 percent -- almost two times higher than the average, or fourteen (!) times higher than tertiary education. The proportions for hard poverty were even more pronounced, especially with respect to tertiary education, which gave a negligible poverty rate, equal to 0.4 percent (Statistical Appendix Table A14). 4.88 The way in which education influences the poverty rate is indirect. Education is correlated with many variables that help people to stay out of poverty. Education provides a better opportunity in the job market, increases mobility, has an impact on family status, etc. Nevertheless, it does not reward all groups equally. For example, those who live in rural areas benefit less than urban residents, and women benefit less than men. Medium poverty headcounts for educated women are higher than for educated men, although the difference is small: 1.1 percent for women with a university degree, 0.9 percent for men with a university degree, 5.2 percent for women with a secondary school diploma, and 4.9 percent for men with a secondary school diploma. Table 4.12: Education and Poverty Education of the household WB2 headcounts [%] Composition of the Poor [%] head Hard poverty Medium poverty Hard poverty Medium poverty Tertiary 0.4 1.3 0.5 0.9 Secondary general 3.3 6.6 11.8 14.1 Basic vocational 10.7 18.0 46.5 47.2 Primary 16.2 24.7 41.2 37.8 Note: Secondary general includes uncompleted tertiary; primary includes uncompleted primary. Source: HBS 2001, own computation. See also Statistical Appendix Tables A14 and A15 in the Statistical Appendix. Poverty Factors Combined 4.89 Poverty factors often overlap. But it would be useful to understand how each single factor works separately, and also to understand the joined or combined impact of selected factors. 4.90 In order to examine these questions, logit models explaining the probability of falling into poverty, both medium and hard, have been estimated. The main poverty factors have been selected as regressors. However, those factors that are: (i) clearly insignificant according to the standard statistical tests; or (ii) that enter into the models with questionable signs have been excluded. This is the case, for example, with variables depicting the disability status of household members or dummies for voivodships. On the other hand, work status and dependency have been measured as shares (accompanied by the household size) and not as integer ordinal numbers. This form seems more convenient for the logit regression analysis, allowing one also to control for household size. 4.91 The estimation results for two LS concepts, (GUS reg and WB2) are displayed in Table 4.13. These results may be given the following interpretation. First, the sign of each coefficient informs us whether the impact of a given variable on poverty (or more precisely, on the probability of being poor) is positive or negative, while keeping all other variables constant. Second, coefficients within a group of dummy variables (concerning, for instance, the socioeconomic category of a household or the education of a household head) may be reasonably compared. This means that they indicate which regressor has a stronger impact on poverty. Third, given the formula of logistic CDF, the probability of being poor for various combinations 87 of poverty factors may be estimated and compared. In fact, poverty headcounts would be estimated in this way.36 4.92 In general, estimates of the logit functions confirm major findings of the analysis based on headcounts. As may be seen, almost all variables in the models are statistically significant and their coefficients have correct signs. The only exception includes three dummies expressing the main source of income, namely, retirement pension, farming and mixed (but farming and mixed are significant in the medium poverty GUS reg model). This gives evidence that living on a retirement pension or on farming per se does not sufficiently explain the probability of being poor. Or, in other words, the poverty of farmers and their families, as well as of retirees, cannot be significantly distinguished from the worker situation, if other poverty factors ­ such as unemployment, household size, education, etc. ­ are held constant. However, the total expenditure approach (GUS reg) shows that falling into medium poverty is in fact more likely for farmer families than for worker families, and even more for mixed families (see signs and absolute values of the coefficients in the GUS reg models). Table 4.13: Logit Models for Medium and Hard Poverty Regressor Medium poverty Hard poverty GUS reg WB2 GUS reg WB2 Place of residence: rural (dummy, ref urban area) 0.458 0.148 0.492 0.216 Soc-ec category of a hhld (dummies; ref worker) farmer * 0.200 -0.049 0.091 -0.128 mixed * 0.491 -0.002 0.216 -0.189 self employed -0.297 -0.545 -0.439 -0.637 retiree * -0.070 -0.088 0.000 -0.071 disabled/survivor& 0.410 0.286 0.393 0.204 social beneficiary 0.819 0.664 0.734 0.571 Household size 0.281 0.388 0.295 0.385 Share of employed -1.092 -1.509 -1.271 -1.637 Share of unemployed 1.862 1.804 1.889 2.073 Share of dependents 1.852 1.631 1.928 1.791 Age of the household head -0.012 -0.017 -0.014 -0.018 Gender of a hhld head: woman (dummy; ref man) 0.400 0.389 0.359 0.377 Education of a hhld head (dummies; ref primary) tertiary -2.730 -2.859 -2.902 -3.218 secondary -1.451 -1.526 -1.451 -1.599 vocational -0.639 -0.686 -0.572 -0.708 Land area > 1 ha (hectares) -0.036 -0.029 -0.030 -0.030 Constant -2.344 -2.429 -2.921 -3.052 Pseudo R2 0.228 0.249 0.232 0.260 Notes: Dependent variable: Probability of being poor. * Small fonts, italics indicate statistically (highly) insignificant coefficients. Coefficient for disabled/survivor variable in WB2 hard poverty mode is significant at only the 10 percent & level. Source: HBS 2001, own computation. 36In addition, pseudo-R2s in the bottom line of Table 4.13 inform us that the goodness of fit given this type of model is satisfactory and that all variables combined explain hard poverty slightly better than medium poverty. 88 4.93 Other estimation results are more transparent. All models give evidence that the rural population is more often driven into poverty than is the urban population. This is shown by a positive sign of a dummy "rural" coefficient. But it may be noted that WB2 models attribute a much smaller impact to the "residence" variable than to GUS reg: for medium poverty, coefficients are equal to 0.15 and 0.46, respectively, while for hard poverty they are 0.22 and 0.49, respectively. 4.94 It is also evident that the higher the share of unemployed and dependents in a household is and the higher the household size itself is, the higher the probability is of being poor (positive signs of the coefficients). It may also be read that, given WB2 concepts, unemployment is more important than dependency in driving people into poverty, especially into hard poverty. The GUS reg approach does not allow us to say which of these two factors is more important. 4.95 According to the estimates, poverty is more likely for social benefit recipients and disability/survivor pensioners than for workers and their families (keeping all other factors constant). It is also more likely for female-headed than for male-headed households, and the gender factor proves fairly significant. It may seem, as well, that people who live mainly on incomes from self-employment would have lower poverty rates compared to workers (negative signs of the coefficients), but models which use WB2 suggest a more intensive impact of self- employment than do GUS reg models. 4.96 The impact of education on the probability of being poor (while keeping all other variables unchanged) is certainly very important. This is reflected in the signs and absolute values of all coefficients regarding education, which show that poverty decreases considerably if the education of the household head increases upward from primary education. This decrease in poverty is especially pronounced when moving from the secondary to the tertiary education level. 4.97 Finally, one may ask what the probability would be of being poor in given circumstances, for example, in a worker family of four living in an urban area, headed by a uneducated male aged 40, with two employed and two dependent children, and no land. For medium WB2 poverty, the answer would be 16 percent, and this is the predicted headcount. Had this household lived in a rural area, the predicted medium poverty headcount would be equal to about 18 percent. Had the household contained only one employed person, one unemployed person and two children (still living in a city), its estimated medium poverty rate would be over 30 percent. Finding poverty rates for other combinations of poverty factors is possible as well. C. CHANGES IN POVERTY IN 1994 - 2001 4.98 Changes in poverty during the period 1994-2001 have been investigated with the use of three living standard measures: GUS, GUS reg, and WB0. These measures were fully described in section A. Both medium poverty and hard poverty have been considered, and delimited in the way discussed. It should be remembered that in this study poverty lines are kept constant in real terms over the whole period examined.37 Overall Picture 4.99 Changes in poverty headcounts are displayed in Table 4.14 and in Statistical Appendix Table A16. In addition, Figures 4.4 and 4.5 in this chapter present the main tendencies, clearly showing that poverty decreased in the years 1994­98, and that it increased thereafter. This is the 37This is not the case in GUS publications. See, for example, GUS (2002), Chapter 9. 89 case with both medium poverty and hard poverty, and it does not depend on the LS concept used.38 4.100 The poverty decline in the middle of the decade was more pronounced than the poverty increase that came later. In 1994-98, the GUS medium poverty headcount dropped by 7 percentage points, from 20 percent to 13 percent, while the rise in 1998 ­ 2001 was equal to 3.3 percentage points. According to WB0, the decline was less sharp, from 18 percent in 1994 to 12.6 percent in 1998, and so the increase thereafter was less sharp. This difference between WB0 and GUS is not surprising. Quite often, WB measures provide a somewhat fuzzier picture. Table 4.14: Poverty in 1994 - 2001: An Overview Headcounts [%] LS concept 1994 1995 1996 1997 1998 1999 2000 2001 Medium poverty GUS 20.0 18.3 16.1 15.0 12.9 14.9 15.1 15.2 GUS reg 19.2 17.3 15.3 14.2 12.0 14.0 14.3 14.2 WB0 18.0 16.9 14.6 13.6 12.6 13.5 14.5 14.8 Hard poverty GUS 12.4 10.8 9.5 9.1 7.7 8.9 9.2 9.6 GUS reg 11.7 10.2 8.8 8.3 7.0 8.1 8.6 9.0 WB0 10.5 9.6 8.2 8.0 6.8 7.7 8.6 8.9 Note: For methodology, see Section A and Methodological Annex. Source: HBS 2001, own computation. See also Statistical Appendix Table A16. Figure 4.4: Medium Poverty, Poland 1994 - 2001 22.0 20.0 (%) 18.0 16.0 14.0 Headcounts12.0 10.0 1994 1995 1996 1997 1998 1999 2000 2001 CSO CSO reg WB0 Source: HBS 2001, own computation. 38GUS published results differ slightly, but they also show first, the decline, and next, the increase, of poverty (see Statistical Appendix Table A16, and GUS 2002 just quoted). The difference between poverty trends presented in this report and published by GUS results from the methodology. GUS, for example, does not keep poverty lines constant in real terms. It uses relative lines, subjective lines, current SA thresholds, and subsistence or social minimums. None of these lines is constant in real term (relative to general CPI). 90 Figure 4.5: Hard Poverty, Poland 1994 - 2001 13.0 12.0 11.0 (%) 10.0 9.0 Headcounts 8.0 7.0 6.0 1994 1995 1996 1997 1998 1999 2000 2001 CSO CSO reg WB0 Source: HBS 2001, own computation. 4.101 Before looking for the determinants of poverty trends, one should consider changes in the methodology of measurements which might have affected the results. In fact, in 1997 ­ 98 GUS introduced some adjustments to the household classification as well as some corrections to expenditure concepts (see notes to Statistical Appendix Tables A17 and A18). However, it would seem that these adjustments could not bias poverty indices significantly. Moreover, the data in Statistical Appendix Tables A17 and A18 show that key statistics based on the HBS (such as sample composition, unemployment ratio, etc.) give a correct view over the whole period under investigation. The share of households living on social income appears to be the only index that might be questioned (it is surprisingly low in 1998, as seen in Statistical Appendix Table A18). Average household size seems a little too low as well, but it is still acceptable. 4.102 The trends in GDP growth rate, and the changes in average income (wages, pensions), material inequalities and unemployment should be considered as among the causes of poverty fluctuations (see Statistical Appendix Tables A16 and A19). Demographic variables and education may be ignored since the period under investigation is rather short. It is easy to see that trends in poverty closely follow changes in the unemployment rate, both registered and BAEL. Figures 4.4, 4.5 and 4.6 provide the evidence for this. The figures show that the "V-shape" of the unemployment rate clearly reproduces the "V-shape" of the poverty rates. This is not surprising, since unemployment has been noted as one of the main poverty factors. 91 Figure 4.6: Unemployment Rate, Poland 1994 - 2001 18.0 16.0 14.0 Unempl [%] 12.0 Rate10.0 Registerde 8.0 6.0 1994 1995 1996 1997 1998 1999 2000 2001 Source: HBS 2001, own computation. Changes in the Poverty Profile 4.103 Statistical Appendix Tables A20 ­ A23 and Figures A5 ­ A10 display changes in the medium and hard poverty profiles measured with WB0. Only major poverty factors (i.e., place of residence, unemployment, source of income, education, and family status) are considered there. 4.104 The poverty profile appears quite stable over the entire eight-year period, despite changes in the overall standard of living and the new social policy rules implemented in the middle of the decade,39 and targeted at some groups in the population. Residents of large cities have always been less poor than those living in small towns, and the difference between their headcounts has not changed remarkably. People living on retirement pensions have been less affected by poverty than many other groups; the same applies to people living alone and couples with no children. However, the relative positions of some groups changed slightly, and some poverty factors seem to have become more important. 4.105 In 1998 and 1999, for example, the distance between rural and urban populations was greater than in other years. At that time, medium poverty headcounts for the rural population were equal to 1.4 (relative to the country average), while at present they are equal to 1.3. This is also the case with farmers: relative poverty rates were considerably higher for this group in 1998- 1999 than they are at present (Table A21 and Figure A5 in the Statistical Appendix). In 1997 and 1998, the poverty of people living on social incomes was relatively higher than in 1999-2001. People living in households with unemployed members were also more affected by poverty in these years. 4.106 Education seems to be more important at present than in the past. This may be seen through the changes in the relative poverty rate for households headed by uneducated persons. This (relative) rate increased steadily during the entire period, from 1.4 in 1994 up to 1.7 in 2001 for medium poverty, and from 1.5 up to 1.8 for hard poverty (Statistical Appendix Tables A21 and A23). Other poverty factors have not revealed similar impacts. 39Changes in social assistance, unemployment benefit schemes, and family and nursing allowances. 92 D. TEMPORAL INCIDENCE AND MOBILITY OF POVERTY 4.107 Having examined the magnitude and factors of poverty it is tempting to see if poverty is a long-lasting phenomenon.40 This is possible, since the Polish HBS includes a rotating panel sub- sample: 4,859 households were interviewed each year during the period 1994-96 and 3,051 households were interviewed during the period 1997-2000. This feature of the HBS allows, to some extent, the analysis of (i) the temporal incidence of poverty; and (ii) mobility in and out of poverty. The size of the panel sub-sample is affected considerably by attrition, hence the difference in size between the two periods. In order to reduce this effect, the weights were rescaled to preserve the socio-demographic characteristics (size of the place of living, number of people in the household, socioeconomic type of the household) of the sample from the last years of the panel, 1996 and 2000, respectively. The living standard concept used in the temporal incidence and mobility analysis of poverty was WB0 (i.e., regionally deflated total household expenditures [cash and in kind] on selected goods and services, less: support for other households, non-current taxes, durables and rents [sewerage/water and energy expenditures were included]). This measure was made equivalent by using the Polish Social Assistance scales. Statistical Appendix Table A24 contains a comparison of the LS measure (WB0) distribution between the panel and the cross-section samples. The latter samples have higher values of the mean, median and standard deviations. WB0 is more equally distributed in the panel samples. This can be associated with a higher rate of attrition among households from the top of the distribution. This is not necessarily a problem in terms of poverty analysis, since the focus is on the other end of the distribution. Table A24 also shows that poverty headcounts in the panel samples differ only slightly from those in the cross-sectional samples. Temporal Incidence of Poverty 4.108 The short time-span of the panel sub-samples does not allow for a proper duration analysis of poverty spells. It is possible, however, to examine the temporal incidence of poverty in terms of the number of years spent in poverty during the time of the panel. Table 4.15 provides the general figures. If it is borne in mind that the poverty headcounts of the panel samples are higher than those of the cross-sectional samples, long-term poverty rates appear to be rather low. The fraction of individuals not affected by poverty at all was similar in both periods ­ about 72 percent. On the other hand, the fraction of people remaining in hard poverty during the entire periods of the panels was about 3 percent. Further analysis was focused on these two groups. Statistical Appendix Table A25 shows how the temporal incidence of poverty varied between socioeconomic groups. Generally, the pattern was similar to that of cross-sectional poverty. Living as a couple with no children, being self-employed, working or retired, living in a household with a head having higher education, and living in the largest cities increased the chance of staying out of poverty during the period of the panel and decreased the chance of remaining in poverty throughout the period. On the other hand, living as a couple with three or more children, being a farmer or living on social welfare, living in a household with a head having only an elementary education, and living in rural areas decreased the chance of staying out of poverty and increased the chance of remaining in poverty. A comparison of the two panels shows that the differences between the socioeconomic groups (established by family type, socioeconomic status and the education of the household head) have increased. Two groups stand out in terms of the chance of being among the long-term poor: those living on social transfers and couples with four or more children. 40Separate analysis of the two panels has been carried out by Panek (2001b). 93 Table 4.15: Poverty Temporal Incidence, 1994-2000, Number of Years in Poverty 0 1 2 3 4 Total Hard poverty 1994-1996 82.5% 9.9% 4.8% 2.8% x 100.0% 1997-2000 81.6% 9.8% 4.2% 2.7% 1.6% 100.0% Medium poverty 1994-1996 71.3% 13.6% 8.5% 6.7% x 100.0% 1997-2000 71.8% 12.6% 7.0% 4.6% 4.0% 100.0% Note: Equivalized WB0 LS measure, individual level. Source: HBS 1994-2000, own calculations. Poverty Mobility 4.109 The temporal incidence of poverty analysis presented up to this point did not take into account the movements into and out of poverty. Table 4.16 and Statistical Appendix Table A26 compare two groups: (i) "In" - those who were not poor during the first year, entered poverty during the second or third year of the panel and remained poor until the end; and (ii) "Out" - those who were poor during the first year, escaped poverty during the second or third year of the panel and managed to remain out of poverty until the end. The general figures in Table 4.16 show that the magnitude of mobility both in and out of poverty was larger during 1994-96 than during 1997-2000. In both panels, and at both poverty levels, the fraction of those moving into poverty was lower than the fraction exiting from poverty (the large fractions of the short-term poor sustain the poverty headcounts in the cross-sections). This difference was smaller in the second panel, which is in line with the reverse of the poverty headcount trend which took place in 1998. Statistical Appendix Table A26 shows the fractions of movers in socioeconomic sub- samples. The fraction of hard poverty entrants was highest among those living as a couple with three or more children, farmers and those living on social transfers, those living in households with a head having an elementary education, and those living in rural areas. Poverty entrance mobility generally had a larger between-group variation than poverty exit mobility. Summary of Panel Analysis 4.110 The analysis of the two panel sub-samples of the HBS shows that the most vulnerable groups in terms of cross-sectional poverty are also affected by long-term poverty and are at the greatest risk of entering poverty. At the same time, they have better chances of exiting from poverty but this indicates that their standard of living is more volatile. The impact of factors influencing the chance of staying out of poverty for the entire period of the longitudinal study has increased between the two panels. Living on social welfare greatly increased the chance of being long-term poor (both medium and hard) and decreased the chance of staying out of poverty. This may be an indicator of the fact that social welfare policy does not help households escape from poverty. The level of mobility has decreased between the two panels. 94 Table 4.16: Poverty Mobility, 1994-2000, Movements In and Out of Poverty In Other Total Out Other Total Hard poverty 1994-1996 4.2% 95.8% 100.0% 6.9% 93.1% 100.0% 1997-2000 2.1% 97.9% 100.0% 3.2% 96.8% 100.0% Medium poverty 1994-1996 6.0% 94.0% 100.0% 9.6% 90.4% 100.0% 1997-2000 3.8% 96.2% 100.0% 4.6% 95.6% 100.0% Note: Equivalized WB0 LS measure, individual level. Source: HBS 1994-2000, own calculations. E. CONCLUDING REMARKS 4.111 Poverty in 2001 varied significantly across regions. The north, southwest and southeast had higher poverty rates than the center of the country. Rural poverty dominated there, and the indirect statistical evidence shows that in most cases this was poverty among former PGR (state farms) workers. 4.112 Unemployment is certainly the most important single factor affecting poverty. The new approach has strongly confirmed earlier findings. Apart from unemployment, the important factors are lack of education and a large number of children or dependents in the family. Age is negatively correlated with poverty, or in other words, poverty decreases with age. There is, however, some evidence of an increase in poverty among the elderly. Farmers are poorer than urban workers, but farming seems to be less pronounced in driving people into poverty than it appeared to be in the light of previous studies. 4.113 The analysis of poverty trends over the last eight years shows that the above pattern of poverty remains relatively stable, with poverty changes determined mostly by fluctuations in the unemployment rate. However, the position of farmers ­ relative to other groups ­improved slightly in 2000-01, and lack of education seems more important as a poverty factor now than at the beginning of the decade. 4.114 Poverty analysis, conducted according to the World Bank standards, shows a picture generally similar to that found in earlier Polish studies. But the differences between rural and urban areas, as well as between the regions, are less pronounced. This finding should be treated with caution, however, for at least two reasons. 4.115 First, given the available HBS data (with incomplete and imprecise information on the quality of consumer durables and housing), the World Bank approach as exercised in this research may underestimate the level of rural poverty. This could be alleviated only if separate, specific studies on living standards were conducted. 4.116 Second, it should be noted that, if (according to the new methodology) people who live in rural areas or in some voivodships appear relatively less poor, this is so only insofar as they stay home. Only then can they take advantage of the lower prices and flows of implicit consumption from their housing and durables. However, this means that they are locked into the poor areas, with lower access to the labor market and, generally, with lower life chances than elsewhere. 4.117 The analysis of the two panel sub-samples of the HBS shows that the most vulnerable groups in terms of cross-sectional poverty are also affected by long-term poverty and are at the greatest risk of entering into poverty. The level of mobility has decreased between the two panels. 95 REFERENCES GUS [1993], Living Conditions of the Population. GUS [1994], Living Conditions of the Population. GUS [1995], Living Conditions of the Population. GUS [1996], Living Conditions of the Population. GUS [1997], Living Conditions of the Population. GUS [1998], Living Conditions of the Population. GUS [1999], Living Conditions of the Population. GUS [2000], Living Conditions of the Population. GUS [2000a], Comparaisons de vie en France et en Pologne. GUS [2001], Living Conditions of the Population. GUS [2002], Living Conditions of the Population. Czyewski, A.B., M. Góralczyk­Modzelewska, and E. Saganowska, (2001), Regionalne zrónicowanie kapitalu ludzkiego. Warsaw: Zaklad Bada Statystyczno-Ekonomicznych GUS i PAN. Gorzelak, G. 2003 [1999], Historia ­ Transformacja ­ Przyszlo: Szkic o strukturze polskiej przestrzeni i jej zmianach, in: Gorzelak, G. et al. ed. Rozwój ­ Region ­ Spoleczestwo, Warsaw-Katowice: Europejski Instytut Rozwoju Regionalnego i Lokalnego. Kawczyska-Butrym, Z. [2001a], Postulowane sposoby rozwizania problemów mieszkaców osiedli bylych PGR, Polityka Spoleczna, 8 (329), pp. 25-26. Kawczyska-Butrym, Z. [2001b], Sytuacja yciowa mieszkaców bylych PGR. Polityka Spoleczna, 8 (329), pp. 15-17. Niemczyk, A. [2001], Poziom rozwoju spolecznego w nowym ukladzie administracyjnym. Przegld Statystyczny, 3-4: pp. 289-299. Okrasa, W. (2000), "Who are Poland's Long-Term Poor? Household Risk-Managing Capabilites According to Panel Data 1993-1996," Statistics in Transition, Vol. 4, No. 4, pp. 841­882. Panek, T. [1996], Wielowymiarowa analiza sfery ubóstwa, in: GUS, Ubóstwo w Polsce w wietle bada budetów gospodarstw domowych, Studia i Analizy Statystyczne. Panek, T. [2001a], Sfera niedostatku i jej determinanty w Polsce, Wiadomoci Statystyczne, 1, pp. 21-32. Panek, T. [2001b], Zmiany w sferze ubóstwa ­ mity i rzeczywisto, Wiadomoci Statystyczne, 8, pp. 39-50. 96 Panek, T. [2001c], Wymiary ubóstwa w Polsce w latach 1996-1999, Wiadomoci Statystyczne, 11, pp. 37-55. Sen, A. [1992], Inequality Reconsidered, Russell Sage Foundation, Clarendon Press. Szulc, A. [1996], Skale ekwiwalentnoci w pomiarach ubóstwa i dobrobytu, in: Golinowska, S. ed. Polska bieda. Kryteria. Ocena. Przeciwdzialanie, Instytut Pracy i Spraw Socjalnych, pp. 204-219. Szulc, A. (2000), "Economic Transition, Poverty and Inequality: Poland In the 1990s," Statistics in Transition, Vol. 4, No. 6, pp. 997­1017. Szulc, A. (2001), Household Equivalence Scales for Poland: How Sensitive to the Method of Estimation Are They? Research Bulletin, Vol. 10, No 1, pp. 5 ­ 24. Tarkowska, E., ed. [2000], Zrozumie biednego. Warszawa: IFiS PAN and typography. UNDP [2000] Raport o rozwoju spolecznym Polska 2000: Rozwój obszarów wiejskich. Warsaw: UNDP. World Bank [1995], Understanding Poverty in Poland, Washington DC. Zablocki, G. et al. [1999], Ubóstwo na terenach wiejskich pólnocnej Polski, Wydawnictwo Uniwersytetu Mikol aja Kopernika, Toru. 97 Methodological Annex: Imputation Procedures Rent Imputation One of the main issues related to housing expenditure has been the wide variety of legal titles to dwellings. Table A.4.1 presents the number of households possessing each type of title. Rent payments implied by each type of legal title to dwelling were not fully comparable. Almost 50 percent of all households occupied own dwellings (where rent payments should not be very high) and another 25 percent (regulated rent and tenancy) rented dwellings outside of the free market (where rent payments also should not be very high). Only a small fraction of households had been renting dwellings on the free market. In order to unify the rental payments of all households, rents have been imputed using theoretical values estimated in a regression using data supplied by households from the cooperative (not encumbered with loans) category of the legal title to dwelling. Rental payments of those households were considered an approximation of actual costs associated with living. Table A.4.1: Legal Title to Dwelling, Poland 2001 Type of title N % property, not encumbered with loans 15,033 47.1 regulated rent 5,315 16.7 co-op, not encumbered with loans 5,065 15.9 Tenancy 3,491 10.9 free-market rent 1,000 3.1 living with family 699 2.2 property, encumbered with loans 538 1.7 co-op, encumbered with loans 429 1.3 renting part of dwelling 179 0.6 social dwelling 107 0.3 Other 45 0.1 Total 31,901 100.0 Source: HBS 2001 (not weighted). The data did not allow determining whether rents included payments for utilities. Therefore, only households that reported such payments separately (see Table A.4.2) were considered further). This sample was further trimmed to exclude outliers of a newly defined variable: the logarithm of real rent (in June prices) per square meter (of usable surface). Observations between the first and the nineteenth ventile were selected. Table A.4.2: Utilities and Rents Paid by Co-op Members, Poland 2001 Frequency [N] rent = 0 rent > 0 Total utilities = 0 139 170 309 utilities > 0 43 4,713 4,756 Total 182 4,883 5,065 Source: HBS 2001 (not weighted). The HBS provided many variables representing potential factors which could explain, in a regression model, the log of real rent per square meter (see Table A.4.3). In order to avoid estimation solely using dummy regressors, three continuous variables were introduced to the model: number of rooms in dwelling, number of people in household, and quarterly rate of 98 unemployment in voivodship. The inclusion of the rate of unemployment was intended to relate the log of real rent per square meter to the general macroeconomic situation of the region. Table A.4.3: Potential Rental Payments' Explanatory Factors Factors Variables' description number of inhabitants in place of living reference category: rural area dummy variable: 500+ thousand inhabitants dummy variable: 100-500 thousand inhabitants dummy variable: 20-100 thousand inhabitants dummy variable: -20 thousand inhabitants number of rooms in dwelling continuous variable number of people in household continuous variable type of building reference category: single household dummy variable: multi household voivodship reference category: Mazowieckie dummy variable: Dolnoslaskie,Lubuskie, Malopolskie, Opolskie dummy variable: Pomorskie, Zachodniopomorskie dummy variable: Lubelskie, Podkarpackie, Podlaskie, Swietokrzyskie dummy variable: Lodzkie dummy variable: Slaskie dummy variable: Kujawsko-Pomorskie, Warminsko- Mazurskie dummy variable: Wielkopolskie year building built reference category: 1979+ dummy variable: -1944 dummy variable: 1945-1970 dummy variable: 1971-1978 type of heating system reference category: electric/gas/other dummy variable: network dummy variable: local dummy variable: furnace ­ wood/coal running water reference category: no dummy variable: yes water closet reference category: no dummy variable: yes bathroom reference category: no dummy variable: yes type of hot water supply reference category: no hot water dummy variable: network dummy variable: locally heated type of gas supply reference category: no gas dummy variable: network dummy variable: container telephone reference category: no dummy variable: yes quarterly rate of unemployment in continuous variable voivodship* Sources: HBS, and *GUS (published). 99 Table A.4.4 contains the estimation results of a weighted least squares regression with a stepwise model selection algorithm. The p-values of entering and staying in the model were not restrictive - respectively: 0.1 and 0.2. The final model resulted in an R2 coefficient of only 7.88 percent; however, after inverting the logarithm and multiplying by the number of square meters the squared coefficient of correlation between the predicted and actual values was 40.64 percent. The estimation results are of limited interest to this study. However, a brief examination proves them to be reasonable: rents increase with size of city and decrease with rate of unemployment. Table A.4.4: Estimation Results, Poland 2001 Regressors included in final model Estimate St. error constant term 0.467 0.073 dummy variable: -20 thousand inhabitants 0.093 0.033 dummy variable: 20-100 thousand inhabitants 0.179 0.032 dummy variable: 100-500 thousand inhabitants 0.225 0.033 dummy variable: 500+ thousand inhabitants 0.302 0.034 dummy variable: Lodzkie -0.076 0.017 dummy variable: Pomorskie, Zachodniopomorskie 0.103 0.016 dummy variable: Slaskie 0.109 0.015 dummy variable: locally heated hot water 0.023 0.010 dummy variable: local heating 0.068 0.034 dummy variable: multi household 0.121 0.061 dummy variable: gas in container 0.029 0.017 quarterly rate of unemployment -0.578 0.145 Source: HBS, own calculations. Imputation of the Consumption of Durables In order to reflect the differences in the levels of welfare between households equipped and unequipped with certain durables, the consumption of such goods has been imputed. The imputed figures, wherever feasible, were taken from HBS (ownership, expenditures), following the intention of making maximum use of information from the survey. A linear depreciation schedule has been assumed, with a 15 percent rate of depreciation in the first year and 9 percent for all other years, reflecting a 10-year durable lifetime (there were some exceptions to this rule, such as sewing machines). For durables bought in the year of the survey (reported in the expenditure data), the initial prices have been taken from HBS. However, extraordinarily low actual expenditures (prices) have been neglected, and the scheme for "old" goods has been applied instead. For "old" durables (i.e., bought in earlier years), four alternative approaches have been proposed: (i) the prices taken were average prices in ex ante (before durable imputation) consumption quartiles; (ii) the prices taken were median prices; (iii) the prices were taken from external GUS data; or (iv) the prices were taken from the market. The choice depended on the number of survey cases with the required data, and on the quality of information on expenditures. Table A.4.5 provides details. Table A.4.6 reports numbers for average household equipment, by residence. It should be noted that rural households are often better equipped with durables than urban households, despite the fact that average consumption (whatever the concept used) is lower in rural areas. But it should be remembered that many durables may be used for production purposes by rural residents (cars, deep freezers, etc.). Urban residents, on the other hand, seem to use durables mostly for pure 100 consumption (with the exception of firm-owned cars). Given the HBS data available, it is not possible to distinguish between the two purposes (production and consumption). Table A.4.5: Durable Consumption Imputation Values, Poland 2001 HBS prices Imputation price Durable Ex ante consumption GUS quartiles Median Mean prices New durables Old durables Q1 Q2 Q3 Q4 tv (color) 928 1 121 1 144 1 290 999 1 145 996 >=200 quartile price Radio 90 166 55 168 50 125 x x median price radio casette recorder 245 224 241 270 169 248 x >=75 quartile price tape-recorder 245 224 241 270 169 248 x >=75 quartile price electric gramophone 90 166 55 168 50 125 x x median price cd-player x 1 350 1 263 798 484 1 036 721 >=400 GUS price stereo 818 719 717 877 649 794 x >=450 quartile price vcr (player) 309 562 711 690 599 638 x >=310 quartile price video camera 238 177 257 399 312 100 x >=100 quartile price still camera 238 177 257 399 312 100 x >=100 quartile price pc + internet 703 1 305 1 039 972 130 1 033 x >=450* quartile price* pc 703 1 305 1 039 972 130 1 033 x >=450* quartile price* printer 703 1 305 1 039 972 130 1 033 x x market price* electric washing machine + dryer 870 769 927 1 092 1 000 936 x x mean price / 4 automatic washing machine 870 769 927 1 092 1 000 936 1 474 >=390 quartile price electric vacuum cleaner 371 341 310 319 280 330 368 >=360 quartile price refrigerator 1 300 1 200 1 236 1 291 1 250 1 260 991 >=300 mean price deep freezer 1 300 1 200 1 236 1 291 1 250 1 260 1 914 x mean price >=300;< microwave oven 407 415 377 351 120 382 588 =1000 mean price food processor 126 94 89 111 72 104 167 x mean price dishwasher 870 769 927 1 092 1 000 936 x >=1000 quartile price sewing machine 2 81 167 43 14 284 806 x GUS price* bicycle (not child) 431 429 458 530 400 466 805 >=400 quartile price car household-owned** 27 16 34 28 544 749 400 303 26 760 27 610 24 139 >=8691 quartile price Notes: x ­ not available; * Depreciation rates exceptions: PC (20% first year, 15% otherwise), printer (16% all years), sewing machine 6% all years); used cars (9% all years). ** If a used car was bought, the imputation price was based on the quartile price >= 1200. If a household used a firm car, the quartile price has been taken as a base. Sources: HBS 2001, own calculations, and GUS 2002, Prices in the national economy in 2001, t. 27, pp.152, 3, 6. 101 Table A.4.6: Average Number of Durables per 100 Households, Poland 2001 Place of residence Durable All households Urban Rural tv (b&w)* 3.4 6.2 4.4 tv (color) 116.4 107.1 113.1 radio 55.6 64.1 58.6 audiocassette recorder 52.9 47.9 51.1 tape-recorder 6.9 6.2 6.7 electric gramophone 2.1 1.1 1.7 cd-player 12.9 6.8 10.8 stereo 46.1 28.3 39.9 vcr (player) 60.6 44.6 55.0 video camera 5.4 2.0 4.2 still camera 63.5 42.5 56.1 pc + internet 10.2 3.2 7.8 pc 13.6 6.7 11.2 printer 15.4 6.0 12.1 electric washing machine + dryer 29.1 66.9 42.4 automatic washing machine 83.4 56.6 74.0 electric vacuum cleaner 96.3 88.4 93.5 electric floor-polisher* 2.2 1.2 1.8 refrigerator 98.7 98.2 98.5 deep freezer 29.4 55.4 38.6 microwave oven 24.7 15.5 21.4 food processor 58.3 53.3 56.5 dishwasher 2.9 1.4 2.4 sewing machine 43.3 47.4 44.8 bicycle (not child) 72.7 123.4 90.6 motorcycle, scooter, motorbike* 2.0 7.2 3.8 car (private) 45.4 56.2 49.2 car (firm) 1.5 1.1 1.4 Note: * Durables not considered in the imputation procedure (assumed zero utility or the price not possible to establish). Source: HBS 2001, own computation. wb13696 P:\POLAND\PREM\Living Standard Assessment\4RED\Part 1\Papers\FINALPLS Vol 2 Chap1-4 0323.doc March 25, 2004 10:40 AM 102 PART II: IDENTIFYING PARTICULARLY VULNERABLE GROUPS 5. REGIONAL INEQUALITIES IN LIVING STANDARDS Grzegorz Gorzelak1 A. INTRODUCTION: HISTORICAL HERITAGE OF POLISH SPATIAL PATTERNS 5.1 This chapter presents the main dimensions of the territorial (regional-local) differentiations in the Polish socioeconomic space. The paper's main focus is on the last period (i.e., the period of the post-socialist transformation after 1989), with special reference to the last period, 1998-2002. 5.2 As demonstrated elsewhere (Gorzelak, Jalowiecki, 2002), the present shape of the territorial differentiations has strong historical underpinnings. The spatial process has shown exceptional stability over time, which caused the French historian Fernand Braudel to label them as belonging to the processes of "long duration." Several regional studies conducted in Poland reveal the strong impact of historical factors on the current socioeconomic spatial patterns. 5.3 The contemporary patterns of the Polish socioeconomic space date back hundreds of years. From the beginning of Poland's history, a division along the east-west axis determined different development paths for individual parts of the country. Starting from the first wave of urbanization in the thirteenth to fifteenth centuries, towns were significantly more frequently established in the western rather than the eastern part of Poland. For many centuries, transport routes in the western part were denser and the level of farming culture was higher. The period of the partitions (1795-1918)2 reinforced those disparities even more. The differences established earlier were increased not only by the uneven distribution of industrialization (which was much more intense in the areas west of the Vistula River), but also by the fact that individual Polish regions belonged to different political organisms, which was reflected in the institutional structures, organizational capabilities, work culture, self-government traditions, etc. Those disparities were not cancelled out during the Second Republic (1918-1939), while efforts at modernization (of which COP -- along with the construction of the harbor and the city of Gdynia -- was the most important undertaking) were interrupted by the outbreak of World War II. 5.4 The development level and the structure of the territories which in 1918 joined to form the independent Polish state were markedly different from the rest of the country. These territories were better equipped with a material infrastructure (roads, railways, communal facilities), and had a more dense urban structure. After World War II, Poland's borders were shifted westward at the expense of the eastern (and less developed) territories taken over by the former USSR by force of the Treaty of Yalta. The western and northern territories were populated by migrants from the eastern regions lost to the USSR and from the underdeveloped, overpopulated agricultural regions of eastern and central Poland. During the period of post-War 1This chapter was written with the assistance of Maciej Smtkowski. 2 The last partition took place in 1795, and Poland was divided among three states: Russia took the eastern part, Prussia, the western part, and Austria, the southeastern part. In what is now Poland (i.e., after its frontiers moved westward after World War II); there is a fourth historical part, the northern and most western part, which for several centuries had been part of the German states. communist Poland (the People's Republic of Poland) the investment effort mainly targeted areas which had already received some investment and the newly established centers for natural resource extraction. Those processes seldom reached the areas of eastern Poland, which continued to be referred to as "Poland B." 5.5 The transformation period (1990-2002) has influenced the regional patterns but has not introduced any "revolutionary" changes. The shift from an industry-driven to a service-oriented pattern of economic development has led to the accelerated growth of strong urban centers, which assumed the positions of regional leaders of growth and structural change. Thus, the traditional dimension of an urban-rural divide has been replaced by a metropolitan - non-metropolitan divide. The traditional east-west divide has not been changed and has even been reinforced. 5.6 As a result of these new processes, Poland's regional differentiations have grown, although not to the point where they exceed the "civilized" patterns of the more developed Western European countries. 5.7 The regional processes are an outcome of the more general changes in the global and national situation. To understand these regional dynamics and structural changes, it is important to analyze the entire national setting. This is undertaken in the first sections. Section C describes the current regional and (for some phenomena) the local patterns of territorial differentiations. Section D provides insight into the dynamics of some of the phenomena and processes in the transformation period, with special reference to its last phase of slower growth and weaker structural changes. Section E summarizes certain major observations and provides an outlook on future regional developments in Poland, together with some policy recommendations. B. THE CURRENT STATE OF THE POLISH TRANSFORMATION 5.8 In the process of the Polish post-socialist transformation three periods can be distinguished that demonstrate the major economic features of the Polish economy in the period 1989-2001 (see Table 5.1). Table 5.1: Dynamics of Basic Economic Categories, Poland, 1989-2001 (previous year=100) Categories 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Gross Domestic Product 100.2 88.4 93.0 102.6 103.8 105.2 106.5 106.1 106.9 105.6 104.1 103.8 101.0 Industrial production 99.5 75.5 92.0 102.8 106.4 112.1 109.4 108.7 110.8 104.6 104.8 107.2 100.3 Agricultural final production 111.0 99.7 106.8 87.7 108.0 89.2 116.3 99.1 101.8 105.2 96.0 96.8 108.6 Fixed capital formation 97.9 89.4 95.6 102.3 102.9 109.2 126.0 120.3 120.8 113.7 106.1 103.9 87.4 Consumption 104.9 84.3 107.5 103.5 104.8 103.3 107.2 108.7 106.1 104.3 102.8 101.7 Exportsa 100.2 113.7 97.6 97.4 98.9 118.3 116.7 109.7 113.7 104.2 97.0 115.5 114.0 Importsa 101.5 82.1 137.8 113.9 118.5 113.4 120.3 127.9 114.9 109.6 97.6 106.6 102.8 Foreign investment . . . 242.7 204.2 93.4 209.2 241.7 110.9 111.2127.8 128.1 84.3 Working population, total 99.0 97.3 94.1 95.8 97.6 101.1 100.3 101.9 101.2 163.3 98.4 97.2 96.8 Working pop., public sector . 93.1 85.4 89.1 90.9 95.1 95.7 96.2 93.7 99.8 92.8 92.6 93.3 Working pop., private sector . 102.1 104.8 102.4 103.4 105.5 103.3 107.1 105.3 92.4100.9 98.4 98.2 Unemployment rateb . 6.3 11.8 13.6 13.7 16.0 14.9 13.2 10.5 103.5 13.1 15.1 17.5 Inflationc 640.0 686.0 171.1 142.4 134.6 130.7 126.8 119.4 113.5 10.4107.3 110.1 105.4 Households' real incomes . . . 99.6 99.4 103.5 106.2 104.2 107.3 109.7102.0 101.4 101.3 104.4 aIn US dollars, current prices. At December 31. December to December. b c Sources: Statistical Yearbooks, Central Statistical Office, Warsaw, 1990-2002. · The period 1990 to mid-1992, when "shock-therapy" took its toll in the form of a recession which brought about a 15 percent decline in GDP (however, the inclusion of 104 the "gray economy," which developed rapidly, would lower the magnitude of the "official" decline), saw a dramatic increase in unemployment and inflation. · Mid-1992 to 1998, when the growth rate was constantly accelerating to reach 6-7 percent yearly. Unemployment ­ after reaching a peak in 1994 of 16 percent - declined to 10 percent, foreign capital investment went up to an inflow of US$10 billion yearly, and inflation declined, dropping to a one-digit figure. · After 1999, a breakdown can be seen in these positive trends. The rate of growth declined to a mere 1 percent in 2001 and 2002, to climb ­ according to different forecasts - to 2.5 - 3.5 percent in 2003, the latter figure being a somewhat unrealistic expectation on the part of the government. There was a crisis in public finances, with an increase in the deficit of the state budget (to over 5 percent of GDP, up from below 2 percent in 1998), unemployment grew up to 18 percent in 2003, and FDI declined (down to US$6-7 billion yearly). Inflation declined to 3.6 percent in 2001 and to below 1 percent in 2003 (some economists even voiced fears of deflation), and the Polish currency is constantly appreciating (with some fluctuations, however), which places a strain on the foreign trade and current account balance. 5.9 Table 5.2 presents the economic performance of Poland in comparison with other post- socialist European countries. 5.10 Figure 5.1 presents the growth/decline trajectories of particular post-socialist countries in the transition period. 5.11 It is striking how clearly the trajectories of the post-socialist transformation in particular countries comply with the concept of the J-curve trajectory (Bradshaw, Stenning, 2000) (see Figure 5.2). These patterns will be seen later in the discussions of the trajectories of regions under transformation. Table 5.2: Dynamics of GDP in Post-Socialist Countries, 1989-2001 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 1989= Countries Previous year = 100 100 Belarus 98 99 90 89 84 90 103 111 108 103 106 - 79a Bulgaria 91 88 93 99 102 102 89 93 104 102 105 104 71 Czech Rep. 99 86 94 99 103 106 104 100 98 99 103 103 93 Lithuania 93 87 61 70 101 103 104 106 104 97 104 106 48 Latvia 103 89 65 85 101 98 102 109 106 101 107 108 69 Poland 88 93 100 104 105 107 106 107 105 104 104 101 127 Russia 97 95 86 91 87 96 97 101 95 103 108 - 62a Romania 94 87 91 102 104 107 104 93 93 97 102 105 79 Slovakia 98 85 94 96 105 107 107 107 104 102 102 103 106 Slovenia 92 92 95 102 105 104 104 105 104 105 105 103 116 Ukraine 96 88 86 86 77 88 90 97 98 100 106 - 38a Hungary 97 88 97 99 103 102 101 105 105 105 105 104 110 aFor 2000. Sources: Various. 105 Figure 5.1: GDP Dynamics in Central and Eastern Europe, 1989 ­ 2001 Sources: William Davidson Institute; based on OECD, Economic Outlook, Vol. 69, July 2001, EBRD, Transition Report, 2001 Update, and Davidson Institute staff calculations. Sources: William Davidson Institute, based on Oecd, Economic Outlook, Vol. 69, July 2001, EBRD, Transition Report, 2001 Update, and Davidson Institute staff calculations. Figure 5.2: Different Restructuring Trajectories, J-curve Patterns q q: growth rate; t: time steep J-curve (similar to Polish case) shallow J-curve (similar to Czech, Hungarian cases) t delayed restructuring (similar to Russian, Ukrainian cases) Source: Based on Author's own computations. 5.12 The J-curve indicates that the restructuring process has to begin with a phase of scrapping or destroying old, obsolete production stock (this may include the institutions and the skills of the labor force) which cannot be effective under the new economic and technological conditions; it is only after this "scrapping" phase that the economic system can enter the path of growth. This pattern follows the idea of the Schumpeterian "creative destruction," which refers to the role of 106 the innovator in economic development. The systems (firms, cities, regions, countries) that are bolder during the "destruction" may be better rewarded in the "creation" or "growth" phase, since the burden of obsolete structures is much smaller and the freedom for creation is greater. 5.13 According to the concept of the J-curve, the recession that occurred in the post-socialist countries after 1990 was the price of restructuring and was similar to that which the West paid for its change in socioeconomic structures after 1973. The bolder the restructuring was, the better were the foundations for future growth. Slowing down the restructuring eventually leads to a slowing down in growth, which happened in Poland at the end of the 1990s and also in other Central European countries a little earlier. And delaying the restructuring, or attempting to avoid it altogether, does not lead anywhere: the "price" of a recession is still paid, but the reward does not come (or comes very late). 5.14 The post-socialist transformation should be demythologized from its ideological underpinnings (although this dimension should not be entirely overlooked) and should be regarded as a normal process of technological and organizational change, one that would have taken place earlier if the former socialist countries had been incorporated earlier into an open global economy. The initial decline and the later downswings in economic performance can be considered as the price of restructuring, similar to that which the West paid for its change in socioeconomic structures after 1973. The J-curve pattern, or the principle of "creative destruction," may just apply to the post-socialist transformation. 5.15 The current stagnation of the Polish economy (and not the recession, as some tend to call it) can be attributed to several factors: · The worsening of the international economic environment, with the special influence of the collapse of the Russian economy and financial system in 1998 and a slowdown in the economic growth of the European Union and the United States. · The slowdown in economic restructuring and privatization during the period of social-democratic-peasant rule (1993-97), which, after a three-year time-lag, resulted in the declining performance of the Polish economy. · The series of mistakes and the economic mismanagement of the Solidarity - Union for Freedom government in its term of 1997-2001, especially in the sphere of public finance, which led to the budgetary deficit of 5.3 percent of GDP. 5.16 The prospects for the next few years are rather bleak. The governmental strategy assumed the growth rate, according to the formula 1-3-5, to reach 5 percent in 2004. However, after just over a year from this announcement, it appeared that these assumptions cannot be reached. The unemployment rate will not be curtailed. The fiscal strain is still strong, and reductions in spending on the social sphere are necessary (in Poland about 14 percent of GDP is spent on pensions and disability allowances -- one of the highest levels in the world, if not the highest). The attempts at deep reform in public finance have a poor chance of success owing to the inability of the political elites to reach agreement on its shape. It now appears that the main barriers to accelerating growth lie in the institutional, financial, and political spheres. 107 C. TERRITORIAL PATTERNS IN POLAND3 Demographic Patterns 5.17 Figures 5.3, 5.4 and 5.5 present the age structure of the Polish population, by municipalities, in 2001. 5.18 The highest concentration of young population occurs in the northern and western territories, as well as in the southeastern territories. In the first case the explanation should be sought in the massive migrations just after the War, which were composed of younger people (as is usually the case with migrants). The high proportion of children in the southeast is the result of the rural demographic reproduction patterns ­ in rural families the number of children is usually higher. 5.19 The highest concentrations of productive age populations are found in urban-industrial regions, which have attracted migrants and in which the number of children is relatively lower. Figure 5.3: Population of Pre-productive Age (0-17) in Polish Municipalities, 2001 Population of pre-productive age, 2001, in percent 27,8 - 34,9 (482) 26,3 - 27,8 (509) 24,9 - 26,3 (491) 23,3 - 24,9 (482) 13,5 - 23,3 (523) Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. 3 The main problem in regional studies in Poland is the unavailability of lengthy time series for the territorial units, especially at the regional and sub-regional levels. The administrative division of Poland has changed twice over the last three decades from 17 to 49 units in 1975 and then from 49 to 16 units in 1999. Only a few variables and for a limited number of points in time, have been recalculated from the old 489 units to the new 16 voivodships. In 1999, 44 (later this number was enlarged to 45) sub-regional statistical territorial units (NUTS3) were created, with a limited set of data collected and published. For example, the GDP for the old 49 voivodships was calculated only for 1992 and then yearly for the period 1995-98, while for the new NUTS3 and the new 16 voivodships, it was calculated only for 1998 and 1999. Also, other time series were calculated for the old 49 units until 1998, and for the 16 new ones, from 1999. About 2,500 municipalities are the most stable territorial units, though their number and boundaries do change, and the scope of data is limited. This instability of the time series of territorial data (also aggravated by the changes in classifications and statistical definitions, also stemming from adapting the Polish statistical system to the EU standards) has led to limited opportunities for in-depth historical analyses of the territorial structure of Poland and its changes. 108 Figure 5.4: Population of Productive Age (men: 18-65, women: 18-60) in Polish Municipalities, 2001 Population of productive age, 2001, in percent 61.9 - 75.0 (522) 59..9 - 61..9 (478) 58..3 - 59.9 (474) 56.1 - 58..3 (497) 45.3 - 56.1 (516) Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. Figure 5.5: Population of Post-productive Age (men: 65+, women: 60+) in Polish Municipalities, 2001 Population of post-productive age, 2001, in percent 17.9-40.3 15.8-17,9 14,2-15,8 12.7-14,2 5.4-12.7 Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. 5.20 The older population has a higher share in the rural regions of central and eastern Poland, with the exception of the southeastern regions, where the population is relatively younger (because of the high birth rate, as indicated), and of the large urban centers, where the share of older people is lower. The highest shares occur in the mid-eastern localities. 109 Economic Territorial Differences 5.21 The level of economic development of the Polish regions is strongly differentiated. The spatial structure of GDP per inhabitant for 2000 in 44 NUTS3 (the smallest territorial units for which this economic category is calculated) is presented in Figure 5.6. 5.22 The ratio of the richest and poorest NUTS3 regions in Poland is equal to 5.4 to 1. The extreme regional values (in relation to the national average) are: 298 percent for the city of Warsaw and 57 percent for the Loma sub-region (northeast of Warsaw). 5.23 It is worth noting that several large urban agglomerations are surrounded by regions with relatively low levels of GDP per inhabitant. This is the case for the city of Warsaw (the ratio is equal to 3.4 to 1), Kraków (2.4 to 1), Wroclaw (2.3. to 1), and Pozna (2.2 to 1). It may be argued that these differences are a territorial-statistical phenomenon caused by "extracting" the city from its immediate surroundings. However, similar discrepancies can be found within larger regions ­ for example, in Mazowieckie Voivodship the difference in GDP per capita between Warsaw and the poorest sub-region with the main city of Radom (located 1,000 kilometers from Warsaw) is equal to 4.9 to 1. 5.24 While the regional differences in Poland are relatively large, it is doubtful that they are much greater than in some Western European countries of similar size, and they are not greater than in other post-socialist European countries. This is shown in Table 5.3. Figure 5.6: GDP per Capita, NUTS3, Poland =100, 2000 GDP per inhabitant 2000 r., NUTS3 Poland = 100 150.1 - 298.0 (5) 100.1 - 150.0 (5) 75.1 - 100.0 (20) 57.0 - 75.0 (14) Source: Produkt Krajowy Brutto wedlug woejwództw i podregionow w 2000 r., (Gross Domestic Product by Voivodships and Sub-regions in 2000), Katowice, 2001, table 16 110 Table 5.3: Regional Differences in GDP per Inhabitant in Selected European Countries, 1998 Countries Number of regions Highest : lowest ratio Poland 16 2.0 : 1 44 5.4 : 1 France (continental) 22 2.0 : 1 Germany 42 3.4 : 1 Italy 20 2.3 : 1 Spain 17 2.2 : 1 United Kingdom 37 3.3 : 1 Czech Republic 8 2.4 : 1 Hungary 7 2.2 : 1 Romania 8 1.9 : 1 Slovakia 4 2.5 : 1 Source: Second Report on Economic and Social Cohesion, European Commission, Brussels 2001, Table 50. 5.25 Fiscal and social policies soften the regional differences through personal redistribution. The regional differences in personal disposable incomes are smaller than the differences in primary incomes, which are similar to the differences in GDP per inhabitant (see Table 5.4). 5.26 The share of social benefits in gross disposable incomes plays the role of the regional "equalizer." The share of these social incomes in total household incomes (national average equals 19 percent) was highest in the poorest eastern regions (Lubelskie, 22 percent) and lowest in the richer regions (Mazowieckie with Warsaw, 14.8 percent). Table 5.4: GDP and Personal Incomes per Inhabitant, 1999 (extreme values for 16 voivodships, current prices, in zlotys) Voivodships with the Voivodships with the Ratio of Categories Poland highest values lowest values the extreme voivodship value voivodship value values Gross domestic product 17,725 Mazowieckie 26.871 Lubelskie 12,146 2.2:1 Gross primary incomes 13,306 Mazowieckie 21,057 Lubelskie 9,325 2.3:1 Gross disposable incomes 12,635 Mazowieckie 17,843 Podkarpackie 9,625 1.9:1 Source: Statistical Yearbook of the Voivodships 2001, Central Statistical Office, Warsaw, Table 5 (253). 5.27 The regional differences in the employment structure seem to be the most important factor differentiating the regional GDP levels. This is reflected in Table 5.5.4 Table 5.5: Gross Value Added per Employee, 1999 (current prices, in zlotys) Voivodship with the Voivodship with the Ratio of Categories Poland highest value lowest value extreme voivodship value voivodship value values Total 40,102 Mazowieckie 51,760 Podkarpackie 25,404 2.0:1 Agriculture, game, fishing 5,727 Zachodnio- 13,045 Podkarpackie 2,127 6.1:1 Industry 49,368 Pomorskie 64,400 Warmisko- 40,072 1.6:1 Market services 63,296 Mazowieckie 74,613 Mazurskie 53,713 1.5:1 Non-market services 34,170 Mazowieckie 40,308 Podkarpackie 30,799 1.3:1 Mazowieckie Lubelskie Source: Statistical Yearbook of the Voivodships 2001, Central Statistical Office, Warsaw, Table 4(252). 4The 16 new voivodships are taken here for analyzing the regional differentiation in Poland. 111 5.28 Labor productivity in agriculture and other primary sectors is about 6 times lower than the average productivity, over 7 times lower than productivity in industry and 10 times lower than productivity in market services. Therefore, in regions with high employment shares of low productivity sectors (agriculture) the overall level of value added must be lower than in regions with high shares of high productivity sectors (services). Low GDP values per inhabitant in the eastern and central Polish regions are the result of high shares in the primary sector in their employment structures: these shares are as follows: Lubelskie (Lublin) 51.9 percent; Switokrzyskie (Kielce) ­ 48.9 percent; Podkarpackie (Rzeszów) ­ 47.3 percent; Podlaskie (Bialystok) ­ 46.5 percent. In the metropolitan regions the shares of the first sector are low, and are equal to: 12.2 percent in lsk Voivodship (capital city, Katowice), 14.8 percent in Zachodnio-Pomorskie (Szczecin), 15.8 percent in Pomorskie (Gdask), and 24.9 percent in Mazowieckie (Warsaw). 5.29 In regions with high shares of the labor force in the low productivity sectors the GDP per inhabitant has to be lower than in regions with high shares of the sectors with high labor productivity. The structural factor is therefore largely responsible for the regional differentiation in Poland. A more detailed picture of the territorial differentiation of the occupational structures5 is presented in Figures 5.7, 5.8 and 5.9. Figure 5.7: Share of Working Population in Agriculture, Forestry, and Fishery in the Total Number of Working Population, 2000, by Powiats % 54 ,6 - 87,3 (132) 34 ,5 - 54,5 (101) 17 ,6 - 34,5 (66) 0 ,5 - 17,5 (74) Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. 5In the Polish statistical system, "employed" refers to those who are employed on a contractual basis. The privately occupied are not considered to be employed. "Working" is the broadest category, comprising all forms of occupation, whatever the legal basis. 112 Figure 5.8: Share of Working Population in Industry in the Total Number of Working Population, 2000, by Powiats % 39 ,3 - 79,5 (97) 29 ,3 - 39,2 (93) 19 ,3 - 29,2 (99) 8,1 - 19,2 (97) Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. Figure 5.9: Share of Working Population in Services in the Total Number of Working Population, 2000, by Powiats % 36,1 - 69,0 (64) 26,1 - 36,0 (118) 16,1 - 26,0 (88) 2,6 - 16,0 (103) Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. 5.30 As can be seen, in central and eastern Poland (former Russian "Kongresówka" and Austrian "Galicia") the shares of the working population in sector I are the highest. Regions with large urban centers demonstrate high shares in industry (especially Upper Silesia) and in market and non-market services (this last sector assumes high shares also in tourist regions). 5.31 The economic structures can be noted in all aspects of socioeconomic processes. For example, Polish academic potential is concentrated in a few big cities, which deliver most of the scientific output and which connect Polish science with the outside world (see Table 5.6). 5.32 A more detailed picture of territorial differentiation in Poland can be obtained through measuring the level of development of the local units (municipalities). Owing to the limitations 113 of data (calculating GDP for such small units would be inappropriate), one may use financial categories related to the local budgets. Figure 5.10 presents the quintiles of municipalities according to the value of own revenues and shares in state taxes (PIT and CIT) per inhabitant in 2001.6 Table 5.6: Shares of the Leading Four Academic Centers in Number of Publications and Citations in the Total Number Noted in the "Philadelphia List," in the Total Number of Publications and Citations of the Respective Type of Polish Academic Establishments, 1999 Universities Technical universities Medical universities academic publica- citations academic publica- citations academic publica- citations centers tions centers tions centers tions Warsaw 23.4 35.7 Warsaw 18.1 21.3 Kraków 15.8 19.7 Kraków 16.7 17.2 Wroclaw 15.6 14.4 Warsaw 14.8 16.7 Pozna 10.1 8.1 Kraków 15.1 19.0 Lód 12.1 11.6 Wroclaw 12.6 9.3 Lód 12.7 13.8 Pozna 9.9 9.9 Total 62.8 70.3 Total 61.5 68.4 Total 52.6 57.9 Source: Gorzelak and Olechnicka, 2003. 6This seems to be a good category of the economic situation of a locality. Own revenues are derived mostly from local taxes, of which the property tax on land and assets owned by companies is the most important part. Shares in the state taxes (PIT and CIT) are correlated with the "intensity" of economic output of the productive sector and the value of personal incomes. It should be remembered, though, that the agricultural population is not taxed by PIT, which lowers the values of this variable for the rural areas. 114 Figure 5.10: Quintiles of Polish Municipalities According to the Value of Own Revenues and Shares in State Taxes (PIT and CIT) in Zlotys per Inhabitant, 2001 6 5 3 - 3 2 0 8 8 ( 5 0 0 ) 4 9 1 - 6 5 3 ( 4 9 3 ) 3 9 3 - 4 9 1 ( 4 9 0 ) 3 0 8 - 3 9 3 ( 4 9 9 ) 1 4 9 - 3 0 8 ( 4 9 6 ) Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. 5.33 Only two extreme quintiles (the top and the bottom) are presented in Figure 5.11. 115 Figure 5.11: Lowest (blue) and Highest (red) Quintiles of Polish Municipalities According to the Value of Own Revenues and Shares in State Taxes (PIT and CIT)per Inhabitant, 2001 "Rich" and "poor" municipalities, 2001 Top 20 percent Bottom 20 percent Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. 5.34 The spatial distribution of the poorest municipalities shows that they are located almost entirely in the former Russian and Austrian parts of Poland under the partitions. The 1815-1918 boundaries are clearly visible. Almost no localities from the Prussian part of Poland and the western and northern territories fall into this group. Rich municipalities are scattered around the entire country, but they are more densely distributed in its western regions. This shows that, on the one hand, belonging to an opulent region ensures "immunity from poverty," while, on the other hand, even in a poor region, thanks to external factors (tourist potential; border location; large industrial plant; vicinity of a big city) and/or to endogenous factors (local initiative and efficiency of action), a relatively high level of development is possible to attain. 5.35 The "red" spots in the figure that denote the richest municipalities clearly mark the regions around the big cities: Warsaw, Lód, the Upper Silesian conurbation, Pozna, Wroclaw, and to a lesser degree, Kraków and the Tri-City of Gdask-Gdynia-Sopot. This reflects the positive impact of a big city on its immediate surroundings. Suburbanization is definitely a factor responsible for this, as is the practice of commuting to work in the city and also the business development in its vicinity. No such strong influence can be observed in the case of the eastern big cities of Bialystok, Lublin, and Rzeszów, which shows the relative weakness of these cites and the localities surrounding them. 5.36 One particular concentration of the poorest municipalities (the belt along the eastern side of the former interstate boundary of Prussia and Russia) is striking. These localities were the most peripheral in the Russian Empire in the nineteenth century: they were hampered by a location along an insulated border, which decreased their development opportunities. Although this border lasted for only 103 years and ceased to exist 85 years ago, its negative influence is still 116 visible. This may be considered as a clear proof of the strength of historical factors in the present structure of Polish socioeconomic space. 5.37 Figure 5.12 demonstrates the density of companies with foreign capital in Polish municipalities. Figure 5.12: Density of Companies with Foreign Capital in Polish Municipalities, 2001 Companies with foreign capital 501 - 9 830 (11) 101 - 500 (41) 21 - 100 (199) 6 - 20 (441) 2 - 5 (618) 1 (426) 0 (751) Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. 5.38 The similarity of the distribution of companies with foreign capital to the distribution of "non-poor" municipalities is striking. It can be seen that one of the most important phenomena representing the transformation process ­ the inflow of foreign capital ­ clearly follows the spatial pattern of wealth and poverty in Polish localities, which is the product of a process much more deeply rooted in history. 5.39 It is also easy to note another manifestation of historical factors in contemporary Polish spatial patterns. Figure 5.13 presents the geography of the voting pattern for T. Mazowiecki (the fist prime minister in democratic Poland, who introduced the reforms in Polish economy) in the first round of the first democratic presidential election in November 1990 (i.e., when most of the hardships and almost no benefits of the so-called Balcerowicz plan could be seen). 5.40 If voting for T. Mazowiecki in the fall of 1990 could be considered an expression of support for the course of reforms and of belief in their success, then the spatial pattern of this voting can be treated as a representation of pro-reform attitudes, and of trust in the adopted course of post-socialist transformation. 5.41 Other research (Gorzelak et al, 1999) has demonstrated that the activities and innovations of local authorities have also been distributed following the same patterns (i.e., the highest in the western and northern territories, followed by Wielkopolska (with its capital in Pozna) and Silesia, and (after a big gap) by the former Austrian, Galicia, and Russian ­ Kongresówka parts of Poland). 117 Figure 5.13: Shares of Votes Cast for T. Mazowiecki in the First Round of Presidential Elections, November 1990 31,0 - 37,8 25,1 - 31,0 19,8 - 25,1 15,6 - 19,8 12,6 - 15,6 9,9 - 12,6 7,3 - 9,9 5,1 - 7,3 3,3 - 5,1 1,8 - 3,3 0,0 - 1,8 brak danych Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. The Labor Market 5.42 The beginning of the transformation process introduced a new phenomenon ­ unemployment (which had not openly existed in the "socialist" reality). In 1994, unemployment approached almost 3 million people, to drop down to under 2 million in 1996 and to exceed 3 million in 2002 (in mid-2002 this number reached 3.3 million). Figure 5.14 presents the spatial distribution of unemployment in Poland at the end of 2002, by powiats.7 5.43 In its regional patterns of unemployment, Poland differs from "typical" cases, in which the poorer and more peripheral a region is, the higher its unemployment is. The Polish northern and western regions are relatively better developed than the eastern and central regions, but they have the highest unemployment in the country. Unemployment in the west and north has been caused by the collapse of the state-owned agricultural farms, which were concentrated in these areas.8 This is the most difficult situation in the labor market, since the unemployed (the former workers on the state agricultural farms) are poorly educated and have a low motivation to search for work: in addition, the job opportunities are extremely low in these areas. However, as observations in the field indicate, there are several sources of income for the unemployed. Working abroad is the most common source, and there are several directions of migrations (see Box 5.1). 7The powiat is a self-governmental unit between the regional (voivodship) and local (municipality) levels. There are 315 "rural" powiats and 65 "urban" powiats. This latter case denotes the bigger cities (over 10,000 in population, in principle) which have at the same time both the status of an ordinary municipality and the status of a powiat. 8 This structure has a clear historical underpinning. In 1945 the northern and western territories were regained by Poland from Germany, and after the relocation of the German population they were filled by the migrants form the eastern territories which were lost by Poland to the Soviet Union. State agricultural farms were then created on the large landholdings of the German junkers, and the private farms of the former German bauer were taken over by farmers. 118 Figure 5.14: Unemployment Rates (registered unemployment) in Powiats, December 2002 27.5 - 40.6 (76) 21.9 - 27.5 (76) 17.7 - 21,9 (76) 14.5 - 17.7 (76) 6.3 - 14.5 (76) Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. Box 5.1: Specialization of Temporary Migrations There are specific territorial "specializations" of temporary migrations from Poland. For example, in the small town of Siemiatycze in northeastern Poland there are special minibus connections to Brussels, which is a popular market for such jobs as office cleaning. Women from some southeastern localities (such as Jaroslaw) migrate to Italy, where they work as nurses or take care of the elderly. The western regions feed the German cleaning (women) and construction (men) labor markets. Traditional migration to the United States is still popular in the rural regions of southern and central-eastern Poland. 5.44 As a second source, part-time jobs, usually in the gray economy, give some income to families in which no one would work officially. Third, a return to very traditional sources of income, such as selling mushrooms and berries picked up in the forests, brings in some money, especially in the summer and fall. Social allowances also contribute (see Box 5.2). Box 5.2: "Soft" Benefits and Motivation for Work It is often argued that the "soft" system of unemployment benefits has negatively influenced attitudes toward work in Poland. As observed during several trips to the northern and western territories, work has often ceased to be one of the most important goals in life, and has become just a source of eligibility for unemployment benefits. Motivation for work has dramatically decreased, and it is common knowledge that it is often impossible to find employees for even simple jobs in localities with massive unemployment. 119 5.45 Finally, theft and smuggling are frequent sources of income, especially in border locations. 5.46 The second cause of unemployment ­ the industrial restructuring -- has manifested itself in some localities in the southeastern and central parts of the country, mostly in the former Central Industrial Region (Centralny Okrg Przemyslowy, COP), which was established in the 1930s in order to develop the industrial-military complex of pre-War Poland. The industrial plants were constructed in small and medium-size towns, and relied on workers who commuted from nearby villages that were suffering from massive agricultural overmanning. The COP has been further developed in post-War Poland, and also relies on labor commuting from overpopulated rural areas. The collapse of the Warsaw Pact and the industrial restructuring have stricken these industrial plants severely, which has led to the dramatic collapse of the towns that hosted them (with some exaggeration, these towns could have been compared to the "company towns" so well known in some industrial regions of the United States). Unemployment has soared and has assumed two forms: open (registered) unemployment in the towns, and hidden unemployment in the rural areas of southeastern and central Poland. It is estimated that in southeastern Poland up to a million persons who derive some income from agriculture are in fact partially or entirely unemployed, which adds (as hidden unemployment) to the overall figures of the jobless. 5.47 Unemployment is low in two types of regions: in metropolitan regions, with diversified economic structures and high rates of growth (such as Warsaw, Poznan, Wroclaw and their surroundings), and in their opposites: the relatively less densely populated regions with high shares of agriculture, where farming provides some shelter against the tight labor market. However, although the registered unemployment is low, the real unemployment is much higher, owing to the massive phenomenon of partial employment. According to some estimates, there are up to 900,000 redundant persons of productive age in the southeastern region of Poland. 5.48 In addition, the industrial region of Upper Silesia has not yet reached the national values of unemployment rates, thanks to the delayed ­ postponed ­ restructuring of its coal-and-steel complex. However, in some parts of the region unemployment is soaring and the signs of urban pathologies can be observed. D. REGIONAL DYNAMICS IN THE TRANSFORMATION PERIOD Demography and the Settlement System 5.49 The structure of the settlement system displays a tendency towards stabilization. During the last decades we can observe a constant decline in domestic migratory flows. The shift from rural areas to urban areas has been steadily declining, to turn into a net outflow from towns in 2000 (see Table 5.7). 5.50 This process has been paralleled by changes in the internal structure of the metropolitan areas (see Figure 5.15). 5.51 With one exception ­ that of the city of Lód and its hinterland ­ we observe a shift from the core to its immediate surroundings (the ring). The core and the ring together, as well as the entire metropolitan region, gain in population. Lód is the only metropolitan region that experienced losses in both the central city and its regions, which reflects the negative effects of the industrial restructuring in this city and, perhaps, the dominance of the industrial socio- occupational structure, and the relative under representation of the upper strata. 120 Table 5.7: Domestic Migrations in Poland, 1971-2001 1971- 1981- 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Flows 80 90 in thousand persons Towns inflow 5460.7 4196.5 346.0 331.2 314.1 281.1 258.9 239.3 243.5 233.5 236.9 241.4 221.3 210.2 outflow 3455.3 2852.5 233.3 224.8 228.4 221.2 219.9 212.4 220.2 217.7 228.3 238.6 225.5 216.7 Rural areas inflow 3470.3 2513.9 183.9 174.2 180.0 175.7 179.0 180.4 183.8 183.5 188.9 191.0 172.8 159.1 outflow 5477.7 3857.9 296.6 280.6 265.7 235.6 218.0 207.3 207.1 199.3 197.5 193.8 168.6 152.6 Balance, towns 2005.4 1344 112.7 106.4 85.7 59.0 39.0 26.9 23.3 15.8 8.6 2.8 - 4.2 -6.5 Source: Statistical Yearbooks, Central Statistical Office, Warsaw. Figure 5.15: Migration Balance in Selected Polish Metropolitan Regions, 1994-1999 (in percent) 8.00% 6.00% 4.00% Core Ring 2.00% Area 0.00% Region Warszawa Kraków Wroclaw Pozna Trójmiasto Lód -2.00% -4.00% Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. Regional Dynamics 5.52 The economic regional differences have been growing during the last decade. Table 5.8 presents GDP statistics for selected Polish old voivodships in the period 1992-98. 5.53 The regional differentiation of GDP per inhabitant is growing rapidly. While in 1992 the richest region ­ the former Warsaw Voivodship - reached 158 percent of the national level of GDP per capita, in 1998 this figure jumped to 214 percent. At the opposite end of the scale we observe a relative decline for the poorest regions, located mostly in the eastern part of Poland. The ratio of the extreme values increased from 1 : 2.4 to 1 : 3.7, by more than half (the growth of the national economy during the period analyzed equaled 39 percent). 5.54 The regions in Table 5.8 are arranged to form coherent groups. The first group of regions is composed of the regions with large urban centers, well equipped in terms of material and institutional infrastructure, with a well educated, skilled and motivated labor force, and with a rich economic structure. In most of these regions we can observe an increase in their relative 121 position in relation to the national average: this was the case in Warsaw, Pozna (mid-west), Kraków (mid-south), and Wroclaw (southwest). The two agglomerations with sea harbors and a ship-building tradition (the Tri-city of Gdask-Gdynia-Sopot and Szczecin) have not performed well, and the western Szczecin region even noted a decline ­ which is surprising, since this region could be expected to take advantage of its border location. Table 5.8: Regional Differentiation of GDP per Inhabitant in Selected Regions of Poland=100 Regions 1992a 1995b 1998b Percent increase, 1992-1998 Warsaw 158 168 214 88 Gdask 105 107 107 41 Kraków 117 112 120 42 Pozna 132 112 134 41 Szczecin 127 118 111 21 Wroclaw 105 118 116 53 Gorzów 86 90 86 39 Zielona Góra 105 95 87 15 Opole 100 101 90 35 Legnica 126 121 116 28 Piotrków 120 94 86 - 1 Plock 149 132 114 6 Katowice 116 127 112 34 Lód 105 105 103 36 Walbrzych 80 78 73 26 Biala Podlaska 79 67 63 11 Chelm 87 64 65 4 Ostrolka 70 71 67 33 Radom 78 75 72 28 Slupsk 77 73 74 33 Suwalki 65 64 65 39 Zamo 73 64 58 10 aGDP in factors cost prices Gross value added. b Sources: Gross Domestic Product and Incomes by Voivodships in 1992. Part I: Methodology and Results of Studies, GUS, Warsaw 1994, Table 20; Gross Domestic Product by Voivodships in Years 1995-1998, GUS, Katowice 2000, Table 9. 5.55 A similar unexpected situation has occurred in the two mid-western regions (Gorzów and Zielona Góra), which declined in relation to the national average in the period 1992-98. There has been a general belief that the exposure to Western (mostly German) influence, in the form of consumer demand and direct investment, as well as the inflow of funds from the EU, would accelerate the growth of these regions (this thesis was also presented in my own writing). It now appears that this has been true in only a few localities: namely, those that had a border crossing or that were located very close to the border ­ to which the consumer demand "poured in" from Germany, especially in the first half of the 1990s. As deeper insight into the dynamics of particular economic sectors reveals, the growth of the market and non-market services in these regions was especially slow, which contributed to a worse than average economic performance (see Box 5.3). 122 Box 5.3: Trade and Border Crossings As field observations indicate, the trade with German consumers has been conducted not only by citizens of the border regions, but in many cases by those who came to this region from other parts of the country. In addition, the products sold ­ mostly food, apparel, household utilities, furniture, decorative goods, etc. ­ have been produced in all of Poland. Thus, the profits were transferred out of the region, and those that left the region were spent in great proportion on new housing, luxury goods, and the pleasures of life (travel, entertainment, etc.). The opportunity given to the region through the opening of the border with a richer neighbor has not been fully utilized to create a sound economic basis that could enable the local firms to compete in the demanding German markets. 5.56 The next group comprises the regions dominated by the raw-materials based industries: brown coal and electricity in Piotrków, copper in Legnica, and oil refineries in Plock. All of these regions declined in relative terms, which shows the changing patterns of Polish economic structures. 5.57 The next group of regions in Table 5.8 is composed of the old industrial regions, which had (or should have) undergone deep industrial restructuring. This was the case with Walbrzych (southwest), where the coal mines were closed down due to the extraction of the available coal deposits. The decline in the relative value of GDP per inhabitant confirms this process. Lód (center, the second largest city in Poland) displays a mixed picture: on the one hand, its traditional industry, textiles, experienced a deep decline in the first phase of transformation, recovering partly in the second; on the other hand, the relatively differentiated economic structures and personal flexibility of its inhabitants enabled the region to develop new economic activities. The case of Upper Silesia is more complicated: the growth in GDP per inhabitant in relation to the national average in the first period may be attributed to the delayed restructuring and the mounting debts of coal mining and heavy industry (steel, chemistry, machinery). The partial rationalization of these industries, which took place in the second phase, is reflected in the rapid decline in its position. It is very likely that this process will take place in the future as well. 5.58 The last group of regions is composed of the three types of regions: the external, eastern periphery (Biala Podlaska, Chelm, Ostrolka, Zamo), the internal periphery (Radom, undergoing industrial restructuring as well) and the regions affected by the collapse of the former state agriculture (Slupsk, in the mid-north, and Suwalki in the northeast). All of these regions but one (Slupsk, where the tourist sector adds to economic performance) have noted steeper or milder decline, which was most rapid in the east. Eastern Poland is thus a typical marginalized, depressed region, poorly equipped with all of the most important factors for development (infrastructure, skills and education, institutions, R&D potential, organizational experience, social capital), and also remotely located in relation to the centers of capital and innovation. 5.59 There are two possible patterns for the growth of regional differences: · The rich regions will develop more rapidly than the poor ones, but all of them will note a positive rate of growth. · The rich regions will grow while the poor ones decline. 5.60 The first situation can be considered acceptable, since this seems a natural process, especially in relatively less developed countries which try to achieve accelerated growth and to 123 carry out deep structural change. The second situation might be regarded as a symptom of deficient development, and definitely calls for deeper reflection. At the end of the 1990s, unfortunately, we have observed the second situation in Poland. Several poorly developed regions on the external eastern and internal (central) peripheries noted an absolute decline in terms of real GDP.9 This is shown in the three illustrations in Figure 5.16, which represent the dynamics of GDP growth in three periods: 1999/1998, 2000/1999, and 2000/1998. Figure 5.16: Growth of GDP in NUTS 3, in Percent, Constant Prices (national deflator) (a) 1999/1998 (b) 2000/1999 (c) 2000/1998 a b c over 112,0 (10) 108,1 to 112,0 (13) 104,1 to 108,0 (9) 100,0 to 104,0 (5) 96,1 to 100,0 (4) under 96,0 (3) Source: Produkt Krajowy Brutto wedlug województw i podregionow, 1999, 2000 (Gross Domestic Product by Voivodships and Sub-regions 1999, 2000), Katowice, 1999-2001 5.61 In 1999, 15 sub-regions (one-third of the total number) noted a real decline in their GDP. These sub-regions were the eastern and central sub-regions plus two industrial sub-regions of Legnica (copper) and Opole (mid-south, machinery and heavy industry) and one northern sub- region (Slupsk). 5.62 In 2000 this picture was partly reversed: some regions that had large cities grew at a slower pace, while some rural, peripheral regions achieved a growth rate greater than the national average (in both of these years the national economy grew by 4 percent). However, there were also poor regions which noted a further decline in their GDP. 5.63 The slowing down of growth in regions with large urban centers could be interpreted as a first sign of the coming recession, which appeared more rapidly in the metropolitan centers than in the peripheral, underdeveloped rural regions. 5.64 This picture is fully consistent with the changes of in GDP per inhabitant among the "old" 49 voivodships, shown in Table 5.8. In general, the eastern and central peripheries are 9Since the price deflators have not been calculated for the territorial units, a general deflator of 1.067 for the whole of the GDP was applied. This deteriorates the picture for particular regions, especially those regions which have high shares of agricultural production, since prices for agricultural production in fact declined by 2.3 percent. Therefore, the real growth in these regions has been greater than indicated in the table, but this difference could not have been too large, since even in the regions with a high share of agriculture it provides only from 10 to 17 percent of GDP. 124 losing distance to the rest of the country, but now we have proof that this is due to their absolute decline and not to just a relative decline. 5.65 Deeper insight into the dynamics of the most disadvantaged region ­ the Polish "Eastern Wall" during the 1990s ­ sheds some light on the nature of this polarization process. This region has been following a "shallow" J-curve, as opposed to the metropolitan regions (with the exception of Upper Silesia) which assumed the "acute" shape of the J-curve. In the period 1986- 90, when the transformation shock of 1990 took place, and the national economy dived by some 10 percent, the eastern (less developed, peripheral) Polish regions relatively improved their economic situation, and in some cases this improvement even had an absolute dimension (i.e., the growth of their GDP). This stemmed from the smaller transformational recession in these regions, which was due to the higher share of private agriculture in their economic structures. This sector "cushioned" the recession of the initial years of the Polish transformation which was mainly driven by industrial restructuring. In the years 1991-92 (i.e., when the economic decline reached bottom, the trends in the first stage of transformation were reinforced. Almost the entire Polish "Eastern Wall" went through the recession of the first transformation stage smoothly, with its GDP dynamics exceeding the national values. During the recovery phase, however, the Eastern Wall proved less capable of accelerating its development. Apart from a very few exceptions (e.g., areas that experienced a revival of tourism and those with major urban centers) all of the eastern voivodships recorded lower GDP dynamics than the national average. This continued during the next five years, until 2000 (the last year with available statistics on regional GDP). 5.66 This pattern can be interpreted in the following way: the relatively smoother performance of the eastern regions in the decline period can be attributed to their agriculture-dominated economic structures, which were not "restructurable" ­ and which were even reinforced by the industrial decline in these regions. This did not create sound conditions for the growth of these regions in the recovery phase. As a result, the agricultural, peripheral regions, which are poorly equipped with infrastructure, and which have a low educational level and obsolete skills, do not demonstrate a strong potential for development in an open economy, exposed to the current paradigm of innovation-driven growth. 5.67 As before, more detailed insight into the regional dynamics is possible through analyzing the municipal level. Figure 5.17 presents the three groups of poor municipalities: those that were among the poorest 20 percent in 1998 only, those that were in this group in 2001 only; and those that belonged to the lowest quintile in both years. 125 Figure 5.17: Changes in the Lowest Quartiles of the Arrangement of Municipalities, on the Scales of Own Revenues and Shares in PIT and CIT per Inhabitant, 1998-2001 present in 1998 and 2001 (stable) present only in 1998 (advanced) present only in 2001 (declined) Source: Based on Bank of Local Data, Central Statistical Office, Warsaw 5.68 As can be seen, some of municipalities to the north of Warsaw were able to improve their relative position. The "backwash effect," represented by the consistent belt of poor localities around Warsaw and Lód, which was clearly visible in 1998, thus became somewhat less dense, which could perhaps be attributed to a strengthening of the "spread effects" of these two big cities. 5.69 Figure 5.18 presents the depth of change in the local budgets (nominal values) among the poorest municipalities. 126 Figure 5.18: Dynamics of Own Revenues and Shares in State Taxes in Relation to the Average Change in the Whole Country among the Poorest Municipalities, 1998-2001 (in nominal terms) growth 20-65 percent (better than average) growth 0-20 percent (average) decline 0-65 percent (worse than average) Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. 5.70 It can be seen that the municipalities around Warsaw have advanced more than the "average" poor municipalities, which has also been the case with some southern localities. At the other pole we find poor municipalities in the east, located far from large urban centers, that worsened their relative position in the period 1998-2001. 5.71 Another way of analyzing the dynamic on the local level is to analyze the changes in unemployment. Figure 5.19 presents this change in powiats in the period 1998-2002. 5.72 A deeper insight into the relationship between the initial rate of unemployment and its changes in the period 1998-2002 (in all powiats the unemployment rates grew) is provided in Figure 5.20. 5.73 The powiats are arranged along two axes: x = rate of unemployment in 1998 in percent; y = change in unemployment rates for 1998-2002, in percentage points. On each axis half of the standard deviation was marked on each side of the axis, which left, in the middle of the diagram, the units with unemployment rates and changes close to the national averages (within the span of one standard deviation). The cases which differ more than plus/minus half of the standard deviation on each axis can be arranged in four groups: · Success: relatively low unemployment in 1998, relatively low increase · Stabilization: relatively high unemployment in 1998, relatively low increase · Failure: relatively low unemployment in 1998, relatively high increase · Catastrophe: relatively high unemployment in 1998, relatively high increase. 127 Figure 5.19: Increase in Unemployment Rate 1998-2002, in Percentage Points (in gray ­ powiats not existing in 1998) 11,1 - 20,1 (80) 8,8 - 11,1 (72) 6,9 - 8,8 (72) 5,2 - 6,9 (71) 1,2 - 5,2 (78) no data (7) Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. 128 Figure 5.20: Relation of Rate of Unemployment in 1998 and its Growth in the Period 1998-2002 20 18 16 "failure" "catastrophe" 12.1998-12.2002 14 % 12 change 10 rate 8 6 4 Unemployment "success" "stabilization" 2 0 5 10 15 20 25 Unemployment rate % 12.1998 Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. 5.74 These four types are indicated in Figure 5.21. Figure 5.21: Typology of Powiats According to the Relation of Rate of Unemployment in 1998 and its Growth, 1998-2002 "failure" (28) "catastrophe" (50) "stabilization" (20) "success" (58) Source: Based on Bank of Local Data, Central Statistical Office, Warsaw. 129 5.75 As can be seen, unemployment grew especially rapidly in the regions already severely hit by a shortage of jobs ­ especially the western and northern regions, with high unemployment caused by the collapse of the former state farms. One of the most serious problem regions ­ that of Radom (south of Warsaw) also belongs to this group, whose problems are caused by the industrial decline and weakness of other sectors. In the eastern and central regions which are agricultural, with little industry and services, unemployment (registered) was lower and increased slowly. 5.76 Among large cities, only those which are suffering from a (delayed) decline in maritime- related industries (Gdask, Gdynia and Szczecin) have fallen into the group of "failures" (i.e., districts with relatively low initial unemployment rates and rapid growth). Other large cities (Warsaw, Pozna, Wroclaw, and even Lód which underwent early rapid industrial decline at the beginning of the 1990s) were able to join the group of success cases or to remain among the average cases. 5.77 As was observed during the analysis of the changes in the local budget, it is evident that the Warsaw region, especially its southern and western parts, can maintain its strong position and its relatively healthy metropolitan labor market. E. CONCLUSIONS: THE POSSIBLE FUTURE DEVELOPMENTS 5.78 At this point it is useful to recapitulate some of the major observations made in this chapter: 1. In spite of deep structural changes in the Polish economy and in Polish society, the regional structure of the country has not undergone deep changes during the transformation process. On the contrary, it has become petrified, since the polarization processes have dominated. 2. There are two dimensions to these processes: (i) the metropolitan ­ non-metropolitan divide, and (ii) the east-west divide. · The economies of the large urban centers (Warsaw, the Tri-city of Gdask- Gdynia-Sopot, Kraków, Pozna, Wroclaw) grew either little more rapidly than the rest of the country, or grew at the same rate as the rest of the country. The big cities have a locational and business advantage, which also manifests itself in a higher standard of living. The metropolitan ­ non-metropolitan divide has replaced the traditional urban-rural divide that was the spatial pattern of the industrial paradigm. This can be explained by the fact that the metropolitan regions were best adapted to benefit from the new economic reality, in which services and modern types of industries develop at the fastest pace, and that only these metropolitan regions are able to benefit from the incorporation of the Polish economy into the global economic structures. The non-metropolitan areas are still dominated by agriculture or industry, which has resulted in their growing slowly, or even declining. · The east-west divide stems from the lower ability of the eastern Polish regions to respond to the new paradigm of development and from their relative peripheral situation in the entire Continent's socioeconomic structure. This has long been the spatial pattern of Poland, since modernization has constantly come from the West (see Gorzelak, Jalowiecki, 2002). 130 3. There are obviously some changes in the regional structure of the country, but they do not strongly influence the general picture. The most visible changes are the following: · The decline of the former state farms and the emergence of persistent unemployment in the peripheral rural areas that are distant from any urban or industrial centers, with, at the moment, a "non-employable" labor force and the resultant self-perpetuating social marginalization and misery. · The restructuring, through decline and closure, of some raw material concentrations (Walbrzych ­ coal; Tarnobrzeg ­ sulphur) and the relative decline (compared to the pace of growth of the national economy) of raw material concentration (Legnica ­copper; Piotrków ­ brown coal; Katowice ­ coal, steel). · The advancement (although not as rapidly as expected) of the western regions, which have gained from the proximity of German consumer and investment capital and have enjoyed an inflow of EU funds within the Phare Crossborder Cooperation Program (about 50 million yearly). 4. The slow pace of changes in regional patterns can be attributed mainly to the stability of the regional dispersion of population, with its virtual lack of changes in occupational structure and with the stability of the relative differentiation of labor productivity in particular sectors. This can be expressed in the following formula: low domestic migrations ­ persistent high occupation in agriculture in less developed regions ­ low productivity in agriculture ­ low level of development in agricultural regions. 5.79 Rapid alterations in these patterns are unlikely unless some changes occur in the general structure of the European Continent. The discussion below substantiates this thesis. 5.80 As indicated elsewhere (Gorzelak, 1998; Gorzelak, Jalowiecki, 2002), the regional spatial patterns for Poland are very stable over time. It is possible to trace the basic east-west divide to the twelfth to thirteenth centuries. This was also the time when the basic urban structure of Poland was shaped, and the industrialization processes have not changed this structure substantially. If this observation is correct, then we should assume a very modest attitude toward attempts to introduce "corrections" to this historically created structure (i.e., to try to accelerate the development of the less developed eastern regions). 5.81 There are two basic questions which should be answered before the goals, means and instruments of regional policies aimed at equalizing the levels of regional development in a country like Poland are formulated and applied (if we consider the "regional policy" as a part of developmental policy, and not as a part of social policy). These questions are the following: 1. What is possible at all, and under what conditions? 2. Out of the possible solutions, which ones are feasible with the limited disposable resources? 5.82 The broad comparative and historical experience (see, for example, Landes, 1998) leads to the following answer to the first question: the underdeveloped regions usually have two kinds of disadvantages: they are peripherally located in relation to the centers of capital, decisions, and 131 innovation, and their internal features do not allow for successful endogenous development.10 At the same time their location makes them unattractive to external investors (for example, why should a major automobile manufacturer want to locate its assembly plant in eastern Poland, far from its supply sources, subject to an unreliable transport infrastructure, in a region that lacks business-supportive institutions, and lacks a well educated and skilled labor force, and that offers only poor living conditions for the managerial staff?) Any public external support aimed at changing this unfavorable situation would have to fail (as has been proved, for example, in such areas as the Mezzogiorno, Middle Appalachia, the Spanish-Portuguese border regions, or, recently, eastern Germany), since this support would not be able to spur economic growth and would have limited social meaning. Therefore, to some extent these regions are "sentenced" to relative underdevelopment, and thus polarization between them and the more advanced regions ­ (which could develop more rapidly) ­ would have to take place. 5.83 This might sound like pessimistic determinism, which, in fact, it is, ­ but only to some extent. We have also witnessed several cases that have broken out of this vicious circle. Examples exist in certain parts of the south in the United States, in Ireland, and in some western regions of post-socialist European countries (this list is unfortunately short). What happened in these regions was that a shift in macro-locational preferences took place, a shift that was independent of the activities of these regions themselves, but was very beneficial to them. Moreover, these regions were the ones that took advantage of these changes, and that anticipated them well in advance and prepared a rich "soil" to accommodate these changes. These were also the regions that used the used external public assistance that was made available to them to accelerate the necessary restructuring in order to be ready for investment from the outside and in order to adapt their own endogenous potential to the new patterns of development. 5.84 To illustrate this line of reasoning, let us take the case of the Polish "Eastern Wall," which will soon become the extreme eastern region of the enlarged EU. This region has only one opportunity for development. This would materialize if the former Soviet republics should become a dynamic market, of interest to global producers (European, American, Asian), that would want to use the locational advantage of that part of the EU closest to their new customers. Eastern Poland would become a focus of their locational decisions, since it would still be offering cheaper labor and land than that found in the core regions, and would offer all of the guarantees stemming from EU membership. This is only a possibility, however, and certain conditions would have to be met if this chance is to become a reality: The Eastern Wall has to be accessible (Poland has to build highways and reliable railways, as well as airports), the labor force must be both skilled and motivated (the educational and training system must be improved), business- supporting institutions must be available, local governments must understand the requirements of investors and must be helpful, and incentive packages (in kind, not in cash) must be prepared, etc. 5.85 This leads to the second question regarding regional policy: which activities should be undertaken in order to accelerate the development of the Polish eastern regions? The simple answer is: those activities which would first of all increase the chances of economic advancement for the post-Soviet republics. (Here, Poland may have only a limited influence, but it would definitely be greater than none.) Then it is necessary to take steps (nationally, regionally, and locally) to be ready to deal with these changes. This should also become Poland's message to the EU when that organization is shaping its structural policies for the next period, after enlargement. This would also serve as a message to all regions that are awaiting their 10It should not be forgotten that the country now operates in a competitive global economy driven by innovations (also called "technocapitalism"), and that it is more difficult for lagging regions to catch up than it was with the resource- driven paradigm. 132 chance for development--they should facilitate its emergence and should be ready to deal with it when it becomes a reality. 5.86 Overall regional policy in Poland should be conducted along the following four axes: education, science, infrastructure, and environment. Such an arrangement should allow for breaking through the almost unavoidable problem of regional policy: the equity ­ efficiency dilemma. "Education" is the goal that equalizes the chance for development and helps in supporting the less-educated regions. "Science," on the contrary, concentrates the resources of the regional ;policies in the better developed regions, which present the best potential for research and innovation. However, enhancing interregional networks of scientific cooperation should be an important objective of science policy, as it could lead to overcoming the split into "central" and "peripheral" science. Stress on infrastructure could remove one of the greatest barriers to development in Poland ­ the poor quality of roads, railways, urban transport systems, and (the relatively less underdeveloped) telecommunications networks. Finally, the "environment" denotes the need for improvement in several areas that are heavily polluted (although the situation improved considerably in this respect after 1990). 5.87 The accession to the EU is an important opportunity historically for Poland. The government's regional policy (among other policies) will be responsible for the efficient use of the resources that will come to Poland from the EU. Regions and localities might become important partners in this process, which calls for the introduction of the third step in decentralization reform: decentralization of the public financial system. 133 REFERENCES Bradshaw, M., and A. Stenning, (2000), "The Progress of Transition in East Central Europe," in J. Bachtler, R. Downes and G. Gorzelak, eds., Transition, Cohesion and Regional Policy in Central and Eastern Europe, Ashgate, Aldershot-Burlington-Singapore-Sydney, 2000. Gorzelak, G. (1998), The Regional and Local Potential for Transformation in Poland, EUROREG, Warsaw. Gorzelak G., and B. Jalowiecki (2002), "European Boundaries ­ Unity or Division of the Continent?, Regional Studies" Vol. 36.4. Gorzelak G., and A. Olechnicka (2003), "The Innovative Potential of Polish Regions," mimeo (in Polish). Landes, D.S. (1998), The Wealth and Poverty of Nations, New York: W.W. Norton. Smtkowskim, M. (2001), New Relation of a Metropolis with Its Hinterland in the Information Economy, Studia Regionalne i Lokalne, Vol. 4(7). 134 6. POVERTYAND RURAL DEVELOPMENT Mark Lundell A. OVERVIEW 6.1 Developments in welfare in the rural sector are important to the welfare of the entire country, as close to 40 percent of the population lives in rural areas and close to 60 percent of the poor are rural. Moreover, with concerns about equity and social cohesion becoming increasingly important in the EU accession period, the welfare of rural dwellers takes on even greater importance. But rural can be a broad or narrow term. Indeed, some think of rural almost exclusively as villages in which farmers tend their fields and homesteads or transport their product to market. 6.2 In fact, rural society in Poland is a much richer and more complicated interplay of mobilizing and modernizing forces, in which a larger share of the historically farming families are increasingly involved in off-farm activities. As this chapter takes a fairly standard numeric definition of "rural" (the EU standard of population density of less than 100 per sq. km), much non-agricultural activity takes place in "rural" areas that are simply less densely populated than the areas in which these non-agricultural activities typically occur. Thus, the chapter's aim is to examine, regionally, the progress and setbacks encountered by farmers and other rural inhabitants in their development processes, recognizing that there is an employment and income dimension as well as a welfare dimension to rural society. These two dimensions can be mutually enforcing, in a "virtuous circle," but they are at work in their separate directions as well. 6.3 Improved rural income from agricultural production and other sources has a direct positive impact on rural welfare. It is also associated with higher productivity and higher wage expectations on the part of employees, as well as with an exit from agriculture that slowly pushes up labor productivity. Without the expansion of the cultivated area (which eventually must come to a halt), the rise of agricultural labor productivity increases the availability of labor for the rest of the economy and the average wage in these sectors. This is the essence of development: labor effort becomes increasingly differentiated into specialized occupations, and the share of labor outside of agriculture increases along with the average wage. 6.4 By the same token, some steps that directly enhance welfare (e.g., steps that provide improved education, health, and other services in rural areas) cannot be seen as merely "consumption improving." Indeed, they also have feedback for rural employment, as they represent investments that make investors more interested in locating their businesses in rural areas (small towns) and more able to find the type of employees they need. The reason for this is that improved rural life helps stem the "rural brain drain." In the final analysis, the operative word is balance: allowing those who prefer to remain in their rural setting to stay and to earn a real wage (given a generally lower cost of living) that is not greatly inferior to that of urban dwellers, while at the same time maintaining the nominal wage discount (compared to urban wages) which makes them competitive in attracting business investment to their rural areas over time. 6.5 Another theme that will be examined is that successful rural development is usually as much a function of good overall economic performance in the economy as a whole as it is of good rural development policies and outcomes. This is because of the strong links backward from the urban to the rural sector. These links take the form of a demand by urban areas for labor that leaves the agricultural sector and commutes to jobs in larger towns, or that out-migrates entirely from the rural sector: both of these movements serve to reduce underemployment in the agricultural sector by providing more land at the disposal of the remaining (full-time equivalent) farmers. 6.6 Overall, the findings of this chapter first demonstrate that a number of regions have been able to reach a type of rural balance in which the income of the mainly farming population can keep pace with that of other rural inhabitants who work largely off-farm. This situation is supported by larger farms, fewer farm workers, and higher labor productivity. The regions that demonstrate this type of balance are Zachodniopomorskie, Lubuskie, Warminsko-Mazurskie, Wielkopolskie, Pomorskie, Dolnoslaskie, and Opolskie. At the other end of the spectrum are regions which are have been slow to make the necessary structural changes to both greater off- farm employment and higher productivity agriculture. These regions are Podlaskie, Malapolskie, Lubelskie, Podkarpackie, Mazowieckie, and Swietokrzyskie. There is variation in this latter group, as some of these regions have above average labor productivity in agriculture but below average off-farm employment (Podlaskie), while others have relatively low agricultural productivity but better developed off-farm employment (Malapolskie and Mazowieckie). B. RURAL POPULATION IN A REGIONAL CONTEXT 6.7 Using the standard definition of "rural" adopted by the European Union (EU), namely a population density of a given administrative unit of less than 100 persons per sq. km, we have used powiat level data available from GUS to calculate the share of rural population in Poland. This level in 2001 was 36.4 percent, which is about 5 percentage points lower than the average for the CEE countries shown in Figure 6.1, and roughly on a par with the average for the Europe and Central Asia region as a whole (35 percent). (These latter estimates are from the World Bank's Development Indicators "Green" Data Book for 2000 and 2001, depending on the country). This definition of "rural" is used throughout this chapter (shown in Figure 6.1 as the LSA methodology), unless otherwise noted. 136 Figure 6.1: Share of Population in Rural Areas in CEE Countries, 2000-01 70.00% 60.00% 50.00% 40.00% 30.00% 20.00% 10.00% 0.00% ) ) ic a ic a ology Book Bulgaria gary ubl ania Hun Croati Rep Rom Sloveni Albania Method Green ech Republ (LSA (WB Cz Slovak Poland land Po Source: World Bank, 2000-2001. Development Indicators Green Data Book, Washington, D.C. and own calculations based on GUS, 2002. Statistical Yearbook of Regions, Warsaw, Poland. C. MAIN DEVELOPMENTS IN AGRICULTURE IN THE 1990S Production and Employment 6.8 The contribution of agriculture, forestry, and hunting to the gross domestic product (GDP) has been falling steadily since 1990, primarily because of the lack of growth in agricultural production volume but also because of declines in prices for agricultural and food products over the period. The decline in production volume resulted from demand reductions in the early 1990s (especially for livestock products) as well as from the restructuring of former state farms, whose production share fell rapidly during the mid-1990s. In turn, the restructuring of state-owned farm holdings led to the exclusion of some marginal farmland from agricultural use, but to the more rational use of the remaining land by the new private cultivators. (This was the case mainly in the north and west, where state farms were historically important.) 6.9 Thus, in the late 1990s the share of agriculture in value added fell from about 7 percent (1995) to just over 3 percent (Figure 6.2). Gross agricultural output has been more or less stable. However, as agricultural goods prices increased about 20 percent more slowly than growth in the prices of agricultural inputs, value added in agriculture fell substantially (about 25 percent from 1995 to 2000). Still, the share of agriculture in overall employment has been more or less constant at 26-27 percent. 137 Figure 6.2: Trends in Agricultural Value Added and Employment, 1995-2000 30.00% 25.00% 20.00% 15.00% Share in Gross Value Added 10.00% Share in Employment 5.00% 0.00% 1995 1996 1997 1998 1999 2000 Source: Foundation for the Development of Agriculture. Rural Poland 2002. Rural Development Report, Warsaw, Statistical Annex. 6.10 Though this is the employment share cited in official GUS publications, it is somewhat overstated as it does not consider the degree of underemployment present in the agricultural sector. Polish rural specialists concur that this degree of underemployment is very high, on the order of 25 percent of those who are solely or mainly employed on their farms1 (or, for very small holdings, these are referred to as "agricultural plots"). This would bring the level of the effectively employed down to 22 percent. Moreover, other analysis shows that in adjusting the employment level in terms of full-time equivalents, the share of agricultural employment in total employment falls to only about 10 percent (in 1996).2 This needs to be considered when comparing the relative productivity of the agricultural sector with the rest of the economy (which one author estimated to be on the order of 60 percent in the mid-1990s3). Land Tenure and Productivity 6.11 Of late (2000), Poland's agricultural land area has totaled about 18.5 million hectares, with 92 percent used by the private sector. In the late 1990s, agricultural land used by the private sector was fairly stable at 17.0 million hectares, but this was a marked increase from the 1990 level, since land leased from the state increased from 0.7 million hectares to about 2.4 million hectares (2000). Since the number of agricultural holdings (above 1 hectare) has been declining from about 2.05 million in 1996 to 1.89 million in 2000, average agricultural land per farm grew from 8.2 hectares to 9.0 hectares (in 2000). Moreover, during this process the distribution of land usage by farm size began to show a less evenly distributed and more bi-modal pattern. The share of total agricultural land owned by farmers with more than 15 hectares doubled from 20 percent to 40 percent, and the share of 1-2 hectare farmers grew slightly as well. At the same time, the shares of agricultural land owned by farmers of 2-5 hectares, 5-10 hectares, and 10-15 hectares all dropped continually in the 1990-2000 period. The current land tenure pattern is similar to that of Portugal in 1989, where this bi-modal pattern was even more distinct (see Figure 6.3). 1Izaslaw Frenkel, "Rural Unemployment in Poland" in Wies I Rolnictwo (2001), p. 81. Redundancy is defined as being able to work elsewhere without affecting the level of production on their farm or plot, and the calculations were based on data from the 1996 Agricultural Census. 2Tadeusz Hunek (2001), "Shaping the Post-Transitional Model of Polish Rural Economy," in Wies I. Rolnictwo, p. 13. 3Hunek, p.15. 138 Figure 6.3: Distribution of Ownership of Agricultural Land 70% 60% Group 50% 1990 Size 40% 1995 of 30% 2000 Share 20% Portugal 1989 10% Land 0% 0% 20% 40% 60% 80% 100% 120% Cumulative Percentage of Farms (smallest to largest) Source: Own calculations, GUS, 2002. Statistical Yearbook of Poland, Warsaw, Poland. 6.12 This is evidence of farm consolidation at the higher end of the distribution, as the number of farms above 15 hectares increased by almost 50 percent since 1990, with the area farmed by this group increasing by 2.7 million hectares. At the same time, the number of farms of 2-15 hectares fell by almost 400,000 since 1990, and their aggregate area fell by nearly 2.5 million hectares. The three groups with 2-5, 5-10, and 10-15 hectares also have the highest degree (almost 50 percent) of mixed farming crops and livestock), whereas the smallest farms and the largest farms are almost 50 percent specialized in the crops sector. In terms of marketed output, the share of the largest farms increased from 39 percent in 1996 to over 45 percent in 2000. The farms with more than 15 hectares also showed the largest share (over 60 percent) of family members who are between 18 and 44 years of age (the so-called "mobile age") and the smallest share of family members in the "after production age" (men over 65, women over 60). All in all, these are signs of an evolving commercial farming structure. 6.13 Just as land and overall agricultural production are becoming more concentrated in commercial farms, similar trends of consolidation in livestock production are evident over the past few years. These trends are most notable in the dairy sector, where a substantially improved integration between farmers and modernizing dairies has been achieved (most notably in northeast Poland).4 In just three years (from 1997 to 2000), the share of marketed milk receiving the top quality grade rose from 7 percent to 45 percent: this has been possible only with increased investments in quality control by the larger farms. In the meat sector, the share of pigs raised in herds of 50 head or more increased in 1996-2000 from 37 percent to 50 percent and a similar trend has been evident in the cattle sector (with the share of cattle raised in herds of 10 head or more having increased from 40 percent to 48 percent). 4Rynek Mleka, "Stan I Perspektywy, Kwiecien" (2002), Instytut Ekonomiki Rolnictwa i Gospodarki Zywnosciowej, Warsaw. 139 6.14 At the other end of the spectrum, and not including the roughly 1 million rural families with less than 1 hectare (about 400,000 hectares total, averaging 0.4 hectares each5), the smallest farms, of 1-2 hectares, represent an expanding share of all farms (almost one-quarter in 2000) and their average size was very stable at 1.4-1.5 hectares over the 1990-2000 period. Still, their land share is 5 percent, and their share of marketed output is less than 5 percent of the total. Moreover, these farms show the largest share of family members in the "after production age" (over 27 percent) and the smallest share of family members of "mobile age" (only 37 percent). 6.15 The dynamics of consolidation show considerable regional variation. Historically, the largest private farms have been in the Wielkopolskie and Kujawsko-Pomorskie voivodships (west central and north central), and it is precisely in these areas that further increases in farm size are most evident. Much of these increases in size are from the sale of state lands there. In the south and southeastern voivodships, farms have had a much smaller, average size, and distribution among family members is producing even greater miniaturization. But the rural families in these areas have long adjusted to the relative scarcity of land resources by pursuing off-farm labor activities (some of which are distant from their homes in other areas of Poland), thereby greatly diversifying their income sources. In central and eastern Poland (Lodzkie, Mazowieckie, and Lubelskie), the atrophy of small to medium-size farms (7-15 hectares) has been most pronounced, with the growth of larger farms coming mainly at the expense of the smaller farms. 6.16 Given that the ultimate aim of this consolidation is to increase labor productivity in the agricultural sector (as these farms are almost all owner-operated), it is informative to examine column 4 in Annex Table A.6.1 (Comparative Rural and Agricultural Employment and Productivity Characteristics by Region). The regional pattern for labor productivity in agriculture is quite clear: the five northern and western voivodships of Warminsko-Mazurskie, Pomorskie, Zachodniopomorskie, Lubuskie, and Wielkopolskie have the highest levels, ranging from 1.3 to 2.3 times higher than the average for Poland. These areas also tend to have agricultural employment shares well below the average (it should be noted here that the agricultural employment levels are not adjusted for underemployment or full-time equivalents). In contrast, the southern and eastern voivodships of Malapolskie, Podkarpackie, and Swietokrzyskie have the lowest agricultural productivity levels as well as the highest agricultural labor shares. (Podlaskie is a bit of an outlier, with a high agricultural labor share but a higher than average agricultural labor productivity, as is Slaskie, which has the lowest agricultural labor share, but a below average level of agricultural labor productivity.) D. TRENDS IN RURAL POVERTY AND UNEMPLOYMENT IN THE TRANSITION PERIOD 6.17 Over the past few years unemployment levels in the rural sector have risen in line with urban unemployment levels, both doubling from the level of roughly 10 percent in 1997 to roughly 20 percent in early 2003 (see Table 6.1). However, the unemployment rates in the rural sector have been 1.5-3.5 percentage points lower than in the urban sector, and this gap widened consistently during this period. Comparing rural unemployment among non-agricultural workers and farmers, we see that the households with a farmland user display unemployment rates about 2.5 times lower than the rural non-farming population. This reflects the high unemployment rates among former state farm workers, who are now mainly households without farmland users. It also reflects the fact that many current farmers are technically employed, though they are really quite underemployed. 5Agricultural Census data for 1996 show that 29 percent of these landowners get their income exclusively from non- agricultural sources, while another 70 percent have mainly non-agricultural sources. Only 1 percent characterize their income sources as mainly agricultural. 140 Table 6.1: Unemployment Rates in Urban and Rural Areas, 1997-2003 Unemployment Rate 1997 1998 1999 2000 2001 2002 2003* Urban 10.7% 11.1% 14.4% 17.0% 19.4% 21.3% 21.1% Rural 9.3% 9.7% 13.2% 14.6% 16.4% 17.8% 19.6% Rural farming 5.8% 5.5% 8.1% 8.6% 10.1% 10.9% 11.9% Rural non- 16.0% 17.3% 21.7% 24.7% 27.6% 29.3% 30.1% farming** *First quarter only. ** Those with a farmland user. Source: GUS, Labor Force Surveys, 1997-2003. 6.18 When one examines the distribution of the rural population's consumption across the national quintiles over the period 1994-2001, it is apparent that, despite the comparatively lower unemployment noted above, the rural sector accounts for a disproportionately large share of the poorer quintiles (see Table 6.2). Over this period, the rural share of the population has been steady at 36-38 percent, while the rural shares of the population in the poorest two consumption quintiles have been much higher: about 56 percent for the poorest quintile and 47 percent for the second poorest. The rural share of the poorest quintile rose in 1998-99, but has since fallen back. This increase in rural poverty was reflected in an increase in the rural share of the total number of poor in Poland to 61 percent in 1999, and a rise in the rural poverty head count to over 22 percent. Table 6.2: Rural Shares of Consumption Quintiles and Poverty Head Count, 1994-2001 By Quintiles Rural Share Rural Poverty Rural Share 1 2 3 4 5 Of Total Poor Head Count 1994 52.7% 45.0% 38.5% 31.6% 24.6% 52.7% 23.1% 1995 54.4% 48.2% 40.3% 33.1% 27.2% 55.1% 22.3% 1996 55.4% 48.3% 40.7% 32.4% 26.6% 56.6% 20.8% 1997 54.3% 46.9% 38.3% 30.4% 23.9% 55.7% 21.1% 1998 57.3% 47.0% 38.9% 29.6% 21.7% 59.2% 19.9% 1999 59.2% 47.1% 38.6% 30.6% 21.7% 61.0% 22.1% 2000 56.7% 48.0% 39.6% 32.9% 22.1% 57.6% 21.3% 2001 56.3% 47.8% 39.4% 31.9% 23.2% 56.5% 22.2% Source: GUS, 1994-2001. Annual HBS. 6.19 This corresponds to the time when the urban unemployment level exceeded rural unemployment by the smallest margin in the past seven years. Since then, as this gap has widened, the rural share of the total number of poor has dropped by 5 percentage points. This implies that the rural sector has been better able to cope with the reduced growth rates of the three years than the urban sector. Still, the poverty head count has been steadily in the 21-22 percent range during this period. This level is only marginally lower than the peak in the rural poverty head count 23 percent in 1994. 6.20 When the incidence of poverty is examined across those rural inhabitants with at least 0.5 hectare, it is clear that those with such access to land have much smaller farm sizes (based on HHS data for respondents with at least 0.5 hectares). It is clear that the poverty head count has not been the highest among the smallest farmers (quintile 1 in Figure 6.4). Indeed, these smallest farmers usually have a poverty head count which is about 2-5 percentage points lower than the poverty head counts of the middle three quintiles. As expected, the largest farmers are consistently those who show the lowest poverty head count, though this is still fairly high at over 15 percent. 141 6.21 In sum, lower unemployment levels in the rural sector do not make for lower rural poverty compared to the urban sector. Lower educational attainment and skills must be bringing rural wages down in comparison with the urban sector in the off-farm sector of rural employment. In addition, although farmers are counted as employed, there is high underemployment as referred to above. This underemployment manifests itself as lower average labor productivity, which will be discussed in detail below and examined on a regional basis. Figure 6.4: Rural Poverty Head Count by Quintiles of Agricultural Land, 1994-2001 30.0% 1 25.0% 2 20.0% 3 15.0% 4 10.0% 5 1994 1995 1996 1997 1998 1999 2000 2001 Source: GUS, 1994-2001. Annual HBS. E. IMPORTANCE AND STRUCTURE OF RURAL EMPLOYMENT 6.22 An examination of the structure of employment and income (both rural and urban) by region provides the backdrop for an analysis of the trends in the rural segments of these regions, because it answers the question, "How important is rural sector employment?" From Figure 6.5, it can be seen that the rural sector (off-farm and agricultural employment) accounts for about 35 percent of employment nationally and 40 percent or more of total employment in fully 10 out of Poland's 16 voivodships, and over 50 percent in 6 voivodships. In only 5 voivodships (Lubelskie, Lodzkie, Mazowieckie, Podlaskie, and Swietokrzyskie) does rural agricultural employment exceed rural off-farm employment. Interestingly enough, the 4 voivodships in which rural off-farm employment exceeds rural agricultural employment by the greatest margins (Lubuskie, Pomorskie, Warminsko-Mazurskie, and Zachodniopomorskie) are those which have the highest agricultural labor productivity levels. 142 Figure 6.5: Share of Rural Off-Farm and Rural Agricultural Employment in Total Regional Employment, 80% 70% 2001 60% 50% 40% 30% 20% 10% 0% skie skieelskie skie zkieolskiewieckOpols ie kie kie kie kie kie ki tal Dolnsko- oslapomo ub r rs L Lubu s s Lod rskie To MalopMazo dkarpacPo Po dlamo Sla rzyskiezurskielpolsmo land Po Kujaw Swi sk etok o-ma Wieiopo Po rmin Rural Off-Farm Employment Agricultural Employment Wa Zachodn Source: Own calculations based on GUS, 2002. Statistical Yearbook of Regions, Warsaw Poland. 6.23 Annex Table A.6.1 shows that, for Figure 6.6: Trends in Rural Non-Agricultural Employment and Poland as a whole, the Total Agricultural Employment, 1995-2000 share of rural employed 5000 Rural Non- persons whose primary 4500 Agricultural 4000 employment is outside of Employment (Incl. 3500 Informal) agriculture is on the Employed 3000 Rural Non-Ag in order of 60 percent. 2500 Enterprises > 5/9 Thus, the "lower bound" 2000 people for the agricultural share 1500 1000 of rural employment is Agriculture, forestry Thousands 500 and fishery 40 percent. The share of 0 those employed 1995 1996 1997 1998 1999 2000 primarily outside of agriculture ranges Source: Own calculations based on GUS unpublished data. regionally from a high of 86 percent in Zachodniopomorskie to a low of 24 percent in Podlaskie. These regional non-agricultural rural employment levels are strongly correlated with the comparative labor productivity in agriculture (0.43). Over the past few years, the level of agricultural employment has been fairly steady at about 4.3 million, with 2.7 million of this in rural areas.6 Given a total rural employment of 5.3 million, this would mean an upper bound for the agricultural share of rural employment of 50 percent (Annex Figure A.6.8). 6.24 Over the same period (1995-2000), rural non-agricultural employment first grew rapidly, gaining over half a million jobs (including small enterprises and the informal sector) between 1995 and 1998 (Figure 6.6). This development was quickly reversed in 1999-2000, as about half of these off-farm jobs were lost. This trend is also visible in the level of rural, non-agricultural (here, official) jobs, which also declined by about 250,000 between 1998-2000. This is an alarming trend in rural employment, as non-agricultural employment accounts for 50-60 percent of rural employment. Indeed, the 10 percent decline in these jobs means that further diversification of income sources for rural inhabitants has been halted (temporarily, it is hoped). 6This measure uses the LSA methodology (the population density definition of "rural") and does not exclude those farmers whose primary source of income may be outside of agriculture (i.e., it includes part-time farmers). 143 6.25 A related development here is the emergence in 2000, for the first time, of net migration from urban to rural areas (illustrated in Annex Table A.6.1). For the country as a whole, the rate is low, at only 0.03 percent of the population (2.8 per 10,000 population), but for some voivodships it is five to ten times greater. These high levels of net migration from urban to rural areas are taking place in rural areas where the share of rural employment outside of agriculture has been highest: Slaskie, Dolnoskaskie, Malapolskie, and Pomorskie. (These net in-migration levels show a very high correlation [0.68] with rural employment outside of agriculture). Moreover, an examination of gender and age categories shows that the net inflow is made up of men in the age group 35-64 (women and all other age groups showed net out-migration from rural areas).7 6.26 Apparently, the reduced growth rates of the service and industrial sectors in Poland in the 1999-2000 period took the form in rural areas of reversing the previous trend of increased diversification of income sources: those voivodships which had made the greatest progress in this area now suffered most, and many male family members who became unemployed returned home. Some Polish authorities cite this as a case of "last hired, first fired," with employers rationalizing that formerly rural workers would have an adequate (agricultural) base to which to return. However, this proposition needs further, rigorous, validation. Figure6.7: Distributionof PopulationandEmployment Typeby 100% Degreeof UrbanConcentration 80% 60% 40% 20% 0% n Populatio Employed oyed ure y s icult vice Industr Ser Unempl Agr BigCities VoivodshipCapitals Urban Districts Rural Districts Source: Own calculations based on GUS, 2002. Statistical Yearbook of Regions, Warsaw, Poland. 6.27 Nevertheless, in 2001, as can be seen in Figure 6.7, the structure of jobs for rural dwellers is quite different from that for more urbanized areas. Despite accounting for over 35 percent of the population, rural districts have only about 20-25 percent of the industrial and service sector jobs and about 65 percent of the agricultural sector jobs. (The other 35 percent of agricultural sector jobs are accounted for by people resident in small town powiats with population densities above the 100 per sq. km threshold.) The rural share of the unemployed is, nevertheless, on a par with the rural share of the overall population. 6.28 However, as noted above, this does not take into account the high degree of underemployment in the agricultural sector. This is estimated at probably over 1 million, or about one-quarter of the able-bodied population in agriculture. Still, outright (open) unemployment is highest in the rural sector among non-farm operators. In total, the open rural 7FDPA, Rural Development Report, p.20. 144 unemployment is roughly 1 million. Range estimates are 700,000 to 1.2 million depending on the source (labor offices versus LFS), with the rates being highest in Zachodniopomorskie, Dolnoslaskie, and Warminsko-Mazurskie (where they exceed 25 percent). In addition to the openly unemployed and the under-employed, the annual increment to the rural labor force is expected to be roughly 50,000 (GUS projections for 2010 in Population Projection, 1999). Thus, the next ten years will be a special challenge as demographic shifts leading to a decline in the absolute level of the labor force will be felt only after 2010-15. Until then, the urban sector will clearly not be able to absorb the full increment. 6.29 Currently, about half of off-farm employment involves commuting to urban areas.8 Still, the 50,000 person increment noted above is probably a low estimate of the number of rural jobs needed per year. Given the 2 million rural unemployed and underemployed farmers needing to exit the sector, partially solving these problems by finding employment for half of them over ten years would amount to generating 100,000 jobs per year. Added to the annual increment to the workforce of 50,000 persons, and assuming that 50 percent of this total will commute to cities, about 75,000 new jobs annually are needed outside of the agricultural sector in rural areas. 6.30 It is of course an open question as to how much off-farm employment can be generated in the rural sector. As Annex Figure A.6.2 shows, rural districts account for only about 15 percent of all enterprises in Poland but about 30 percent of individuals conducting economic activity (data from 2001, not including farming). Indeed, the number of rural enterprises per 1,000 population in rural areas is only about 1.75 for Poland as a whole (Annex Figure A.6.3). This ratio is significantly above 2.0 in only Lubuskie (3.4), Zachodniopomorskie (2.9), Pomorskie (2.7), and Dolnoslaskie (2.5). In the southern and eastern voivodships, this ratio is on the order of 1.0, with the lowest level being in Malapolskie (0.7). Enterprise assets and investments in the rural sector are also only about 15 percent of the totals for Poland, and the level of enterprise assets per off- farm employee is on the order of 80,000 PLN (Annex Figures A.6.4 and A.6.5). (The regional variation of this indicator is much less than the ratio of enterprises per 1,000 population.) 6.31 In a regression analysis with off-farm employment as the dependent variable, the results indicate much greater responsiveness to self-employment as opposed to enterprise employment. Off-farm employment is highly inelastic with respect to enterprise assets (only 0.16), whereas this elasticity is much higher with respect to self-employment (0.77). Apparently, at least to date, self-employment is playing a larger role in off-farm employment than larger more formal enterprise job creation. Of course, the level of wages may also differ between these two types of jobs. This is examined below. F. STRUCTURE OF RURAL INCOME 6.32 Considering agriculture's importance to regional income, Figure 6.8 shows that only in Podlaskie and Swietokrzyskie voivodships does agricultural income contribute close to 15 percent of total average regional income per capita, and these two regions have among the lowest per capita income levels, as well as the highest shares of agricultural employment. Even more telling is the very low contribution of agricultural income to regional income in Podkarpatskie voivodship, at only 4 percent, despite the 45 percent agricultural employment share. 8FDPA, p.19. 145 Figure 6.8: Sources of Average Monthly Per Capita 900 Income of Households, by Region in 2001, PLN 800 700 600 500 400 300 200 100 0 oslapomo skie skieels e ie ie skielpols kirskie To tal r Kujaw Dolnsko- Lub LubuskiLodzalop kie kie kie kie ols iec omo land M Mazow Opodkarpackie lsk Po PodlaPom skieorskieSlaskokrzyskiezurWie Swietinsko-ma rm chodniop Po Wages (Hired Work) Agriculture Self Employment Wa Social Benefits Za Other Source: GUS, 1994-2001. Annual HBS. 6.33 Figure 6.8 also reveals the greater importance of self-employment: its importance is almost double that of agriculture nationally (9 percent versus 5 percent), and it is a lower contributor to income than agriculture only in Lubelskie, Podlaskie, and Swietokrzyskie. However, these data (GUS Statistical Handbook of Regions, 2002) are aggregated across rural and urban areas, whereas Annex Table A.6.1 provides the breakdown of income sources by rural and urban areas separately (calculated from HBS data for 2001). Figure 6.9: Value of Monthly Consumption, PLN Per Capita 1,400 1,200 1,000 800 Rural 600 Urban 400 200 0 1 2 3 4 5 Total By Sub-Sample Quintiles Source: GUS, 1994-2001. Annual HBS. 6.34 The rural-urban disaggregation (Annex Table A.6.2) shows the constancy of "social transfers" and "other sources" at about a 40 percent share of household income, for all but the poorest quintile in both rural and urban areas. (Quintiles are based on consumption expenditures for rural and urban sub-samples separately.) (See Figure 6.90.) Where income sources differ substantially is in the categories of agricultural revenue and wages and salaries. In rural areas, this share is about 27 percent, except for the lowest quintile where it is only 22 percent. This agricultural revenue share (which includes own consumption of home-produced food products) of the lowest quintile probably reflects the smaller plot sizes of the poorest farmers. Surprisingly, 146 the agricultural revenue shares of the richest two quintiles are not significantly different from those of the second and third quintiles. This is the result of two factors. First, all rural quintiles are composed of a relatively large share of non-farmers, and second, even larger farmers are successful at diversifying their income away from agriculture. (The larger shares of the category "Other sources," which includes self-employment, should be noted.) Wages and salaries also show a fairly constant share of rural income of about one-third. 6.35 In urban areas, despite the presence of roughly 2 million "farmers," the share of agricultural revenue in urban income is only 1-2 percent. This reflects the low level of agricultural income from these mainly household plots, as well as the small share of employment for which part-time "farmers" account. The larger shares of "wages and salaries" exceed the comparative shares for rural households by a full 25 percentage points, but comparing rural and urban quintiles (one by one) there are no significant differences in the shares of "social transfers" or "other sources." What is clear is that, again, when a comparison is made across similar quintiles between the rural and urban sub-samples, urban household consumption levels are on average about 16 percent higher than those for rural households. This difference rises from the first quintile (+7 percent) through the fourth quintile (+27 percent), but is the lower for the fifth quintile (+21 percent). 6.36 In drawing final conclusions about the relative importance of various types of employment in the rural sector, it is key to examine the income share accounted for by employment type compared with its share of the people employed. (The data for these calculations are calculated using the LSA rural definition [i.e., population density], and are aggregated so that industry and the service sector account for "wages Figure 6.10: Rural Shares of Employment and and salaries" whereas "persons Earned Income by Employment Type, 2001 conducting economic activity" 60.00% account for self-employment.) 50.00% Here we examine only earned 40.00% income and leave out social 30.00% transfers. Thus, in the rural sector, agriculture provides about 35 20.00% percent of earned income (Figure 10.00% 6.10), although it accounts for about 0.00% 50 percent of those employed. This Agriculture Wages and Self-Employment disparity is partly a function of the Salaries underemployment factor discussed Employment Income w/o Social transfers earlier.9 Self-employment provides about 18 percent of earned income, Source: GUS, 1994-2001. Annual HBS. while its employment share is about 15 percent. Wages and salaries accounts for one-third of employment, but generates 47 percent of earned income. Thus, making the comparative labor productivity calculations implied by the above shares, one can say that, in the rural sector, non- agricultural wage labor is about twice as productive as agriculture, whereas self-employment is about 70 percent more productive than agricultural employment. 6.37 These findings are similar to the regionally disaggregated results presented in Annex Table A.6.1 on comparative income levels. Those results demonstrate that the disposable income 9If one were to exclude from the calculation the roughly 1 million people underemployed in agriculture and distribute this reduction across urban and rural areas proportionally to existing agricultural employment levels, the share of agriculture in rural employment would fall to 44 percent. 147 of farmers ranges from 50 to 90 percent of that of non-agricultural workers in the same region (with the notable exception of Zachodniopomorskie, where the level is 150 percent of non- agricultural workers' income). Summary 6.38 The composite index presented in Annex Table A.6.1 tries to summarize the impacts of agricultural productivity and non-agricultural rural employment ­ a proxy for rural development. This index is a product for each region of the regional share of employment outside of agriculture relative to the national average times the regional agricultural productivity level (again, relative to the national average). This index summarizes the regional perspective: the lowest levels are (in ascending order) in Podlaskie, Malapolskie, Lubelskie, Podkarpackie, Mazowieckie, and Swietokrzyskie (all below 80 percent, with 100 percent being the average for Poland), where rural development is weakest. The highest levels (in descending order) are in Zachodniopomorskie, Lubuskie, Warminsko-Mazurskie, Wielkopolskie, and Pomorskie (all above 160 percent), where progress to a dominant off-farm economy with a high productivity agricultural sector has been strongest. The same pattern is substantiated by the variation in the ratio of a region's agricultural share in valued added relative to its agricultural share of employment (Annex Table A.6.1, last column on the right). G. WELFARE AND INVESTMENT LINKAGES: AVAILABILITY OF INFRASTRUCTURE AND SERVICES 6.39 One aspect of rural welfare is the increased food security that comes from being able to reliably produce a larger share of one's household's food consumption. From the HBS data set (2001) it has been calculated that rural inhabitants in Poland produce a share of their own food consumption that is about five times more than their urban counterparts (Table 6.3). For the poorest three rural quintiles, this share is on the order of 25 percent of all food consumed, falling sharply to just above 15 percent for the richest rural quintile. This level is still about three times that of the urban cohort, which ranges fairly evenly for all quintiles at about 5 percent. Table 6.3: Share of Food Consumption Own Produced, by Quintiles,* Rural and Urban Own Produced Food 1 2 3 4 5 Total Rural 25.3% 24.5% 22.5% 19.1% 15.9% 22.5% Urban 5.8% 5.7% 4.7% 3.9% 3.4% 4.5% *Quintiles are defined over equivalent adults using household weights. Source: GUS, 1994-2001. Annual HBS. Education 6.40 If they are to gain productive off-farm employment, one particularly important hurdle that rural inhabitants will have to overcome is their low level of education when compared with their urban counterparts. The 1996 Agricultural Census and later GUS surveys (summarized in Table 6.4) show the following comparative trends in educational attainment in urban and rural areas. In 1995, the main differences were that the rural population share with only a primary education was 24 percentage points higher than that for urban areas, whereas the urban population with a completed secondary or tertiary education was 27 percentage points higher than that in rural areas. In 2001, these gaps persisted between the educational levels of the urban employed and the rural employed in farming. (In fact, the gaps mentioned above were both on the order of 40 percent when urban employed and rural farming populations were compared.) Interestingly enough, the gaps in vocational training were insignificant. 148 6.41 This indicates that rural inhabitants have adequate interest in and access to vocational training, but given the income differential observed earlier, there must be a wide differential in the focus and quality of vocational training offered to rural students versus urban students. Since re-qualification programs such as supplemental short-term courses, job retraining, or consulting services for prospective non-agricultural workers and potential small businessmen are limited in rural areas, the refocusing of vocational training on non-agricultural work and general business skills is clearly a priority that needs to be given much greater attention and funding. Here more focus needs to be allocated to computer skills, basic business competency, and foreign languages. Health 6.42 Data from the HBS (2001) indicate that private household health expenditures (on hospitals and sanatoria) in rural areas are only about one-third of those of their urban counterparts (Figure 6.11). If public health care services were generally deemed to be adequate then this would not be a significant problem. However, the availability of doctors' services in rural areas is quite unfavorable in comparison with urban areas. As shown in Annex Figure A.6.7, this availability in rural areas is about half of that in urban districts, and only about one-fourth of the level available in big cities and in voivodship capitals. Table 6.4: Educational Level Reached by Shares of the Population (of those age 15 and above) Urban Rural Rural Urban Rural Rural 1995 1995 1995, Employed, Employed Employed Ages 20- 2001 Non- Farming, 39 Farming, 2001 2001 Primary or 31% 55% 22% 7% 14% 44% Incomplete Primary Basic Vocational 25% 28% 45-51% 29% 44% 38% Complete Secondary 34% 15% 25-30% 44% 34% 18% Tertiary 10% 2% 3% 20% 8% 1% Sources: 1996 Agricultural Census Report, and FDPA, Rural Poland ­ Rural Development Report, 2002, Table A19, p 112. 149 Figure 6.11: Private Household Expenditures on Hospitals and Sanatoria per Adult 2.5 2 1.5 1 0.5 0 t 2 3 4 al atsk laskie ie ie ie kie ie Rur elskieuskieLodzkapolskioweckOpol ie e ie i skie Poores Richest Urban laskie arskie zyskie urskie orsk nos pom Lub Lub Pod Slaskokr maz Dol Mal Maz Podkarp Pomossk iet Sw insko- Wielkopos opom hodni Kujawsko- Warm Zac Source: GUS, 1994-2001. Annual HBS. 6.43 Indeed, as Table 6.5 shows, the distribution of rural household expenditures for health services is much less variable than that of urban households. Though the share of total expenditures for health services purchased by the poorest rural quintile is more than twice that of the poorest urban quintile, these shares do not increase with income nearly as much as in urban areas. Nor are they nearly as variable as the level of overall consumption expenditures are for the progressive rural quintiles of consumption. This indicates that the income elasticity of these expenditures in rural areas is very low. This appears to be more a problem of adequate access than of inherent preference by rural inhabitants, though this supposition requires more rigorous analysis. In sum, it would appear that rural access to health care is particularly inferior to that in urban areas. Table 6.5: Distribution of Household Expenditures for Health Services by Income Quintiles* 1 2 3 4 5 Total Urban 5.3% 10.3% 17.4% 25.9% 41.0% 100.0% Rural 13.5% 20.1% 22.1% 22.7% 21.5% 100.0% Total 8.1% 13.6% 19.0% 24.8% 34.5% 100.0% *Quintiles are defined over the whole sample, not within sub-samples. Source: GUS, 1994-2001. Annual HBS. Social Assistance 6.44 In rural areas, the main issues regarding social assistance revolve around the difficulty of characterizing need and of preventing leakage of social assistance to those who are not truly needy. Since farmers generally do not file income tax returns on their net profits, it is very difficult for local government to accurately evaluate need in an objective manner. Moreover, given the relatively large share of the importance of the off-farm and often informal employment of rural inhabitants, this problem is further compounded for rural gmina social assistance administrators. First of all, the official characterization dictated to these officials of net agricultural income at 200 PLN/hectare per month (roughly 250 PLN per capita per month) is high given the results reported in Figure 6.7. This overstates agricultural income significantly and 150 leads to the disqualification of primarily agricultural households from social assistance when the need is probably justified. 6.45 Thus, much greater flexibility is needed at the gmina level in determining need, with a commensurate increase in the oversight function by the powiat level. Second, there is the issue of why 85 percent of the funds allocated by the central budget (which provide 97 percent of total social assistance expenditures) are rigidly allocated with little ability for re-distribution across types of social assistance. Here, a greater role for block grants is likely to be warranted for rural social assistance. Ultimately, the issue is: When will an income tax declaration system be more adequately included in the determination of social assistance eligibility? This is not a long-term issue but can be addressed in the medium term. Physical Infrastructure 6.46 As can be seen from Figure 6.12, investments in the provision of rural potable water have been largely successful in meeting the needs in rural areas, and the same is generally true for the rural road network. Indeed, rural Figure 6.12: Total Area (km2) and Infrastructure (km) infrastructure is in 100% good enough shape 80% to permit a 60% reasonable 40% development of tourism: tourist 20% accommodations 0% per 100 population Area Water Line System Sewerage System Gas Line System in rural districts are Big Cities Voivodship Capitals Urban Districts Rural Districts on a par with those in urban districts Source: Own calculations based on GUS, 2002. Statistical Yearbook of Regions, Warsaw, Poland and large cities (Annex Figure A.6.6). The key areas in which investment in rural infrastructure appears to have been inadequate are those of rural telephone services, and, even more, in the provision of rural sewerage. The former will be key to improving the development of knowledge-based employment, distance working, tourism and other rural SMEs. Better rural sewerage will be important for agro-processing certification and for the attraction of investment in new medium and large enterprises. H. THE NEED FOR REDEFINING THE GOALS OF RURAL POLICY 6.47 The main goals of the rural development policy articulated in various government agricultural policy documents in the 1990s have been the fostering of the income parity of rural citizens with their urban counterparts as well as increases in productivity in the agricultural sector which would lead to greater competitiveness. In the late 1990s (and in 2001, as seen from the survey data), disposable income per capita in rural areas was about 75-85 percent of that in urban areas, while the average net agricultural income per farmer (or farm worker) was about half the average non-agricultural sector wage.10 However, in one key way the objectives of parity and increased productivity are not consistent, because the increase in a farm's labor productivity depends to a large degree on the increase in the farm's land to labor ratio, which will mean some farms will inevitably shrink and yield less income for their farmers, forcing them to look to off- 10FDPA, Rural Poland 2000: Rural Development Report, 2001, pp. 44-45. 151 farm sources to make up the loss. As noted above, this consolidation process progressed in the 1990s, and farmers recognize that this is the future path in agriculture. (Polls of farmers in the late 1990s revealed that only 50 percent want to stay in farming and about 45 percent would consider selling their land in order to invest profitably in off-farm enterprises and/or help their children make a start.)11 6.48 Indeed, the government recognizes that policies must be pursued to promote farm productivity and income as well as to boost off-farm income. The main instruments used to achieve the former are input subsidies, credit subsidies and guarantees, and agricultural price support through market intervention. The main instrument for boosting off-farm income has been the KRUS system, for by allowing such low KRUS premiums for participating contributors, the KRUS system in effect subsidizes the income of current workers by 180 PLN per worker per month (with the actual KRUS contribution averaging about 50 PLN/month per contributor).12 However, in the long run, as urban income per capita continues to grow, it is not feasible to try to maintain parity of rural income with urban income levels through transfer payments. Instead, the higher income of rural families will be achieved most frequently through off-farm employment. 6.49 Thus, the key issues are how best to generate off-farm employment and farm productivity in the long run, while in the medium term using income transfers to cushion the effects of the recent fall in agricultural GDP on the small and medium-size farmers. The analysis undertaken in this chapter first reveals that a number of regions have been able to reach a type of rural balance under which the income of the mainly farming population can keep pace with that of other rural inhabitants who work largely off-farm. This is supported by larger farms, fewer farm workers, and higher labor productivity. Needing further investigation is the question of whether these regions reached this balance as a result of exogenous improvements in agricultural productivity which generated higher average labor productivity, or whether the surrounding non-agricultural rural economy and nearby urban centers pulled excess farm labor out of the agricultural sector, thereby reducing underemployment in farming and raising average labor productivity. With the increased unemployment level nationally, this latter "pull" factor has been significantly reduced. 6.50 Clearly, there are regions where agricultural labor productivity is much lower than the average in Poland, and where the off-farm share of employment is still low. These regions tend to be the voivodships of southeastern and eastern Poland. In these areas, a concerted effort by local communities will have to be made to enable more willing rural inhabitants to capably run small businesses and to attract medium and large enterprises to these areas to provide a greater supply of off-farm wage jobs. 6.51 In this regard, the state has a responsibility to ensure that public expenditures are used efficiently when targeted to the generation of new jobs by employers. Under the current MARD and SAPARD investment expenditures and the credit subsidy programs, the emphasis has not been on economic diversification and non-agricultural employment: MARD and other rural programs focus only about 11 percent of resources on economic diversification and non- agricultural employment, and 5 percent on vocational training. The future structure of the EU structural and cohesion funds is due to be finalized in mid-2003 between the European Commission and the Government of Poland, and is therefore currently not clear, but indications are that rural infrastructure will continue to absorb more funding than economic diversification 11Rural Poland 2000, pp.60-61. 12One can estimate the KRUS subsidy by assuming that if the KRUS system had the same ratio of contributors to pensioners as ZUS (roughly 3:1, rather than the current ratio of 0.8) and premiums covered payouts to pensioners, then the premium would have to be 17 billion PLN distributed over 6 million "needed" contributors, or roughly 230 PLN/month. The subsidy is therefore about 180 PLN/month per contributor. 152 and non-agricultural employment. This needs to be re-considered, as the levels of the provision of communal services in rural Poland rose in the 1990s and in most areas are reasonably high compared to those in urban areas. Whatever the level of rural infrastructure investment, future priorities for increased investment should be concentrated in health care provision, sewerage provision, and telecommunications. 153 REFERENCES Central Statistical Office (2002), Demographic Yearbook of Poland, Warsaw. Central Statistical Office (2002), Statistical Yearbook of Poland, Warsaw. Central Statistical Office (2002), Statistical Yearbook of Regions, Warsaw. Foundation for the Development of Agriculture (2000), Rural Poland 2000, Rural Development Report, Warsaw. Foundation for the Development of Agriculture (2002), Rural Poland 2002, Rural Development Report, Warsaw. Frenkel, Izaslav (2001), "Rural Unemployment in Poland," in Wies I Rolnictwo Village and Agriculture Selected Papers 4 (113):71-90. Glowny Urzad Statysczny (2001), Rocznik Statysczny Rolnictwa. Warsaw. Hunek, Tadeusz (2001), "Shaping the Post-Transitional Model of Polish Rural Economy," in Wies I Rolnictwo Village and Agriculture Selected Papers, 4 (113): 10-23. Kowalski, Andrzej, J. Rowinski, and M. Wigier (2002), Polish Food Economy, IER, Warsaw. Rynek Mleka (2002), "Stan I Perspektywy, Kwiecien" Instytut Ekonomiki Rolnictwa I Gospodarki Zywnosciowej, Warsaw. Warsaw Agricultural University Research and Implementation Center (2000), The Strategic Options for the Polish Agro ­ Food Sector in the Light of Economic Analyses, Warsaw. World Bank (2000-2001), Development Indicators Green Data Book, Washington, DC. 1996 Agricultural Census Data. 154 Annex Table A.6.1: Comparative Rural and Agricultural Employment and Productivity Characteristics by Region Disposable Non-Ag Share of Comparative Income of Disposable Income Net urban to Employment Agriculture's Ag Rural Labor Farmers of Farmers rural Share Share in Value Employment Employment Productivity Relative to Relative to Non- migration Weighted by Added Relative Share in Outside of in Ag as National Ag Workers' per 10,000 Comparative to Agricultures' Voivodship, Agriculture*, Share of Rest Average of Disposable Income population Ag Labor Employment Region 1999 2001 (Q1) of Economy Farmers in Same Region (in 2000) Productivity Share, 2000 Dolnoslaskie 15% 74% 19% 71% 50% 17 159% 0.19 Kujawsko-pomorskie 26% 57% 16% 112% 87% -2.3 105% 0.15 Lubelskie 51% 37% 17% 85% 58% -18 70% 0.12 Lubuskie 15% 82% 31% 106% 78% 6.7 284% 0.26 Lodzkie 31% 52% 14% 102% 70% -0.2 84% 0.13 Malapolskie 34% 67% 9% 87% 56% 15.4 69% 0.09 Mazowieckie 25% 45% 14% 115% 62% 7.7 72% 0.13 Opolskie 27% 69% 18% 90% 67% 4.4 141% 0.21 Podkarpackie 45% 60% 10% 87% 66% -1.7 71% 0.08 Podlaskie 45% 24% 18% 107% 88% -42.9 50% 0.16 Pomorskie 14% 73% 20% 100% 73% 15.3 168% 0.19 Slaskie 11% 79% 14% 131% 93% 31.5 124% 0.15 Swietokrzyskie 47% 49% 14% 96% 73% -7.6 77% 0.12 Warminsko-mazurskie 24% 65% 29% 97% 86% -27.4 216% 0.29 Wielkopolskie 25% 61% 27% 112% 76% 8.4 187% 0.23 Zachodniopomorskie 14% 86% 35% 207% 149% -0.6 344% 0.32 Total 27% 58% 15% 100% 69% 2.8 100% 0.14 *Employed with primary source of income outside of agriculture. Sources: FDPA, Rural Poland ­ Rural Development Report, 2002; pp.107, 113, 116, 123, and 124. Last column from Statistical Yearbook of Regions, 2002, p. 303. Table A.6.2: Income Sources for Rural and Urban Households - Shares, by Quintiles Agricultural Wages and Other (Incl. Rural Revenue Salaries Social Transfers Self-Employment) Total Revenue 1 22% 32% 37% 9% 100% 2 28% 32% 31% 9% 100% 3 25% 34% 31% 11% 100% 4 27% 33% 27% 13% 100% 5 27% 35% 23% 16% 100% Total 26% 34% 28% 13% 100% Urban 1 1% 51% 37% 11% 100% 2 2% 55% 32% 11% 100% 3 2% 57% 31% 11% 100% 4 1% 57% 29% 13% 100% 5 1% 62% 20% 17% 100% Total 1% 58% 27% 13% 100% Source: HBS 2001. 155 Figure A.6.1: Average Monthly Wages, PLN 3000 2675.48 2500 2333.96 2000 1889.04 1741.57 1500 1000 500 0 Wages Big Cities Voivodship Capitals Urban Districts Rural Districts Source: Calculated from Regional Statistical Yearbook, 2002. Figure A.6.2: Number of Enterprises, Total and with Foreign Capital, and Persons Conducting Economic Activity 100% 80% 60% 40% 20% 0% # of Enterprises # Foreign Capital Persons Conducting Enterprises Econ Activity Big Cities Voivodship Capitals Urban Districts Rural Districts Source: Calculated from Regional Statistical Yearbook, 2002. 156 Figure A.6.3: Number of Rural Enterprises per 1000 Population 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00 oslaskieomor skie ie kie zurskie ki tal ols kie To mors Doln ko-p MalapoMazowieckieOpolsPodkarpackiePodlaskiPom ie e lsk skie lsk kie orskie laskie Lube Lubu Lodzkie SSwietokrzysko-ma Wielpiopo Poland Kujaws Enterprise by Pop Foreign Enterprise by Pop Warmins Zachodn Source: Calculated from Regional Statistical Yearbook, 2002. Figure A.6.4: Enterprise Assets, Investments, and Gmina and Powiat Revenues 100% 80% 60% 40% 20% 0% Enterprise Assets Enterprise Gmina Revenue Powiat Revenue Investments Big Cities Voivodship Capitals Urban Districts Rural Districts Source: Calculated from Regional Statistical Yearbook, 2002. 157 Figure A.6.5: Rural Enterprise Assets and Investments per Off- Farm Employee, PLN Assets Investments 250,000 14,000 200,000 12,000 10,000 150,000 8,000 100,000 6,000 4,000 50,000 2,000 0 0 al Dolnko-po oslasmorLubeLubu LMaMa kieskielskieskie kielskieckie kieackilaskiersSkielaskrzyskzurskieomo kieTot e ie ie odzlopozowieOpolsarpPodPomo lskirs Podk ko-maodn WielppoPoland io Kujaws Swietok Warmins ch Za Enterp. Assets by Off Farm Empl. Enterp. Investments by Off Farm Empl. Source: Calculated from Regional Statistical Yearbook, 2002. Figure A.6.6: Tourist Accommodations Provided per 1000 Population 1600.00 1400.00 1200.00 1000.00 800.00 600.00 400.00 200.00 0.00 Tourist Accomodations Big Cities Voivodship Capitals Urban Districts Rural Districts Source: Calculated from Regional Statistical Yearbook, 2002. 158 Figure A.6.7: Doctors per 10,000 Population 60.00 45.00 30.00 15.00 0.00 Doctors Big Cities Voivodship Capitals Urban Districts Rural Districts Source: Calculated from Regional Statistical Yearbook, 2002. Figure A.6.8: Employment Shares by Level of Urbanization 100% 80% 60% 40% 20% 0% Big Cities Voivodship Urban Districts Rural Districts Capitals Agriculture Wages Self Employed Source: Calculated from Regional Statistical Yearbook, 2002. 159 7. THE ECONOMIC STATUS OF WOMEN 13 Katarzyna Tyman-Koc 7.1 At present, an evaluation of Poland's economic situation offers a fairly gloomy forecast for those social groups which, because of either poor financial standing or difficult circumstances, have to apply for social welfare funds -- whether in the form of unemployment benefits, re- training funds, or means for launching a new business. Owing to the shortage of public funds, the development of institutional care for small children and others requiring such care has been unsatisfactory, which, in turn, is bound to affect family income patterns. The insufficient financing of institutions for the care of children and dependent persons exerts a significant influence on the rate of the professional activation of women. However, apart from the traditional family-related reasons, women's incomes, professional status, and promotion prospects are affected by other factors. A set of stereotypes in which a woman is perceived as predestined for certain activities and only suitable for certain positions in the professional hierarchy remains in force. Recent changes to the Labor Code, which are meant to promote equal rights for women and men in the labor market, have not yet been fully internalized into the social awareness of the general public. This situation raises the following questions. 7.2 Does the law discriminate between women and men in the fields of labor relations, social insurance, and support for families with children? Are stereotypes connected with the perception of female and male social roles still in force? Do stereotypes affect the professional and economic status of women? 7.3 The subsequent sections of this chapter attempt to provide answers to these questions. The chapter begins with an analysis of the way stereotypes function, followed by a discussion of labor-related legislation and employee rights with regard to parenthood. The next section focuses on business initiatives among women, and women's entrepreneurial spirit and willingness to take financial risks. Another issue crucial to the diagnosis of the present state of affairs is the question of whether the fact that women are ready to take on financial risk, as well as the risk connected with running one's own business or performing public duties, translates into the gradual advancement of women in the workplace, in business, and in politics. In other words, the question is whether the "glass ceiling" phenomenon is a fact or merely a theory far from Polish reality. Yet another relevant subject discussed is the extent to which the professional status of women is reflected in their financial circumstances in old age. Related to this is the question of which variables pertaining to the way women function in the labor market appear to be most important for the size of the future benefit, and what changes should be introduced in this respect. Finally, the picture is completed with an evaluation of the economic situation of women living in rural areas, and of the extent to which the environment they inhabit makes their socioeconomic status different from that of women living in urban areas. A closing section provides conclusions drawn from the chapter. 13Based on the Report,­ Ple a szanse ekonomiczne, (Gender and Economic Opportunities), Eugenia Mandal, Boena Balceak ­ Paradowska, Ewa Lisowska, Boguslawa Budrowska, Irena Wóycicka, Janina Sawicka, Maria Parliska, WB, 2003 160 A. STEREOTYPICAL PERCEPTIONS OF FEMALE AND MALE ROLES 7.4 Gender-related stereotypes are oversimplified judgments and ideas concerning the behavior of women and men, shared by the society as a whole, and imprinted during the processes of growing up and of socialization within the society. In stereotypical perceptions, all individuals are allocated a so-called "group description" whereby they are no longer unique; instead, their rich variety is reduced to a few categories. 7.5 Stereotypes of what is masculine and what is feminine operate on two planes: the open (direct) plane ­ when a person's gender serves as a basis for prejudice and discrimination, for example in hiring and promotion decisions, and the hidden (concealed) plane ­ when a person is assessed and approved of exclusively on the basis of his/her achievements in the fields seen as "typical" of a given gender (e.g., in cases in which women are expected to hold subordinate positions or perform duties which match the traditional definition of female roles). Both female and male gender stereotypes have a complex structure and consist of four components, namely: stereotypes connected with mental characteristics, roles, physical appearance, and occupation. 7.6 Furthermore, there are many negative stereotypes (also known as "myths") with regard to women and work. They include the following beliefs: · Women do not need to go to work because they have an alternative source of income (i.e., their husbands). · Women do not need to earn much and will work for very low salaries. · Women themselves do not want to have professional careers. · Female employees are less diligent, flexible, and available with regard to working hours. · Women are different from men in terms of personal qualities, which means that they should be given "feminine" jobs (e.g., they should work as teachers rather than as lawyers). · A woman will never make a good boss. · The presence of women in the workplace has a detrimental effect on employee concentration. · Women "steal" jobs that would otherwise be available for men (read: women are to blame for growing unemployment). · A woman is not entitled to have a job as much as a man, because she is not the head of the family. · And in any case, women take unemployment better than men, and being unemployed is more harmful to a man than to a woman. 7.7 Some of the consequences of stereotypical perceptions are various phenomena related to women in the workplace, such as: the conflict between family and work, the different 161 expectations with regard to professional careers, discrimination in hiring decisions, the "glass ceiling" (promotion barriers), the "glass escalator" (promotion barriers in "female" branches of the economy), the "sticky floor" (barriers connected to low-status jobs with little or no promotion prospects), occupational segregation, lack of mentors, exclusion from the circle of informal contacts, "token" status, and sexual harassment. B. THE LEGAL AND INSTITUTIONAL FRAMEWORK 7.8 The transition to a market economy in Poland brought about changes in the economic structure: from then on, the principle of effectiveness was deemed a foundation stone of the economy. A completely new situation emerged for the labor market, with an excess of resource supply over actual labor demand. Unemployment appeared and started to grow (Table 7.1). Under such circumstances, the status of women began to deteriorate. The female employment rate went down, although its drop dynamic was similar to that observed among men. Those sections of the national economy that began to expand and grow absorbed a predominantly male workforce. As for women, many of them worked for the public sector and in those branches of the economy that have traditionally been considered as female-dominated (education, health care). Table 7.1: Employment Rate in Poland in the Years 1995­2002 Specification 1995 1996 1997 1998 1999 2000 2001 2002 Total 50.7 51.2 51.5 51.0 48.0 47.4 45.5 44.6 Men 58.5 59.4 59.8 58.9 55.9 55.2 52.5 50.8 Women 43.7 43.8 44.0 43.9 40.7 40.3 39.0 38.9 Note: Data for 2001 and 2002 refer to the fourth quarter; data for 2002 to the second quarter. Sources: GUS (Chief Statistical Office)1999; GUS 2002. 7.9 The average female remuneration is lower than its male counterpart (Table 7.2), although in recent years this gap has gradually been narrowing, especially in the private sector and in small companies. By and large, the wage gap results from female domination in budgetary sections of the economy, where the level of gratification is generally relatively low. It is also caused by the difference in educational choices between men and women: as a rule, women tend to graduate from the arts. Table 7.2: Average Gross Remuneration of Full-Time Employees According to Ownership Sectors, Gender, and Level of Education, October 2001 The share of female Public sector Private sector remuneration in Level of education relation to male remuneration Men women men women Public Private sector sector Higher 3,350 2,488 4,964 3,716 74.3 74.9 College 2,336 1,802 2,610 2,231 77.1 85.5 Secondary vocational 2,497 1,898 2,261 1,838 76.0 81.3 Secondary comprehensive 2,409 2,097 2,435 1,997 87.0 82.0 Vocational 2,238 1,386 1,745 1,291 61.9 73.9 Primary and lower 2,092 1,336 1,655 1,315 63.9 79.5 Source: GUS 2000. 7.10 Disparities in the status of women and men in the labor market may also result from discriminatory practices. Women face a greater threat of unemployment, particularly long-term 162 unemployment, even though as a group they are better educated than men and are willing to improve their skills and qualifications. 7.11 The tendency to reject female job applicants is related to the issue of employer expectations regarding employees. Namely, employers are very much concerned not only with employee qualifications and skills, but also with employee availability. Owing to their maternal and family obligations, women are seen as less available time-wise than men. The fact that women are perceived predominantly through their family obligations may result from a stereotypical way of thinking, but it also reflects the dominant, traditional model of family life. 7.12 Women are the main beneficiaries of employee rights that are intended to reconcile professional obligations and family duties. Some of these rights are connected with the maternal function of women (health protection, provisions for periods of pregnancy, and for childbirth and breastfeeding). Other prerogatives (such as childcare leave, time off to provide care, part of maternity leave) are available for men as well. However, these provisions are used almost exclusively by women. 7.13 From the employer's standpoint, employee rights are responsible for the increase in labor costs incurred by work absences that qualify for the benefit or compensation paid out by the employer for time not spent at work. In addition, employee absence may cause organizational problems. Figure 7.1: Number of People Taking Childcare Leave, 1993-2000 (in thousands) 400 336.1 350 284.4 300 234.8 250 204.9 184.6 170.7 200 151.8 138.8 150 100 50 0 1993 1994 1995 1996 1997 1998 1999 2000 Source: GUS (2001b). 7.14 The legal provisions that are meant to reconcile professional and family duties in fact work as barriers to hiring women. Many women are aware of that fact; consequently, they try to exercise the rights for which they are eligible as rarely as possible. In the joint opinion of both employers and female employees, the development of childcare units (day-care centers, nursery schools) side by side with other forms of institutional care for schoolchildren (extracurricular afternoon activities), and an increase in their accessibility (a change in fees), would be a step in the right direction. It would also be helpful to introduce more flexible solutions with regard to employment schemes and the structure of working time. 163 C. BUSINESS INITIATIVES AMONG WOMEN 7.15 According to statistical data, a significant, stable trend can be observed: women are becoming more interested in running their own businesses, both in the form of self-employment and as employers providing workplaces for other people. In 1998, the number of women with their own registered economic activity outside of farming was almost five times the number in 1985 (see Table 7.3). During the same period, the corresponding indicator for men barely doubled. 164 Table 7.3: Employers and the Self-employed Outside of Individual Farming, by Gender, 1985 and 1998 (in thousands) Employers and the self-employed 1985 1998 Total 574.5 1,574.4 Women 131.7 632.5 Men 442.8 941.9 Source: Ewa Lisowska's calculations based on the 1997 Statistical Yearbook, p. 128, and the 1999 Statistical Yearbook, p. 129. 7.16 In the years 1992-2002, there was an increase in the representation of businesswomen among the total female workforce, as well as in the total population of business people. At the same time, the representation of women in the total self-employed population in Poland is among the highest in comparison with other European countries (see Table 7.4). Table 7.4: Women in the Total Population of the Self-employed in Poland and in Selected EU Countries (average data for years 1990-97) Country Representation in percentage Belgium 28.9 Finland 31.1 France 26.0 Greece 19.4 Spain 26.8 The Netherlands 32.8 Ireland 20.1 Germany 28.3 Sweden 25.7 Great Britain 24.8 Italy 23.4 POLAND 34.0 Sources: OECD Small and Medium Enterprise Outlook, 2000, and E. Lisowska's calculations based on data from GUS Statistical Yearbooks. 7.17 The groups of businesswomen and businessmen are similar in terms of age, level of education, and number of people they employ. The most conspicuous discrepancies are found in their acquired trade or profession: men are usually engineers, technicians or skilled workers, whereas women are typically economics or arts graduates. As far as branch of activity is concerned, both women and men predominantly choose agriculture, followed first by trade and repairs, then by manufacturing and intangible services (women ), or construction and tangible services -- for example, transport (men). If we focus on business people from urban areas, we see that businessmen typically operate in four branches: trade, construction, industrial processing, and transport (75 percent of the total). As for businesswomen, their first choice is trade (46 percent), and then their activities are dispersed in many different branches of the economy. 7.18 Irrespective of gender, women and men list the same incentives that motivated them to start their own businesses, such as: the need for independent decision-making, the need for a higher income, qualities of entrepreneurship and willingness to take risks. More often than men, women are motivated by the threat of unemployment, the lack of other options, and the need to 165 prove their worth (i.e., to show a spouse or a partner that they can successfully run their own company as well as the household (see Box 7.1). 7.19 In the last decade in Poland, entrepreneurial attitudes were by and large a result of the economic transformation with all its consequences, such as the collapse of state-owned companies, the rise of unemployment, and the decrease in workforce demand, which affected women to a greater extent than men. Still, it is important to note that women who start a business are often driven by the wish to take independent decisions, to prove their worth in their profession, and to pursue a higher income. Box 7.1: Key Factors Motivating People to Start a Business 1. Independence - 91% (89%) 2. Decent income - 84% (89%) 3. Inborn entrepreneurial spirit - 70% (75%) 4. Previous work experience - 64% (70%) 5. Good opportunity to earn money - 56% (46%) 6. Inclination to risk - 48% (51%) 7. Need to prove one's worth to a partner - 40% (27%) 8. Unemployment threat - 35% (27%) Note: figures for men are in brackets Source: Findings of the 1995 survey research, "Polish business '95," conducted by the author within the statutory research of the Warsaw School of Economics World Economy College. Of a sample of about 1,050 business people, 305 sent in their answers to the questionnaire, including 143 women. 7.20 Many women have entrepreneurial skills, are unafraid of risk and are ready to face the business challenge. Under favorable circumstances, those characteristics may surface and become activated. In addition, they can be reinforced by appropriate individual education and by state policies intended to promote the equal status of women and men in the labor market. During the last decade, when a woman started her own business it was usually because she had lost a job in a state-owned company and had few chances to find other employment -- not because of a favorable state policy. 7.21 Survey research (Lisowska, 1996; SME Foundation Report, 2000; Demoskop 2001) has shown that the major barriers to the development of business initiative in Poland are economic in nature and issue from the following causes: · The market situation, namely: low demand owing to the restricted income of the population; problems with access to a quality workforce; growing competition and the unethical behavior of many business people. · The economic policies of the state, including the fiscal policy, which exerts a negative rather than a positive influence on the growth of entrepreneurship, since employers carry an excessive burden of taxation on their income and their employees. · A lack of capital and no access to sources of financing. · The limited access to foreign markets, owing to problems in meeting the requirements of foreign standards and the cost of promotion in those markets. 166 7.22 Other obstacles to the development of a small and medium business sector are educational barriers, such as the following: · A limited access to information concerning reforms in the pipeline, changes in legislation, and business support programs. · The high costs of specialist training courses available on the market. · The lack of cheap, reliable and easily accessible (e.g., on the phone or the Internet) consulting and advisory points. 7.23 Apart from the barriers mentioned above, women must also cope with social barriers connected with the traditional perception of female and male roles in the society, together with the lack of widespread approval of those women who choose to pursue their professional career alongside of or instead of a family life. In Poland it is still commonly believed, even among the political and economic elite, that when unemployment is high, men are more entitled to work than women (Siemieska, 1998). 7.24 Regarding the issue of access to financing sources, women in Poland are offered neither special loans nor credit lines, nor more favorable conditions with regard to the granting of loans or credit guarantees. Women can take advantage of the available forms of financing on the same terms as men. However, it is more difficult for women to comply with the requirements and obtain loans, bank credits or venture capital financing. 7.25 Government policy towards the SME sector (with regard to the part of the sector that is currently implemented and with regard to provisions in the document "Government policy towards the SME sector until 20002") does not contain any programs designed especially for women. 7.26 As can be seen in many European countries, as well as in Canada and the United States, such programs are indispensable given the presence of discrimination against women in the labor market. In order to improve the status of women at work, there are temporary provisions that are preferential to women in terms of access to information and training, participation in conferences and international fairs, and access to financing sources. Poland needs such solutions if it is to see decreased unemployment and the growth of entrepreneurial attitudes among women. 7.27 In today's world of growing competition, business people must take on their own continuous training and improvement, and must provide the same for their employees, side by side with introducing new technologies. Businesses that do not grow quickly will lose the position of important market players. Computers and access to the Internet have become standard in any company that hopes to grow and to keep up with the competition. On the basis of the data available, it would be difficult to specify how many of the firms run by women have access to the Internet, how many of them are innovative, and whether businesses run by women have more difficulty gaining access to new technology in comparison with firms run by men. Further research is needed in this respect to determine certain issues and to specify the findings. 7.28 As far as access to training schemes is concerned, the market offer is rich and varied; in addition, women's organizations provide such training courses for their members. From the point of view of small business owners, the only drawback is the cost of some courses, since women are well aware of the need for self-improvement and would willingly participate in training schemes. Continuous professional training is another issue that is as yet unresolved in Poland. 167 First, the general public has to realize that the extremely rapid changes in the contemporary world and the ongoing process of globalization require permanent educational efforts, irrespective of a person's age. Second, there is a lack of relevant programs in the courses offered by schools and universities. D. BARRIERS AND LIMITATIONS TO WOMEN'S CAREER PATHS: THE "GLASS CEILING" PHENOMENON 7.29 Women in Poland, as in many other countries, encounter a "glass ceiling" which keeps them from achieving leadership positions in the hierarchy and from undertaking the highest- ranking duties. The term "glass ceiling" refers to the obstacles encountered by women in managerial positions: it stands for a situation in which "promotion prospects are within sight and yet beyond one's reach." Various data from different fields of political, economic, and educational activity testify to the presence of the "glass ceiling" phenomenon in Poland, which prevents women from holding high ranking, prestigious positions involving authority, a high financial status, and human resources and financial management. 7.30 The findings of the "glass ceiling" research in the study have made it possible to pinpoint certain barriers and limitations which stand in the way of women's professional career advancement, and the roots of those barriers go deeply into Polish culture. We can distinguish between three types, or three consecutive levels, of barriers encountered by women who wish to have a successful career: · The first types of barriers are internal barriers and inhibitions: many women lack self-confidence with regard to their strengths, skills, and abilities; they feel insecure and unfit to hold high ranking positions. · The second types of barriers are related to the widespread tendency to attribute to women the traditional gender-related roles, with all the consequences of this approach. Social expectations related to maternity are deeply internalized by women; but at the same time this burden that women carry is by no means taken into account in organizational and institutional planning. Consequently, women either start a career after their children have been brought up (women councilors' path), or postpone maternity for a later date (women managers' path). · Finally, there are external barriers connected with discriminating attitudes and behavior, the rules governing the functioning of professional and political circles. A wide variety of limitations can be observed in this respect. 7.31 One of the obstacles encountered by women in their professional careers is the homogeneity of managerial staff, especially that of higher ranking staff, which exerts an influence on the attitude and behavior of male managers and directors, on the methods of employee recruitment and selection, and on performance assessment tools. Basically, men decide among themselves who should be appointed to positions of authority. In many cases a woman would not even be told that there was a vacancy. Women are excluded from the flow of information and are not considered as potentially eligible candidates. 7.32 Another obstacle issues from male solidarity versus lack of solidarity among women. Men tend to support one another, especially in favorable political circumstances. One woman councilor says: Typically, men look down on women and see them as inferior. (...) if you give a man a choice, he will always choose another man, unless he knows the woman really well and 168 knows that she is much more suitable for the job. (...) They are mates, they have this solidarity. They know each other, they function well together and cooperate with each other. They respect their female colleagues, they know that we are out there, but they like to keep good things to themselves. 7.33 Women repeat time and again that at work they can rely only on themselves and their own resources; the support of their male colleagues is out of the question. 7.34 Both female and male interviewees pointed to many irregularities in mutual relations. Generally speaking, in this respect women seem to be at a disadvantage. To begin with, there is a double standard of work performance assessment: the results achieved by women are evaluated according to different standards from those achieved by men. Male shortcomings and vices are treated with greater leniency and understanding. A male interviewee admits that: a man might be relatively inexperienced and have some other weaknesses, and he will still get away with it, but with a woman it's a completely different story. 7.35 Many women object to the patronizing attitudes exhibited by men: Whenever my male colleagues run out of arguments to defend their standpoint, they always say, come on, she's a woman, we aren't going to argue with her; which really is, I don't know...I don't want to say [it's] contemptuous, but it definitely is patronizing. This condescending approach also surfaces when men begin to lecture women: For instance, my male colleagues tell me what I should and shouldn't do. They preach to me, and yet they never do it to each other. (...) When I speak about something at work, a man will often interrupt and say that I should stop talking. Apparently, there is a stereotype of telling a wife what to do, and a similar habit of instructing a female colleague. 7.36 Stereotypical views regarding the way women function may also deprive women of being promoted to high ranking positions in the professional hierarchy and in public life. As a rule, women are perceived as more emotional. Moreover, men still appear to cherish the stereotypical image of a woman who is a gentle, kind, sensitive individual. If a woman dares to go beyond the limits of this stereotypical image, she is in for severe criticism. One female interviewee says: A guy won't think twice, he'll just bang his fist on the table and shout at the employee, but I have to stop and think first. For them it's like playing the part of a "true man." On the other hand, I often hear "It does not become you, you are a woman," but it has nothing to do with that. Another stereotypical remark made about women is that they do not exhibit enough initiative and entrepreneurial spirit. 7.37 Both women and men appear to share a widespread negative perception with regard to women in charge, which proves that the common stereotypical belief that a woman cannot be a good boss is deeply rooted in the society. A woman who used to be a boss herself actually thinks that women make better bosses than men, because they are more involved and because they have to fight twice as hard as men if they want to keep their position. But the very same woman, when asked whether she would rather report to a woman or a man, will opt for a man, because "she is used to working with men, and in any case, she might have a bit of a male brain herself." In concluding, the woman in question declares that women deserve to be appointed to the highest positions, as they are in no way different from men. When she promotes someone at work, she never takes gender into consideration. 7.38 All in all, it can be stated that the issue of the "difference" between women and men has not been given adequate attention, and that this may lead to all kinds of misunderstandings. It would be difficult to assess, beyond all doubt, whether women are considered well suited for high ranking positions and duties. According to most female interviewees, women are suited to 169 exercise authority and to hold the highest posts. This belief is often based on personal experience, among other things. When asked to enumerate the positive qualities of women, female interviewees point out that women are well-organized, their relations with people are full of honesty, they show a lot of respect for others and for their time, they are efficient as negotiators and have typically feminine "soft" management skills and, finally, they are down-to-earth, diligent and not prone to corruption. E. WOMEN IN THE RETIREMENT PENSION SYSTEM 7.39 Since 1990, the Polish retirement pension system has been subject to many changes. The most significant of these changes were implemented in 1992 and in 1998 (retirement system reform). Although they were in many ways reasonable and necessary, these changes did have a detrimental effect on the situation of women, especially with regard to: provisions including the periods of care for a small child in the base of the retirement pension size; provisions governing pension size; and provisions concerning the right to the minimum retirement pension. 7.40 The core change introduced into the public (first) pillar of the retirement insurance system consists of a new, totally different method of benefit calculation. The new formula is based on the concept of "Notional Defined Contribution" and was first implemented in Sweden. Under this scheme, an individual account is established for every person insured, which features the size of retirement rights capital at a given point in time. Retirement rights capital is the sum of contributions paid into the account of the person insured, subject to annual indexation. Under the new formula, there is a closer correlation between the size of the benefit and the amount of contributions paid during the whole period of the person's professional career. 7.41 In the second (private, capital) pillar, the size of the benefit is contingent on the sum of collected contributions, minus administrative costs, plus the return on investment of the capital accumulated on the account of the person insured; it also depends on the remaining lifespan expectancy at the time of retirement. In the first pillar, the new formula stipulates a close correlation between the size of the benefit and the amount of accumulated premiums. Analogically, in the second pillar the size of the contributions directly translates into the size of the future retirement benefit. And, in turn, the amount of accumulated premiums depends on the earnings received during a person's professional career, the length of the employment history (which also depends on the actual retirement age), and the length of other periods with contributions of premiums. Furthermore, the size of the future benefit is related to the actual age at retirement ­ this factor determines the average remaining lifespan expectancy, which exerts a significant influence on the size of the benefit in both the first and the second pillar. 7.42 After the reform, the statutory retirement age for women and men remained the same: at present, analogically to the previous system, it is 60 years for women and 65 years for men. Table 7.5: Size of Future Retirement Benefit for Women and Men Depending on Retirement Age, in Relation to Average Benefit Size (assuming that both the first and the second pillar use universal lifespan expectancy charts) Retirement age Women Women Men Male benefit = 100% In relation to average benefit 1 2 3 4 55 70% 44% 64% 56 71% 47% 67% 170 57 71% 51% 71% 58 72% 54% 76% 59 72% 58% 81% 60 72% 62% 86% 61 72% 66% 92% 62 72% 71% 98% 63 72% 76% 105% 64 72% 81% 113% 65 72% 87% 120% Source: Simulations based on the model of the Social Policy Budget, Market Economy Research Institute. 7.43 According to the simulation, the average female benefit, depending on the retirement age, will vary from 70 to 72 percent of the average male benefit, assuming that they retire at the same age (Table 7.5, column 2). This discrepancy is directly related to the difference in compensation size, the length of employment history, and the length of other periods included in the base of retirement pension size. 7.44 The first benefits paid under the provisions of the new system will appear only in 2009, and thus there is enough time to design and carry out changes that will verify some systemic solutions and adjust them to the advantage of women. An absolute priority is the gradual introduction of an equal retirement age for both women and men, preceded by a thorough and reliable information campaign. When provisions for the payment of benefits from the second pillar are implemented, they should be based on universal lifespan expectancy charts. Finally, there should be room for provisions granting access to the deceased spouse's capital accumulated in the second pillar with regard to women under 60 years of age who are eligible for the family pension benefit. F. SOCIOECONOMIC SITUATION OF WOMEN LIVING IN RURAL AREAS 7.45 In the past, the development of rural areas was identified with the development of agriculture, which provided employment for the vast majority of the rural population. In the course of time, as both the productivity of agriculture and the efficiency of the farming workforce went up, the percentage of the population involved in agriculture went down. The future prospects for rural communities depend less and less on farming activity, which calls for a new way of thinking about employment opportunities in the countryside, including those pertaining to the situation of women. As can be inferred from statistical indicators such as employment and unemployment rates, the number of people whose occupations are counted among the total population of farmers remains relatively high; moreover, this group has the lowest unemployment rate. On the other hand, the highest unemployment rate in rural areas affects people without farms (as a rule, these people were employed by the former sector of state farms, now closed down). This problem is especially painful in the western and northern regions of Poland, which had the lion's share of state ownership in the farming sector. In addition, the level of so-called hidden unemployment is relatively high among farmers. 7.46 Not so long ago, women played a most important role in Polish farming, since men were involved (either part-time or full-time) in other kinds of activities outside of farming. They either constituted a separate category of people working in two jobs, or they quit agriculture 171 permanently and earned their income working for heavy industry, transport, or the construction sector. Consequently, women were bound to take over a lot of duties previously performed by men. In the course of time, when many young women from rural areas found employment in industry and the service sector, women started to leave the countryside. As a result, the demographic structure of the rural population changed to a great extent, and the workforce began to age even more. A gap between the sizes of the female and male populations in the productive age range appeared, and many men were unable to find a female partner. 7.47 During the 1990s, after the transition to the new economic system, the outflow of the workforce from farming to other sectors came to a halt. Numerous factories had to close down, and redundant workers started coming back to the countryside. Following the changes in the labor market, high unemployment, and the establishment of an institutional framework of job agencies, new unemployment benefits appeared that were paid out to people registered in the Labor Offices. Women who take this benefit are usually young, under 35 years of age. Once they lose their title to the benefit, a majority of those women are financially supported by their parents: as graduates without any professional experience, they have difficulty in finding their first job. In comparison to the situation in other countries, in Poland the proportion of the female workforce that is in the farming population is quite high, and all women who work on farms must combine family duties and work. The conflict between these two roles, typical of all working women, is aggravated by the hierarchy of authority present in the families and communities of the rural areas. In rural families there is another dimension to women and work: women who contribute to the family income frequently act as partners in their own right during the decision-making process. Consequently, a woman receives more respect and her sphere of authority changes. 7.48 In recent years the country has experienced a growth in unemployment and a decrease in farming profitability. Driven by these circumstances, some women have decided to start their own trade, manufacturing, or service businesses. Their "business college" is either the previous workplace or the farm. There is a wide variety of business initiatives launched by women: in crafts, agro-tourism, processing, and the sale of farm produce. This line of work appeals particularly to young women, since it allows them to combine professional obligations with running the household and taking care of the children. A mixture of farming and work outside of the agricultural sector provides support for less profitable farms and prevents the depopulation of some regions (in other words, their economic and social degradation). As for the reasons behind the business initiative of women in rural areas, the overriding factor is the need to gain additional income. 7.49 Polish women living in rural areas have high hopes for the improvement of their economic status with Poland's accession to the EU. In the long term, farming is likely to become more profitable, and the all-round development of rural areas should bring about more opportunities for employment. G. CONCLUSIONS 7.50 When we look at the country's economic situation and at related issues pertaining to the labor market, we may draw fairly depressing conclusions with regard to both the present and the future status of women. As of the end of the first quarter of 2003, the rate of unemployment in Poland was 18.7 percent, exceeding the EU average by 10.5 percent.14 Women are still a dominant group among the unemployed, although the growth dynamic in this respect is getting 14Unemployment registered in the first quarter of 2003, Statistical data and studies, Chief Statistical Office, Warsaw, 2003. 172 smaller (this is also the case among men). The longer the period of unemployment is, the larger is the share of women in the total population of the unemployed: 39 percent of all unemployed women have been unemployed for more than 24 months, and thus they have very few chances of finding a new job. 7.51 Such a high level of unemployment (especially long-term unemployment) among women cannot be satisfactorily explained by a slower rate of economic growth. And it would be difficult to attribute it to labor legislation, which guarantees equal treatment to all employees, irrespective of gender. Finally, the suggestion that women tend to exhibit little economic initiative and tend to avoid the risks connected with running one's own business does not seem to be a sufficient explanation. 7.52 From the research on which this chapter is based, it becomes evident that the overriding reasons behind the inequalities affecting women's economic activity are the historically determined, culturally conditioned, stereotypes, which in the course of time became reflected in discriminatory privileges in the area of labor legislation and social security. Although the provision of most of those privileges has been extended to employees of both sexes, in the minds of employers it is women who remain responsible for providing care in the family, which automatically makes them less available for work. The same line of thought lies behind the efforts to maintain different retirement ages for the sexes. It is seen as a way of compensating women for their unpaid labor connected with running the household. As a result, the female retirement benefit is much lower -- especially under the new system, where the retirement benefit is related not only to the size of the contributions but also to the length of the insurance history. 7.53 If we look at this from the point of view of the potential beneficiaries of the social welfare system, and of those social groups which are most likely to enter the poverty zone, it is almost certain that, owing to income distribution mechanisms, older women (i.e., those no longer professionally active) will be permanent clients of social welfare. This would be the outcome of the stereotypical perception of social roles affecting market mechanisms as regards working income ­ a legal one, outside of the gray zone -- the income which is subject to social insurance. 7.54 To change this scenario, first and foremost we must consider education. Social stereotypes are predominantly passed on and imprinted in children's minds in the process of education, both at home and in the relevant institutions. Although external institutions have a limited influence on the way children are brought up at home, they have a lot to say with respect to the system of education and the choice of profession. Another step would be to provide gender-neutral employee assessment, and to introduce flexible employment schemes (but not part-time employment), which would make it easier to reconcile family duties and personal obligations. Finally, all of these efforts would not be complete without a well-developed network of institutional care for children and other family members requiring care. The last step should be to implement the same retirement age threshold for both sexes, so that women and men have equal chances to accumulate enough capital to be financially secure in their old age. 173 wb13696 P:\POLAND\PREM\Living Standard Assessment\4RED\Part 2\Papers\FINALPLS Vol 2 Chap6-7 0301.doc March 24, 2004 1:08 PM 174 PART III: A CASE OF "JOB-LESS GROWTH"? 8. HOW FAR IS WARSAW FROM LISBON? Tito Boeri, Pietro Garibaldi, and Mauro Maggioni A. Introduction 8.1 Poland is set to enter the European Union (EU) with the highest unemployment rate among member countries. Over and above the huge challenges that such historical enlargement poses, Poland faces the task of tackling its ailing labor market. Within the EU, labor market reforms and structural reforms in general are certainly a top priority in light of the high unemployment rate, which, although declining significantly from its peak in the late 1990s, remains unacceptably high. But at the policy level in Europe there is a clear shift of emphasis from lowering unemployment to "creating more jobs," by which policymakers seem to mean "increasing the employment rate." For example, the Presidency Conclusions of the Extraordinary European Council, Lisbon, March 2000, place much of the emphasis on increasing the labor market participation of member countries. 8.2 In Lisbon, quantitative targets were specified for the employment rate. Member countries are expected to reach a 70 percent employment rate by the year 2010. In addition, specific targets were set for particularly weak employment groups, such as women (a target of a 60 percent employment rate) and the elderly (a target of 55 percent). In this scenario, it is natural to look closely at the current distance between the Polish labor market and the European targets. 8.3 Beyond the quantitative assessment of the distance from Lisbon, there are at least three practical reasons why economists and policymakers may wish to empirically analyze employment and employment growth. First, employment is easier to measure than unemployment, because it does not depend on subtle distinctions between individuals who are in the labor force and those who are not. Second, for a given level of unemployment, higher net employment growth results in higher output and lower financial pressures on the social security system. (The latter point is explicitly recognized by the European Council in its Conclusions.) Third, a much richer analysis can be conducted by using employment rather than unemployment as the main variable of interest. In particular, data on unemployment do not ascribe workers to a particular sector or type of contract, whereas data on the composition of employment by sector and by type of contract are available. 8.4 Looking at the Polish labor market, a substantial amount of work has been carried out on the rise of unemployment (see, for example, Churski [2002] and Newell and Pastore [2000]). Less emphasis has been given to employment, with the notable exception of the work of World Bank (2001), which carried out estimates of job creation and job destruction in the Polish manufacturing sector. Further, it appears that the measurement problems concerning unemployment statistics appear particularly serious in the case of Poland, where a persistent and widening gap can be seen between the registered and the LFS- based unemployment rate. 8.5 This chapter looks at the basic facts on employment, and analyzes the Polish employment performance by a number of factors, including sector, age, gender, ownership status, and full-time and part-time employment (section B). The key statistic in this respect is the fall in the employment rate by 8 percentage points in the last five years, which resulted in a loss of some 1.5 million jobs. Section C considers the role of growth, and argues that in light of the sizable GDP growth performance the Polish economy seems to suffer from a dramatic "job-less growth" syndrome, a labor market disease typical of Europe in the late 1990s. Since growth is not likely to be the smoking gun that explains the low employment performance, Section D asks whether such behavior is the result of inadequate labor market institutions. The discussion considers in some detail the role of Employment Protection Legislation (EPL), but argues that EPL in Poland is not rigid by international standards, even though the tight legislation for collective dismissal is likely to have played a role in delaying restructuring in large firms. The chapter then looks at the behavior of wages and finds some evidence of real wage resistance in the aftermath of the Russian crisis. However, after reviewing the development of collective agreements, union activity, and delays in wage payments, the chapter argues that real wage rigidity is not likely to be the result of the industrial relations system. If anything, the chapter observes a dramatic decline in the union presence, sustained delays in wage payments, and a decline in the number of collective agreements. Section E summarizes the discussion and presents conclusions. B. EMPLOYMENT PERFORMANCE OVER THE LAST TEN YEARS: THE KEY FACTS 8.6 The dynamics of total employment in Poland over the last ten years is described by two cycles--a first cycle through 1994, and a peak in 1998 (Figure 8.1). The first cycle is in line with the employment dynamics of transition economies, and should not be surprising. The peak in 1998 is more worrying, since it coincides with a dramatic fall in employment that is still ongoing. The total employment changes over these two cycles are impressive, and are certainly visible in Figure 8.1 (left scale) and in Table 8.1. Whereas Poland gained almost 1 million jobs between 1994 and 1998, it lost almost 1.5 million jobs between 1998 and 2002, a period during which unemployment increased by almost 5 percentage points. 176 Figure 8.1: The Dynamics of Total, Private and Public Employment 15500.0 10000.0 Private Employment 9000.0 (right scale) 15000.0 Total Employment 8000.0 (left scale) 14500.0 7000.0 6000.0 14000.0 Public Employment 5000.0 (right scale) 13500.0 4000.0 13000.0 3000.0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Years total employment private sector state sector Source: Poland Statistical Office. Table 8.1: Average Employment by Type of Contract, Ownership Status and Employment Status, 1992-2002 (Average Employment: Thousands of Workers) 1992 1998 1999 2000 2001 2002 Total Employment 15,179.3 15,355.5 14,757.0 14,526.0 14,206.5 13,802.0 By Type of Contract full time 13,502.0 13,760.5 13,182.0 12,950.3 12,748.8 12,312.3 part time 1,677.3 1,594.8 1,574.5 1,575.5 1,457.8 1,489.3 By Ownership Status private sector 7,585.0 9,428.8 9,292.0 9,524.0 9,667.8 9,332.0 state sector 7,594.3 5,926.8 5,465.0 5,002.0 4,538.8 4,470.0 By Employment Status 15,179.7 15,355.3 14,756.0 14,525.8 14,206.8 13,793.0 employees 10,582.0 11,190.5 10,782.0 10,546.5 10,226.0 9,882.0 elf-employed 3,584.7 3,436.0 3,338.0 3,255.0 3,235.8 3,140.3 unpaid family workers 1,013.0 728.8 636.0 724.3 745.0 770.7 Source: Poland Statistical Office. 8.7 The composition of both private and public employment sheds an important light on the current employment situation (Table 8.1). Public employment fell throughout the decade (as should be the case with a transition economy), even though its rate of change increased during the 177 last four years, as Poland undertook important public sector restructuring which was clearly delayed at the early stage of the transition. Cumulatively, public sector jobs account for the full employment losses since 1998, since total public employment fell from some 6 million jobs in 1998 to 4.5 million jobs in 1998. This clearly suggests that private employment was stagnant, and did not experience any net job creation for four years. This is certainly surprising, and deserves a careful policy analysis, which is undertaken in part in Section D. The contribution to Employment Growth by ownership, contract type, and employment status is given in Table 8.2. 8.8 Because of the large employment losses, the distance from the Lisbon target is widening over time (Figure 8.2). Through 1994, the employment rate was in the 65 percent region, in reasonable distance from the EU Lisbon targets. Since 1998, the employment rate has fallen by some 8 percentage points, dropping to 58 percent in 2001. While the dramatic employment losses bear most of the responsibility, the fall in the employment rate is also partially accounted for by a sizable supply shock, as the baby boomers of the early 1980s enter the working age population. 8.9 Looking at the employment rates by age and gender, it is clear that the employment losses hit every age and gender group in the economy (Table 8.3). Nevertheless, some interesting differences emerge. In terms of gender, it seems that male workers were hit disproportionately. The female employment rate was in line with the 60 percent Lisbon target in 1992 but fell to some 53 percent a decade later. The cumulative male loss was even greater, and reached 10 percentage points. When we look at differences in terms of age groups, other differences emerge. The youth employment rate collapsed, losing more than 10 percentage points in a decade. While this was partly due to the entrance of the baby boomers in the labor force, it obviously signals a dramatic hiring freeze. While prime age workers contained their losses, which amounted to 5 percentage points, substantial losses were suffered by the older workers. Indeed, the hump shape age employment profile is a typical phenomenon of high unemployment European countries. 8.10 We next look at employment growth by broad economic sectors. Up to 1998, Poland followed the employment dynamics typical of a transition economy, with fast job creation in the service sector, mild employment changes in industry--and sustained job cuts in agriculture (Table 8.4). The last sector accounted for a remarkable share of total employment in the early phase of the transition, with some 24 percent of total jobs. After 1998, at the time of the Russian crisis, not only was there a sustained period of job cuts, but there was also a reduction in the reallocation of jobs out of agriculture toward the service sector (Figure 8.3). Indeed, the share of jobs into agriculture rose from 18 to 20 percent in the last four years, and such a phenomenon is likely to mask an increase in hidden unemployment, with people going back to agriculture as a residual protection against poverty. 178 Figure 8.2: Polish Distance from Lisbon Target, 1992-2001 75.0 70.0 Lisobn Target Male Empl. Rate Distance from 65.0 Lisbon Total Empl. Rate 60.0 Female Empl. 55.0 Rate 50.0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Source: Poland Statistical Office. Table 8.2: Contribution to Employment Growth by Ownership Structure, Type of Contract and Employment Status, 1998-2002 1998 1999 2000 2001 2002 Total Employment Growth 1.18 -3.90 -1.57 -2.20 -2.85 By Ownership Structure Contr. Private Employment 2.32 -0.89 1.57 0.99 -2.36 Contr. Public Employment -1.15 -3.01 -3.14 -3.19 -0.48 By Employment Status Contr Employees 2.05 -2.66 -1.60 -2.21 -2.42 Contr Self-employed -0.47 -0.64 -0.56 -0.13 -0.67 Contr Unpaid -0.40 -0.60 0.60 0.14 0.18 By Type of Contract Contr. Full Time 1.21 -3.77 -1.57 -1.39 -3.07 Source: Poland Statistical Office. 179 Table 8.3: Employment Rate by Age and Gender, 1992-2001 1992 1997 1998 1999 2000 2001 Total 65.6 64.3 64.5 61.9 60.2 58.3 By Gender Male 71.3 70.2 70.2 67.1 65.1 62.8 Female 59.7 58.1 58.6 56.4 55.0 53.4 By Age 20-24 53.2 49.1 49.0 45.1 41.6 37.1 25-29 68.8 71.5 72.9 69.7 68.7 66.9 30-34 75.1 76.0 77.1 75.5 73.9 73.4 35-39 79.3 78.4 79.9 78.5 76.4 74.3 40-44 79.9 79.0 79.0 76.1 75.1 74.0 45-49 76.3 75.0 75.1 72.8 71.4 69.6 50-54 64.2 63.6 64.3 61.1 59.9 58.2 55-59 42.4 42.3 41.0 37.9 37.0 38.1 60-64 27.6 25.2 24.0 21.7 20.4 19.6 Source: Poland Statistical Office Figure 8.3: Employment Reallocation across Broad Economic Sectors, 1994-2002 0.55 0.50 Services 0.45 0.40 0.35 Industry 0.30 0.25 Agriculture 0.20 0.15 1994 1995 1996 1997 1998 1999 2000 2001 2002 Source: Poland Statistical Office. 180 Table 8.4: Employment Structure and Employment Growth by Broad Economic Sector 1994 1995 1996 1997 1998 1999 2000 2001 2002 Sectoral Composition agriculture 3,508.5 3,768.8 3,303.2 3,106.8 2,917.8 2,659.2 2,718.9 2,714.7 2,693.1 industry 4,695.5 4,559.3 4,747.1 4,857.0 4,927.7 4,627.7 4,486.8 4,337.7 3,967.1 services 6,454.0 6,462.9 6,918.2 7,213.0 7,509.9 7,470.1 7,320.3 7,154.1 7,141.8 Total 14,658.0 14,791.0 14,968.5 15,176.8 15,355.5 14,757.0 14,526.0 14,206.5 13,802.0 Shares agriculture 0.24 0.25 0.22 0.20 0.19 0.18 0.19 0.19 0.20 industry 0.32 0.31 0.32 0.32 0.32 0.31 0.31 0.31 0.29 services 0.44 0.44 0.46 0.48 0.49 0.51 0.50 0.50 0.52 Contribution Growth agriculture - 1.78 -3.15 -1.31 -1.24 -1.68 0.40 -0.03 -0.15 industry - -0.93 1.27 0.73 0.47 -1.95 -0.96 -1.03 -2.61 services - 0.06 3.08 1.97 1.96 -0.26 -1.01 -1.14 -0.09 Total - 0.91 1.20 1.39 1.18 -3.90 -1.57 -2.20 -2.85 Source: Poland Statistical Office. 8.11 Table 8.5 reports the contribution to growth of 14 economic sectors and shows that in the last few years negative employment performance was observed across most sectors. Indeed, 11 of the 14 sectors recorded in the Table 8.5 sectors experienced negative growth, the only exceptions being defense, health, and real estate. Further evidence of the reduced job reallocation of the last few years can be seen in the sizable fall in the coefficient of variation, which fell from 0.5 in 1995 to 0.13 in 2001. Table 8.5: Contribution to Employment Growth by Sectors, 1995-2002 1995 1996 1997 1998 1999 2000 2001 2002 Agriculture -1.16 -0.24 -1.31 -1.24 -1.68 0.40 -0.03 -0.15 Mining 0.04 -0.19 -0.15 -0.11 -0.41 -0.16 -0.13 -0.05 Manuf 0.09 0.05 0.33 0.15 -1.02 -1.00 -0.49 -1.68 Electric, gas, water supply 0.23 0.04 0.01 -0.06 -0.11 0.12 0.04 -0.07 Construction -0.05 0.15 0.55 0.49 -0.42 0.08 -0.45 -0.81 Trade and repair 0.73 0.55 0.64 0.90 -0.22 -0.34 -0.25 -0.37 Hotel, restaurant 0.21 0.05 0.05 0.08 0.00 0.12 0.09 0.04 Transport, storage, communication 0.42 0.22 0.30 0.15 -0.41 -0.01 -0.29 -0.08 Financial intermediation -0.09 -0.07 0.17 0.31 0.19 -0.05 -0.28 -0.15 Real estate & business 0.62 0.50 0.18 0.17 0.27 0.17 0.66 0.29 Public admin and defense 0.04 0.41 0.23 -0.03 -0.03 -0.05 -0.10 0.36 Education 0.01 -0.21 -0.19 0.25 0.33 -0.11 -0.40 -0.30 Health & social 0.28 0.32 0.23 -0.01 -0.17 -0.61 -0.24 0.17 Growth of other services -0.46 -0.38 0.34 0.13 -0.21 -0.13 -0.33 -0.06 Total Growth of Employment 0.91 1.20 1.39 1.18 -3.90 -1.57 -2.20 -2.85 Standard Deviation 0.46 0.29 0.47 0.46 0.53 0.35 0.29 0.51 Coeff. Variation 0.5103 0.24057 0.33667 0.39211 0.13604 0.22371 0.1334 0.18042 Source: Poland Statistical Office. 181 8.12 Finally, employment performance is analyzed in terms of skill composition (Table 8.6). Two facts emerge. First, as of 1998, employment losses were no longer concentrated in jobs assigned to unskilled workers (with primary or less than primary education) but concerned also workers with relatively high skills (secondary education). Second, skilled jobs continued to grow throughout the decade at a sustained pace, even after the change in regime that occurred in 1998. Nevertheless, despite the strong absolute growth of skilled jobs, their relative importance is still small, since skilled jobs account for just 5 percent of the employment pool. Table 8.6: Structure of Employment and Employment Growth by Educational Attainment, 1992-2002 Structure of Employment by Educational Attainment 1992 1994 1996 1998 1999 \1 2000 2001 2002 Employment Structure Primary or less 8,718.0 8,176.3 8,177.3 7,974.5 3,643.3 7,096.5 6,831.5 6,376.0 Secondary 4,942.0 4,903.8 5,081.5 5,461.0 2,746.8 5,414.3 5,263.5 5,237.0 Tertiary 1,519.3 1,577.8 1,710.0 1,920.5 988.3 2,014.8 2,111.8 2,188.7 Total 15,179.3 14,657.8 14,968.8 15,356.0 7,378.3 14,525.5 14,206.8 13,801.7 Shares Primary or less 0.57 0.56 0.55 0.52 0.49 0.49 0.48 0.46 Secondary 0.33 0.33 0.34 0.36 0.37 0.37 0.37 0.38 Tertiary 0.10 0.11 0.11 0.13 0.13 0.14 0.15 0.16 Contribution to Growth Primary or less -2.11 0.17 -0.86 -4.48 -1.29 -1.82 -3.21 Secondary 0.27 0.54 1.15 0.21 -0.54 -1.04 -0.19 Tertiary 0.26 0.49 0.90 0.36 0.26 0.67 0.54 Total -1.59 1.20 1.18 -3.90 -1.57 -2.19 -2.85 1/ Data refer to first and fourth quarter only. Source: Poland Statistical Office. C. WHY IS THE DISTANCE FROM LISBON INCREASING? 8.13 Having documented the fact that Poland's distance from the Lisbon employment target is increasing, the chapter asks its key question: Why is this the case? Is it a growth problem or a problem of the functioning of the institutions? Answering this question is crucial, since policy recommendations are likely to be very different depending on the answer. If the dismal employment performance is a slow growth problem, the policymakers should focus on the determinants of the growth process, with labor market institutions not likely to be high on the list. Conversely, if low employment expansion is not a growth problem, focusing on labor market institutions is more likely to be important. As is shown immediately below, the basic empirical evidence suggests that Poland's problem is not likely to be a growth problem. 182 Poland: A Case of Job-less Growth 8.14 In absolute terms, GDP growth performance in Poland has been positive. Since 1998, the year in which the sustained employment losses took place, Poland registered an average quarterly GDP growth above 3 percent, reaching 6 percent in the year 2000 and falling to 2 percent in 2001. During the same period, quarterly employment growth was always negative (Figure 8.4). Perhaps surprisingly, it can be seen that GDP growth and employment growth are negatively correlated over the last five years--a phenomenon that in the macro labor literature is called "job- less growth." Figure 8.4: Quarterly Growth Rate in Industrial Production and Unemployment Rate, 1992-2002 20 20 unemployment rate 18 15 16 14 10 12 5 10 8 0 6 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 4 -5 Russian Crisis quarterly growth rate of industrial 2 production -10 0 Source: Poland Statistical Office. 8.15 Continental European countries over the late 1980s and early 1990s suffered from job- less growth. During the 1980s, countries such as Italy, Spain, and Germany featured a sustained growth of the order of 3-4 percent per year but simultaneously experienced employment losses and a dramatic increase in unemployment. Caballero and Hammour (1997) analyzed such experience in some detail and argued that changes in labor market institutions observed in Europe over the 1970s were responsible for the labor market dynamics observed throughout the 1980s. 183 8.16 Interestingly enough, the experience of Europe over the last five years changed dramatically. Figure 8.5 reports average GDP and employment growth over the last five years for the EMU-11 countries. Two observations emerge. First, this GDP growth in EMU countries over the last five years has been obviously mild, reaching an average level of 1.5 percent. Second, during the same period European countries experienced sustained employment growth, with an average of 1 percent per year. In other words, GDP growth and employment growth are currently positively correlated in Europe, and many scholars argue that euro zone countries are undergoing a period of growth-less job creation. In light of the European experience, it is obvious that Poland's performance in the labor market is worrisome. Figure 8.5: GDP and Employment Growth in Poland and EMU Countries, 1993-2002 8 gdo growth productivity growth 6 4 gva 2 lab prod wp 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 employment growtj -2 -4 Source: Poland Statistical Office. 8.17 The current situation in Poland can also be seen in Figure 8.4, where we report the growth rate of industrial production and the unemployment rate over time. Industrial production is certainly a good short term indicator for the state economy. Indeed, at a time corresponding to the Russian crisis, which took place in August 1998, we observe a dramatic negative aggregate shock, with a fall in the growth rate of industrial production from some 10 percent to a negative 5 percent. At the same time unemployment started to rise. Yet it is clear that the economy recovered fairly rapidly from the aggregate shock. Nevertheless, unemployment continued to rise (and employment to fall) during the subsequent three years. In 2002, conversely, Poland experienced a further contradictory period. The nature of the more recent slowdown is somewhat different, since it can only partly be blamed on exogenous shocks and on the worldwide economic slowdown. Aggregate investment, which grew very rapidly during 1998 and 1999, contracted sharply from the year 2000 (featuring a 10 percent drop in 2001), and eventually took the 184 economy into recession. Consumption growth, conversely, remained positive throughout, but at a modest 2 percent in 2001. 8.18 As a side product of the job-less growth, the economy experienced substantial gains in average productivity. As shown in Figure 8.6, between 1998 and 2000 productivity growth in Poland rose from 4 percent to 6 percent, mainly as a result of aggregate employment losses. In addition, however, it should be remembered that the reallocation of employment toward high skilled jobs, outlined in Table 8.5, certainly contributed to the productivity gains. Figure 8.6: Productivity Growth in Poland, 1993-2002 8 gdp growth productivity growth 6 4 employment growth 2 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 -2 -4 Source: Poland Statistical Office. D. THE ROLE OF LABOR MARKET INSTITUTIONS 8.19 If GDP growth is not likely to be the key smoking gun for an understanding of the dismal employment growth, then a good question is whether labor market institutions could be responsible for the slowdown. Among the labor market institutions that affect prices and quantities in the labor market, at least two have a potential direct impact on labor demand: Employment Protection Legislation and wage setting. The two institutions are analyzed in turn, and the discussion asks first, whether such institutions represent an obvious obstacle to employment growth, and second, whether any institutional change can be observed around 1998, the year in which the key change was seen in the employment growth process. 185 Employment Protection Legislation 8.20 Employment Protection Legislation (EPL) is a multidimensional institution that refers to the set of laws, rules, and regulations aimed at restricting employers' freedom to adjust labor. The analysis of EPL received considerable attention from macroeconomists, labor economists, and policymakers in the 1990s, since rigid EPL in Europe was seen as an obstacle to labor demand and as partly responsible for the job-less growth disease observed in most European countries. 8.21 The accumulated empirical evidence and theoretical analysis have greatly improved the overall understanding of the effects of EPL on the aggregate labor markets. The main empirical regularities are as follows. EPL reduces unemployment inflows and outflows, but it has ambiguous effects on aggregate employment stocks. In addition, EPL has important effects on the composition of employment, since countries with stricter EPL are associated with higher youth unemployment and higher self-employment. These empirical regularities, recently surveyed by the OECD (1999), are broadly in line with the existing theoretical models (Bentolila and Bertola [1990] and Bertola [1999]). 8.22 Assessing the strictness of EPL in absolute terms is not an easy task, since it is difficult to judge any set of rules without an obvious benchmark. The OECD (1999) recently measured the strictness of EPL in member countries in quantitative terms, and offers a set of indexes aimed at assessing the multidimensional aspects of EPL. In fact, to judge the overall strictness of EPL it is necessary to judge the regulations concerning individual dismissal, temporary work, and collective dismissal. Riboud et al. (2002), in their recent assessment of employment protection legislation in Poland, have elaborated a set of EPL indexes for 10 Eastern European countries. As far as Poland is concerned, the information used by the World Bank was drawn from the OECD study, and the information in that study is fully consistent with the data presented in the current report. 8.23 To obtain an index for the overall strictness of EPL, it is necessary to first consider the strictness of three sub-measures that refer to the strictness of regular employment, fixed term employment, and collective dismissal. Table 8.7 presents these three sub-indexes for the Czech Republic, Hungary, and Poland. As is clear from the table, each index is the average of various dimensions, and to each dimension the OECD assigns a score referring to the strictness of that particular dimension. For example, considering the strictness of EPL for regular employment requires considering the amount of advance notice (in terms of weeks) and the size of the severance payments to be given to individual workers dismissed through no fault of their own. In addition, the index considers the regular inconveniences and the difficulties of dismissal. These three sub-indexes yield the overall index, whose value is reported in the top part of Table 8.7. A similar deconstruction is carried out for the other two sub-indexes, in the middle and lower parts of the table. The information in the table refers to the late 1990s. 186 Table 8.7: Employment Protection Legislation in Poland, the Czech Republic and Hungary OECD Ranking Strictness of Employment Protection for Regular Employment Difficulty of Regular Notice and Severance Dismissal Overall Inconveniences Payment for no-fault Strictness Czech Republic 18 24 19 24 Hungary 16 14 11 11 Poland 22 8 11 12 Strictness of Employment Protection for Fixed Term Employment Temporary Fixed Terms Work Overall Agencies Strictness Czech Republic 5 1 5 Hungary 6 24 21 Poland 7 11 10 Collective Dismissal Additional Definition of Additional Delays Other costs Overall Overall of Collective Notification in days Strictness Czech Republic 4 2 83 0 4.3 26 Hungary 3 2 47 0 3.4 18 Poland 3 1 26 2 3.9 22 Source: OECD. 8.24 The last column in Table 8.7 gives an index of overall strictness in terms of OECD ranking. Out of 27 OECD countries, Poland falls in the middle range, much lower than the ranking of the European countries with the most rigid EPL, such as Italy and Portugal, but also lower than the ranking of countries such as France and Germany. Strictness regarding notice and severance payment is particularly low for Poland. In comparison with Hungary and the Czech Republic, the strictness of EPL for regular employment for Poland is similar to that for Hungary, but is much less rigid than that for the Czech Republic. A similar relative position is observed for fixed term employment. In Poland, temporary work agencies are established and fixed term contracts are allowed, even though up to 2001 there was a limitation on the number of renewals. 8.25 The situation is somewhat different for collective dismissal, where the Polish legislation appears definitely more rigid. In terms of absolute ranking, Poland appears among the strictest range of countries. The details of the strictness of EPL for collective dismissal are provided in Table 8.8. As of 1999, the number of workers necessary to initiate a dismissal procedure was 10 percent of the labor force, which becomes a sizable absolute number for large establishments. In addition, the delays imposed on the procedure -- linked to the obligation to discuss alternative solutions with the trade unions -- are substantial (45 days). Arguably, in a transition economy, restructuring firms need a fairly flexible collective dismissal procedure. In some ways flexibility is more important for collective dismissal than for individual dismissals, since large scale restructuring is likely to involve collective dismissals. 187 Table 8.8: Collective Dismissal in Poland, the Czech Republic and Hungary Country Definition of Notification of Notification of Delays involved Type of negotiation Selection criteria Severance pay collective dismissal employee public authorities required representatives Czech Employers planning Duty to inform Notification of Information to trade Consultation on Obligation to take No special Republic to dismiss several competent trade district labor union and PES office 3alternatives to account of social employees for union body. office. months before redundancy and considerations regulations for reasons of structural implementation. measures for finding (e.g. mothers, collective change or new jobs. adolescents, reorganization. disabled). dismissal. Hungary 10+ workers in firms Committee to be Notification of 30 days delay after Consultation on Law lays down No special 20-299 set up, including local employment notification of principles of staff union regulations for Works Council or office. employment office, if reduction, and ways participation, but collective trade union at least 10 persons are to mitigate its effects. no specific dismissal. representatives to involved; selection criteria consult on for dismissal. procedures and benefits. >10% in firms 100- 90 days if 25 % of 299 workforce or 50+ employees are involved. 30+ workers in firms 300+ Poland 10%+ of workers Duty to inform Notification of Information to trade Agreement to be Law lays down 1m <10years of competent trade local employment union and PES 45 daysreached with trade union service; union. office. before union on alternatives participation, but implementation. to redundancy and no specific in firms < 1000 ways to mitigate the selection criteria 2m<20y employees effects. for dismissal. 100+ workers in 3m>20y firms with 1000 employees and above Source: Poland Statistical Office. 8.26 In 2002 the Labor Code was amended and some of the constraints on employment adjustments were removed. In particular, the strictness of EPL for collective dismissal was reduced, as was the regulation applied to small firms. In addition, fixed term contracts, already available in the late 1990s, were further liberalized. Furthermore, several important changes were introduced. First, firms experiencing severe financial problems now have the right to suspend temporarily (for up to three years) the provisions of collective agreements/employment contracts. Second, with reference to small firms, it was agreed that firms employing up to 20 workers will no longer have to prepare written internal regulations governing the conditions of work. Third, the cost of mass layoffs was reduced. Finally, restrictions were lifted on the use of fixed-term contracts, so as to abolish the limit on the number of permissible renewals. In addition, in the same period there was an increase in working time flexibility, so that firms now have the possibility of redistributing total working time hours within a four month period (previously it was three months). Moreover, the cost of overtime was lowered and the limit on overtime work was increased. 188 8.27 To sum up, it is clear that EPL is not overly rigid by international standards, particularly as regards the regulation concerning regular employment and the use of temporary contracts. What is somewhat rigid is the legislation for collective dismissal. Can EPL be considered as one of the main obstacles to employment growth since 1998? Most likely it is not the smoking gun that explains low employment growth in the private sector. However, strict EPL on collective dismissal has probably delayed adjustment in privatized firms, which hoarded labor up to the Russian crisis. And it should be borne in mind that some reforms to the Labor Code have recently been introduced. Real Wage Resistance 8.28 Excessive wage increases may be an alternative explanation for the job-less growth experience of Poland over the last few years. Indeed, aggregate data on wage growth show that sustained real wage growth was also observed in the years in which aggregate employment declined dramatically (Table 8.9). This section examines this mechanism further, and studies the dynamic relationship between wage growth and employment growth at the sectoral level. Table 8.9: Nominal and Real Wage Growth Across Sector, 1997-2000 1997 1998 1999 2000 Nominal Wage Growth Mining and quarrying 19.9 14.1 10.5 9.4 Manufacturing 21.8 14.7 11.8 9.8 Electricity, gas, and water supply 17.8 14.4 11.5 10.9 Construction 25.8 18 12 9.4 Wholesale and retail trade 23.9 16.7 15.2 7.9 Hotels and restaurants 25.7 17.3 10 7.5 Transport, storage, and communication 23.8 18.6 14.4 13.2 Financial intermediation 25.4 19.8 12 20.7 Real estate and business activities 22.9 14.8 14.8 12.5 Public administration and defense 20.5 17.6 13.5 10.3 Education 23.5 14.5 12.4 18.5 Health and social work 21.9 14.9 10.1 9 Cpi Inflation 14.9 11.8 7.3 10.1 Real Wage Growth Mining and quarrying 5 2.3 3.2 -0.7 Manufacturing 6.9 2.9 4.5 -0.3 Electricity, gas, and water supply 2.9 2.6 4.2 0.8 Construction 10.9 6.2 4.7 -0.7 Wholesale and retail trade 9 4.9 7.9 -2.2 Hotels and restaurants 10.8 5.5 2.7 -2.6 Transport, storage, and communication 8.9 6.8 7.1 3.1 Financial intermediation 10.5 8 4.7 10.6 Real estate and business activities 8 3 7.5 2.4 Public administration and defense 5.6 5.8 6.2 0.2 Education 8.6 2.7 5.1 8.4 Health and social work 7 3.1 2.8 -1.1 Cpi Inflation 14.9 11.8 7.3 10.1 Source: Poland Statistical Office. 189 8.29 In most models of the labor market, wage growth and job creation (or employment growth) in the private sector are variables that are jointly endogenously determined by labor demand and by the wage function. In general, equilibrium wage growth is positively correlated to productivity growth and to changes in variables that affect the worker's position at the bargaining table, such as unemployment benefits, union strength, etc. Job creation, or employment growth, the other endogenous variable, is also correlated to productivity growth, but it is negatively correlated to the variables that affect the worker's bargaining position and outside option. Thus, if real wages are flexible and we have data on employment growth at the sectoral level, we would expect that the sectors that experience larger productivity growth would simultaneously experience larger employment growth and larger wage growth. As a result, if we have data on employment growth and real wages, we should observe a positive relationship between real wage growth and employment growth. These are exactly the data used in Figure 8.7. 8.30 Interestingly enough, in 1997 and 1998 employment growth and real wage growth were positively correlated across sectors, as is clearly shown in the two top panels of Figure 8.7. In some way, the relation between employment and real wage growth was in line with an equilibrium situation in the labor market, with employment and real wages growing faster in sectors with greater productivity growth. But what would happen to the equilibrium in the labor market, in terms of wage growth and employment growth, if a negative productivity shock were to hit the economy in a manner similar to the recession in Poland in 1998? If real wages were flexible, we would expect that a similar positive relationship would also hold during recessions, so that the points in the two top panels in Figure 8.7 would shift down along the 45 degree lines. This does not seem to have happened in Poland, as shown in the figure. In the two bottom panels of Figure 8.7, the relationship between employment growth and wage growth becomes much less clear over time, since positive wage growth continues to be observed also in sectors that experience sizable employment losses. We can label this phenomenon real wage resistance. In other words, as the recession hit Poland in 1998, real wages continued to grow in several sectors. 190 Figure 8.7: Wage Growth and Productivity Growth across Sectors 1999 Wage Growth (vert. Axis) versus employment growth 1998 Wage Growth (vert. Axis) versus employment growth (horiz. (horiz. Axis) Axis) 9 9 8 8 7 7 6 6 5 5 4 4 3 3 2 2 1 1 0 0 -0.1 -0.05 0 0.05 0.1 0.15 0.2 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 1999 Wage Growth (vert. Axis) versus employment growth 2000 Wage Growth (vert. Axis) versus employment growth (horiz. Axis) (horiz. Axis) 9 12 8 10 7 8 6 6 5 4 4 2 3 0 2 -0.1 -0.05 0 0.05 0.1 1 -2 0 -4 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 Source: Poland Statistical Office. 191 Collective Bargaining and Industrial Relations 8.31 Is the real wage resistance observed in the previous section linked to the structure of collective bargaining in the Polish economy? A complete answer to this question is obviously difficult to provide, and may need specific research with data at the establishment level. Nevertheless, as a starting point for answering this question, we can review the structure and development of industrial relations in Poland. Overall, we can argue that developments in industrial relations, union activity, and collective agreements are not consistent with strong real wage rigidity in the aftermath of the 1998 recession documented in the previous section. Conversely, we can observe evidence of substantial delays in wage payments, a decline in the numbers of collective agreements, substantial delays in renewing labor contracts, and a dramatic collapse in union density. Other mechanisms, which should be explored in future research, are likely to be responsible. 8.32 The political and economic transformation in 1989 led to substantial changes in Polish industrial relations, and the structure of collective agreements has since then been framed in a way similar to the European standard, with specific laws regulating the activity of trade unions, and employers' organizations, and the settlement of collective disputes. This process led to the 1994 reform of the Labor Code--a piece of legislation that has existed in Poland since 1974. With the 1994 reform, collective agreements became the key instrument for establishing labor contracts in Poland, with an important distinction being made between multi-establishment and single establishment collective agreements. While collective agreements became the center of labor market negotiation, it was decided that a few dimensions (such as employment protection legislation, maternity and childcare leave, collective dismissals) could not be negotiated within collective agreements. As was discussed above, some of these institutions were further modified in 2002. 8.33 Since 1994, information on collective agreements has been centrally recorded. Data from the State Labor Inspectorate show that by the first half of 2001 a total of over 11,000 collective agreements had been concluded (over 9,000 of which were still in force at the end of the year). The large majority of agreements are firm level agreements, an institutional framework which can easily accommodate wage differentials across firm sectors. Quantitatively, after an initial bulk of registrations that took place following the institutional changes, the aggregate number of registrations slowed down. But beyond this natural slowdown there are many collective agreements (albeit expired) that have not been renewed. The number of multi-establishment collective agreements is much smaller and amounts to just around 100, with most contracts referring to the public sector and the budgetary sector. 8.34 The effectiveness of collective agreements as an institution for regulating labor relations is called into question when we take into account the fact that employers appear to easily breach the provisions of such agreements. In 2001, 50 percent of the inspections showed irregularities from the firm's standpoint, which did not comply with the agreed terms (193 inspections were reviewed). 8.35 Surveys of private sector enterprises in Poland have also found a growing tendency among employers to delay the payment of wages in recent years. For example, in 2001 the State Labor Inspectorate (Pastwowa Inspekcja Pracy) found that nearly two-thirds of the companies audited breached regulations in the area of timely payments. While evidence of late payment was initially restricted to small firms, it has recently also become common in large enterprises. In 2001, the State Labor Inspectorate carried out more than 1,700 audits of enterprises employing a total of 155,000 persons. The inspectors found that pay regulations were violated in 62 percent of 192 the audited enterprises. The average back wages amounted to PLN 1,000 per employee, with a growth of 34 percent over the previous year. A large share of employers (45 percent) reported that cash flow problems represent the most important reason for wage arrears. Overall, it appears that delays in the payment of wages have become increasingly common. In the 115 defaulting enterprises analyzed to date, wages have been temporarily suspended in 66 percent of the cases, and not paid at all in 9 percent. Recently privatized enterprises and health care institutions (hospitals and day care clinics) are strongly represented on the list of defaulting employers. This bulk of evidence on delays in wage payments and wage arrears suggests that real wages are probably more flexible than reported in official data. 8.36 Employers' organizations are a rather recent element in industrial relations in Poland. These organizations appear to have an important institutional role at the aggregate level, where employers have become involved in a dialogue with state authorities and trade unions (see the discussion in the next paragraph on the Tripartite Commission). Conversely, at the industry/regional level and at the enterprise level, employers' organizations are less visible and rarely involve themselves in bipartite dialogue with trade unions. There are two reasons for this phenomenon. First, employers are still relatively weakly organized. Second, at the enterprise level most of the bargaining power rests with the employer, who does not feel the need to join larger employer organizations. 8.37 Poland's Tripartite Commission for Social and Economic Issues was established as a forum for national social dialogue in 1994, under a "State Enterprise Pact" involving the government and the trade unions. Initially, the weakness of the employers' representation (especially of the private sector employers) and the conflict between the NSZZ Solidarno and OPZZ trade unions were major obstacles to the work of the Tripartite Commission. This conflict and the involvement of the trade unions in direct political activity hampered the work of the Commission for some years. Since 2001 the present left-wing coalition government has introduced new legal regulations and has revitalized the Commission. Despite some problems, the body is now functioning relatively well. Decline of Trade Union Density 8.38 The last and most important element to consider in the Polish industrial relations system is the trade unions. Over the past 20 years, trade union density in Poland dropped from 80 percent of the workforce to 14 percent in 2002. As is shown below, this sharp decrease in density has been influenced by a number of economic and political factors. As a result, Poland, whose trade unions played a highly important role in the 1980s, is now characterized by much weaker unions than most other European countries. 8.39 A number of concomitant factors account for these facts. First, the role and authority of the most important union, Solidarnosc (an institution that played a key role in the collapse of communism) has contributed to the observed decline. Indeed, at the early stages of the transition, the union leaders argued that Solidarnosc should no longer increase its membership, since a strong concentration of industrial workers could reduce the incentives to undertake key structural reforms. Second, the Polish economy experienced the same factors that led to the erosion of trade unions in many other European countries, notably, a sharp drop in employment in those branches of industry that traditionally had the highest number of trade union members (mining, metalworking, the machine industry, etc.) (Boeri et al. 2001). Third, the number of workers employed in small and medium-size enterprises and in part-time work began to grow. As in most industrialized countries, a union presence in small to medium enterprises is much less important. 193 8.40 Data collected by CBOS toward the end of 2001 (CBOS, 2002) show that, in late 2001, 7.9 percent of the Polish adult population belonged to a trade union, with the figure standing at 14.1 percent among employed workers. The highest trade union density is observed in mining (43.8 percent), transport (27.3 percent) and education (27.5 percent), with lowest density observed in agriculture (3.5 percent) and building (3.6 percent). Among the youth workers, only 2.4 percent belonged to trade unions, and in the 25-29 age group only 6.8 percent of workers belonged to trade unions. Among the 50 and above age group, 17.0 percent were union members. 8.41 Gardawski et al. (1998) argue that trade unions have been eroded as companies have become increasingly smaller and privatization has taken place. Examining companies in terms of the number of workers employed and the type of ownership, Gardawski et al. found that all large state-owned enterprises (with over 250 employees) had trade unions present (two unions, on average). Trade unions were present in 75 percent of state-owned medium-size companies (50- 250 employees), and in 50 percent of small enterprises (under 50 employees). The situation was somewhat different for privatized companies. Although unions were present in all large enterprises, unions were almost totally absent in small enterprises. Finally, the situation in new private companies was the most unfavorable for unions. A trade union was present in only 5 percent of large private companies. E. CONCLUSIONS AND POLICY IMPLICATIONS 8.42 In this chapter we have looked at Poland's performance in terms of aggregate employment growth over the last ten years, focusing on Warsaw's distance from European targets in terms of its employment rate, currently set at 70 percent for the year 2010. We have shown that since 1998 the Polish economy has experienced a fall of 8 percentage points in its aggregate employment rate, which was accompanied by a collapse in employment across different age groups and genders, and most skill groups. As a result, Poland's distance from European employment targets is continuously increasing through a prolonged period of "job-less" growth, an aggregate phenomenon experienced by most European labor markets in the early 1990s. 8.43 The discussion has looked at the role of labor market institutions as a possible cause of this aggregate outcome, and has analyzed in some detail the Employment Protection Legislation (EPL) and the industrial relations system. The analysis has shown that the existing EPL legislation for individual dismissal is not rigid by international standards, even though the legislation on collective dismissals has been traditionally tight and may have contributed to the delay in restructuring in large firms, which probably hoarded labor up to the 1998 recession. 8.44 The discussion has found some evidence of real wage resistance in the aftermath of the 1998 recession, since positive real wage growth was also observed in sectors that experienced a marked reduction in employment growth. Our analysis suggests that such resistance does not appear to be the result of excessive union pressure and ill-functioning collective agreements. If anything, we observe a dramatic decline in the union presence, sustained delays in wage payments, and a decline in the number of collective agreements. 8.45 From here, research should branch out in several directions. First of all, careful further research should be carried out on the role of labor market institutions. Notably, it is important to assess whether other labor market institutions, such as payroll taxation, minimum wages, and unemployment benefits, could account for the worrisome job-less growth of the Polish economy. Recent work by Riboud et al (2002) suggests that payroll taxation is indeed quite high by international standards, and may certainly be responsible for a depressed labor demand. Yet no substantial changes in the structure of taxation has taken place since 1998, and it remains difficult 194 to rationalize the change in regime observed after 1998. Other labor market institutions, notably the minimum wage, are not particularly tight by international standards. Conversely, the structure of unemployment benefits is likely to have contributed to the recent increase in unemployment. This mechanism is emphasized by Boeri and Garibaldi (2003) in a paper that looks at the labor supply dimensions of the current increase in unemployment. The Labor Market Effects of Partial Deregulation in the Goods Market 8.46 It is important that future research should also look at the role of deregulation in the goods market. Over the last four years the flow of privatization in Poland has gained momentum, as is reflected in the reduction in the total number of state-owned enterprises, from more than 3,300 in 1997 to some 2,000 in 2001. Much remains to be done, but there is certainly more restructuring in former state-owned firms, which hoarded labor in the early years of the transition, as witnessed by the estimate provided by Gora in OECD (1993). And what are the labor market consequences of deregulation in the labor market? Blanchard and Giavazzi (2003), in a recent paper, have offered insights into the labor market effects of goods market deregulation that are likely to be relevant for Poland. Goods market deregulation, in one form or another, leads to the entry of new firms and to the reduction in overall firms' mark-ups, with favorable effects on workers, since lower mark-ups lead to an increase in real wages and an increase in employment. But the positive effects of deregulation on employment work mainly in general equilibrium, since they take place despite a reduction in rents at incumbent firms. Indeed, in general equilibrium the reduction in prices more than offsets the reduction in rents in incumbent firms, with positive effects on employment. 8.47 But the situation becomes more complicated if deregulation affects only part of the economy, for example, because the rest of the system remains heavily regulated. In this case, employment in existing firms may fall, with an adverse effect on unemployment. This mechanism can partially rationalize the sluggish employment performance of Poland and deserves to be studied in detail. 8.48 At the policy level, our analysis of labor market institutions, and employment protection legislation in particular, suggests that no dramatic policy changes are required. Recent changes in the Labor Code are welcome and point in the right direction. In general, policies that reduce the strictness of collective dismissal procedures are likely to be beneficial to Poland, where the restructuring and deregulation of large state-owned enterprises is taking place. As far as the industrial relations system is concerned, Poland features a fairly decentralized wage bargaining system and an institutional set-up that is appropriate for accommodating firm-specific shock. As Poland moves into the European Union, a flexible wage system can certainly be beneficial. In this respect, we believe that an aggregate form of negotiations, involving the Tripartite Commission, may certainly be useful for implementing structural reforms without social tensions, as was the case with the pension reform to a large extent, but it should not necessarily be used for nationwide wage agreements. Pressures to coordinate at the pan-European level after entry into the EU would also be stronger if negotiations were centralized nationwide. 8.49 Finally, inasmuch as the current labor market situation in Poland is the result of partial and incomplete deregulation in the goods markets (an interpretation which deserves further research), it is clear that policy actions should help Poland to fully "cross the river," so that the country can reap all of the benefits of market deregulation. 195 REFERENCES Bedi, A. (2000), "Sector Choice, Multiple Job Holding and Wage Differentials: Evidence from Poland," Journal of Development Studies, Vol. 35 (1): 162-179. Bentolila, S. , and G. Bertola (1990), "How Bad Is Eurosclerosis," Review of Economic Studies, Vol. 57, pp. 381-402. Bertola, G. (1999), "Microeconomic Perspectives on Aggregate Labor Markets," in Handbook of Labor Economics, ed. O. Ashenfelter and D. Card, North Holland, pp. 2985-3028. Blanchard, O. and F. Giavazzi (2003), "The Macroeconomic Effects of Labor and Product Market Deregulation," forthcoming Quarterly Journal of Economics. Boeri, T., and P. Garibaldi (2003), Dealing with an Increasingly Stagnant Unemployment Pool, mimeo, World Bank and Bocconi University. Boeri, T., A. Brugiavini, and Lars Calmfors (2001), Trade Unions in Europe, Oxford University Press. Boeri, T., L. Calmfors, and A. Brugiavini (2001), "What Unions Do to the European Welfare State," in The Role of Unions in the New Millenium, Oxford University Press. Caballero, R., and M. Hammour (1997), "Jobless Growth: Appropriability, Factor Substitution and Unemployment," Carnegie-Rochester Conference Series on Public Policy, Vol. 48, Issue 1, pp. 51-94 (with M. Hammour), Elsevier Science. CBOS (2002), Composition of Trade Unions Towards the End of 2001, CBOS Centrum Badania Opinii Spolecznej, Warsaw, 2002. Available on line at www.cbos.pl. Churski, P. (2002), "Unemployment and Labor Market Policy in the New Voivodship System in Poland," European Planning Studies, Vol. 10, No. 6. Czyesky, A. (2002), Economic Growth and Labor Demand, paper presented at the 22nd NBP Conference on "Monetary Policy in an Environment of Structural Changes." Gardawski, J., B. Gaciarz, A. Mokrzyszeswki, and W. Pankov (1998), Collapse of the Rampant Trade Unions in Privatized Economy, IPA & Ebert Foundation, Warsaw. International Monetary Fund (2002a), "Republic of Poland: Selected Issues and Statistical Appendix," IMF Country Report 02/128. International Monetary Fund (2002b), "Republic of Poland: 2002 Article 4 Consultation, Staff Report." IMF Country Report 02/127. Jackson, J., and B. Mach (2002), Job Creation, Destruction and Transition in Poland 1988- 1998: Panel Evidence, William Davidson Working Paper No. 502. Keane, M., and E. Prasad, (2002), "Changes in the Structure of Earning During the Polish Transition," International Monetary Fund Working Paper 496. 196 Mornilov, E., and B. van Ark (2002), New Estimates of Labor Productivity in the Manufacturing Sector of the Czech Republic, Hungary and Poland, Research Memorandum GD-50. University of Groningen. Newell, A., and F. Pastore (2000), "Regional Unemployment and Industrial Restructuring in Poland," mimeo, University of Sussex. Newell., A., and M. Socha (2002), The Rising Non-Manual Wage Premium in Poland, Discussion Paper 89, University of Sussex. OECD (1993). Employment and Unemployment in Economies in Transition: Conceptual and Measurement Issues, Paris. OECD (1999). Employment Outlook, Paris. Riboud, Michelle, Caroline Sanchez-Paramo, and Carlos Silva-Jaregui (2002). Does Eurosclerosis Matter? Institutional Reform and Labor Market Performance in Central and Eastern European Countries in the 1990s, World Bank SP Discussion Paper No. 0202. World Bank (2001). "Poland's Labor Market: The Challenge of Job Creation," Document No. 23033. World Bank (2003). "Poland: Toward a Fiscal Framework for Growth Report," Document No. 25033. 197 9. DEALING WITH AN INCREASINGLY STAGNANT UNEMPLOYMENT POOL Tito Boeri, Pietro Garibaldi and Mauro Maggioni 9.1 The time profile and the composition of Polish unemployment have been extensively characterized in the literature on transition (Blanchard et al., 1995; Boeri, 1994; OECD, 1994). This literature has pointed out that the rapid buildup of Polish unemployment at the beginning of the 1990s was mainly the by-product of low unemployment outflows rather than of large employment inflows associated with labor shedding in the early stages of transition. A stagnant unemployment pool, it was noted, involves a permanent decline in the labor supply. And unemployment decreased only mildly when Poland exited the so-called transitional recession. An additional problem associated with the spread of long-term unemployment is that it puts pressure on the welfare systems by pushing off of the unemployment compensation rolls those who have exhausted the maximum duration of benefits. 9.2 In Chapter 8 ("How Far Is Warsaw from Lisbon?") we have argued that job losses in the second half of the 1990s were largely concentrated at the lower end of the skills distribution, as Polish employers appeared to take advantage of the recessionary environment to "cleanse" their wage rolls, further reducing the degree of "labor hoarding" inherited from the previous system. In this chapter we analyze the other side of the coin of these labor market developments, namely, whether this new wave of job losses has substantially altered the composition and dynamic features of Polish unemployment. Our key conclusion is that the unemployment pool has become even more stagnant recently than it was in the early 1990s. Among the reasons for this situation are the presence of a large informal sector (which means that a significant portion of registered and even survey unemployment is a statistical error) and the fact that social transfers provided to unemployed individuals at a flat rate, and without taking into account the large cross-regional differences in the cost of living, reduce the incentives to seek a formal job for those who have less marketable skills, notably in rural areas. 9.3 In the following discussion, Section A reviews unemployment dynamics in the last part of the 1990s, while Section B analyzes the evolution of job finding rates and their responsiveness to inflows of new waves of jobseekers as well as vacancy rates. Section C considers the decline in the coverage of unemployment insurance associated with the buildup of long-term unemployment and the pressures this places on other components of the Polish welfare system, which are not tailored to ease the transition back to work. In Section D, some inferences are made as to the role played by labor demand, the presence of a sizable informal sector, and other design features of the Polish social welfare system in making the Polish unemployment pool stagnant. Finally, Section E discusses the potential for Active Labor Market Policies (ALMP) to combat long-term unemployment, while Section F supplies a brief conclusion. 198 A. UNEMPLOYMENT DYNAMICS 9.4 Figure 9.1 characterizes the evolution of Polish unemployment since the inception of the transition. Data are drawn from the Unemployment Register, the only source measuring in continuous time the unemployment inflows and outflows. Figure 9.1: The Evolution of Polish Unemployment Since the Inception of Transition, 1990-2002 250000 20 150000 15 outflow rate and 50000 10 inflow -50000 5 unemployment monthly 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 -150000 0 monthly inflows monthly outflows to job unemployment rate Source: Labor Force Surveys for unemployment rate; Unemployment Register for data on unemployment inflows and outflows. 9.5 Administrative data have a number of drawbacks (see Socha and Weisberg, 1999). Their most serious limitation is that they are affected by changes in the regulations concerning eligibility for and duration of unemployment benefits. However, the regulatory changes that took place in Poland during the 1990s are likely to have only mildly affected the measurement of unemployment inflows and outflows to jobs, the two flows plotted in the figure. Insofar as the reforms reduced the maximum duration of benefits, they may have boosted cancellations from the live register of individuals who did not show up at labor offices after exhausting their benefits. But the recording of inflows and of outflows to jobs should have been mildly affected by policy changes. Clearly, these reforms have affected job search incentives: entitlements effects have been identified in unemployment to employment transitions in Poland (see, for example, Boeri and Steiner, 1998). But this is another issue. We are dealing with measurement here rather than with the effects of reforms on actual flows. 9.6 With the above caveats in mind, two facts highlighted by Figure 9.1 are relevant. First, a key factor behind the decline of Polish unemployment in the 1994-97 period was the pickup of outflows to jobs. The unemployment rate went down dramatically in the 1995-97 period as outflows to jobs almost doubled with respect to their levels in 1993. It should be stressed that not only total outflows to jobs, but also job finding rates (outflows to jobs as a proportion of registered unemployment), increased dramatically over this period (from about 2 percent in 1993 to 5 percent in 1997, on a monthly basis (see Figure 9.2). 199 Figure 9.2: Monthly Outflows to Jobs as a % of Unemployment, 1991-2002 6 5 4 rates wo 3 Outfl2 1 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Source: Unemployment Register. 9.7 At the same time, unemployment inflows were still increasing or were stationary. Thus, the reduction in the jobless rate in the mid-1990s was definitely driven by flows from unemployment to employment. 9.8 Second, a similar development did not occur in the second part of the 1990s and at the beginning of the year 2000. While the rise of Polish unemployment in the late 1990s was again driven by the inflow side, the outflow to jobs failed to pick up in the following years, which explains why registered unemployment, rather than falling after the crisis, reached a new peak in 2002. 9.9 Overall, since the beginning of the transition, inflows into the live register have been on an upward trend, with a marked rise in 1994-95 and 1999. Not only different waves of labor shedding, but also demographic factors, namely, the entry into the labor market of increasingly large cohorts of school-leavers, are likely to have played an important role in this context. Reductions in unemployment have therefore been driven mainly by the outflow side. Job finding rates in particular picked up markedly towards the mid-1990s, contributing to a strong reduction in unemployment. The absorption of the unemployment created in the late 1990s is likely to require, again, a strong increase in flows from unemployment to employment. This would mean a stronger labor demand and also a strong response in the job finding rate to vacancy creation. The next section assesses the changes that have occurred in the matching process in Poland since the beginning of the transition. B. LONG-TERM UNEMPLOYMENT AND THE MATCHING PROCESS 9.10 As shown above, Polish unemployment continued to increase in recent years in spite of the decline in unemployment inflows, because the probability of finding jobs declined. Or, to put it another way, long-term unemployment increased. The risk under such conditions is that a vicious circle is set in motion whereby long-term unemployment induces declines in the effective labor supply, and this reduces the scope of wage adjustment in preventing the absorption of unemployment, which, in turn, means more long-term unemployment, etc. In other words, long- 200 term unemployment can be following a self-fulfilling path. There are theoretical arguments for this: for instance, a long duration of unemployment may be associated with the loss of skills (Pissarides, 1992) and lower search intensity. Alternatively, employers may be tempted to use the length of periods of unemployment as a screening device (Blanchard, 1991), in which case a larger proportion of long-term unemployment in the pool may in itself reduce the outflow rates. To assess the likelihood that these self-fulfilling effects are operating in Poland, we estimate a matching function in which we allow the two pools--long-term and short-term unemployment--to exert a different effect on overall job finding probability. 9.11 In particular, in Table 9.1 estimates of a matching function in Poland are reported for the initial transition years in which the unemployment stock is broken down into two separate poolsthe short-term unemployed and the long-term unemployedand the assumption is made that these two pools are perfectly substitutable up to a "distributional" parameter (w), which embodies the effects of unemployment duration on job finds. In particular, the posited functional form for the matching function is as follows: O=AV [w Ust + (1-w) Ult - - -/ ] where O denotes outflows to jobs, V denotes the stocks of vacancies and Ust and Ult denote the stocks of the short-term and long-term unemployed, respectively. By taking logarithms and using a Taylor expansion around = 0, one then obtains: o = 1+ 2 v + 3 ust+ 4ult+ 5(ust-ult)2 where a lower case denotes logarithms. This equation was estimated with fixed-effects against panel data on 16 macro-regions in the period 1999-2002. Unfortunately, we could not have monthly data on unemployment duration by region. Thus, we proxied short-term unemployment at t by monthly inf1ows into unemployment between t-l and t (i.e., the unemployed having experienced a spell shorter than one month) and obtained the "long-term unemployed" as a residual (the difference between total unemployment at t and proxy short-term unemployment). 9.12 Based on estimates of the parameters 2, 3 and 4, it was then possible to recover the underlying distributional factor w as 3 /(3+ 4) and the elasticity of job finds with respect to total unemployment and vacancies. Clearly, the higher w is the less "employable" are the long- term unemployed with respect to those with unemployment spells shorter than one year. As shown by the table, the coefficient for long-term unemployment is not significantly different from zero at conventional (95 percent) confidence levels. Thus, the estimated value of the distributional parameter is 1. Analogous estimates based on 1992 Polish data broken down by 59 voivodships and with a proper measure of district-level long-term unemployment yielded a distributional parameter of 0.73, that is, significantly lower. 201 Table 9.1: Matching and Unemployment Duration (regression results, with fixed effects) coefficient standard error 95% confidence interval Vacancies 0.09 0.015 0.064 0.126 Short-term unemployment 1.37 0.416 0.555 2.189 Long-term unemployment -0.027 0.028 -0.083 0.029 w (distrib. parameter) 1 nobs 576 R2 0.94 Source: Monthly data on unemployment and vacancy stock and flows by macro-region from the Polish Unemployment Register. 9.13 All in all, the displayed estimates suggest that long-term and short-term unemployment enter the matching function differently in Poland. A change in the composition of unemployment involving more long-term joblessness would seem to reduce significantly the elasticity of job finds with respect to unemployment. Further estimates based on proper duration data by district may yield better (and different from zero) estimates for the elasticity of matching with respect to short-term unemployment. The data at hand suggest that outflows to jobs are highly responsive to an increase in the number of those entering unemployment every month, while those staying longer have hardly any effect on job finds. This negative duration effect may actually be simply the by-product of heterogeneity in the two pools rather than genuine state dependency. Longer durations of joblessness indeed typically involve individuals with personal characteristics which are less marketable (e.g., low educational attainments), as discussed in the next section, where we look at the characteristics of the pool of those with long durations of unemployment. C. CHARACTERISTICS OF LONG-TERM UNEMPLOYMENT 9.14 Table 9.2 provides information on the evolving characteristics of Polish long-term unemployment. In particular, average yearly data from the Labor Force Survey (Poland was the first country in the region to regularly introduce such a survey) in 1992 and 2002 are used to analyze the risk for different socioeconomic groups of falling into long-term unemployment. 202 Table 9.2: The Changing Profile of Polish Long-term Unemployment, 1992-2002 2002 U rate Share in U Share in LTU rate Relative risk of LTU being in LTU POLAND 20.41 11.05 0.541319 Women 20.60 47.64 52.74 10.78 0.523301 Youth 15-24 43.77 27.79 22.82 17.01 0.388622 Primary/low 26.76 20.39 22.90 12.92 0.482810 education Rural area 17.67 44.78 48.08 8.21 0.464629 Families with 34.42 68.63 43.63 15.01 0.436084 children 1992 POLAND 15.97 4.82 0.301816 Women 14.78 50.26 52.19 4.45 0.301083 Youth 15-19 31.15 8.29 7.25 6.20 0.199037 Unskilled 15.84 7.26 8.00 6.23 0.393308 Primary/low 12.08 23.28 23.29 5.26 0.435430 education Source: Average yearly data from Polish Labor Force Survey. 9.15 Data on unemployment in rural areas are not comparable over time, as the LFS in 1992 tabulated the number of inhabitants in the center of residence of each individual, while this information was not provided to researchers (for privacy reasons) for subsequent waves of the survey. Thus, in 2002 we had to define rural employment and unemployment on the basis of other (less precise) indicators, namely, the characteristics of the macro-region in which the individual was living. In particular, we defined as a rural area one in which the agricultural share in employment is larger than the national average plus half-a-standard deviation--clearly a much more restrictive definition of rural areas than the concept adopted in 1992. 9.16 With the above caveats in mind, a few facts are worth noting. First, the increase in Polish unemployment would seem to be entirely accounted for by the rise in the long-term unemployment rate, which was about 5 percent in 1992 and rose to 11 percent in 2002. The share of nemployment, being long-term indeed, climbed from 30 percent to more than 50 percent over this ten year period. This is consistent with the macro picture on stock-flow adjustment offered in Section A. 9.17 Second, Polish long-term unemployment became increasingly concentrated on specific groups. This can be better grasped by looking at the last column on the right-hand-side in Table 9.2, which displays relative risk measures (the ratio of the share in total long-term unemployment of any given group to its share in total unemployment). Among the most vulnerable groups were women, persons with low levels of education, and persons living in rural areas. Their relative position deteriorated over time as is suggested by relative risk measures well above one unit and increasing from 1992 to 2002. Although the share of these "vulnerable" groups in Polish unemployment did not increase from 1992 to 2002, their relative position in the labor market would seem to have markedly deteriorated over time. 203 9.18 Overall, there are indications that compositional effects ­ associated with a worsening of the relative position of specific socioeconomic groups ­ have been an important factor in the rise of Polish long-term unemployment. Ad hoc policies will have to be developed to cope with these groups, as is discussed further below. Before turning to policies for vulnerable groups, it is important to assess the role played by pure duration effects, which may be at work, jointly with selection, in reducing the probability of re-employment as the spell of joblessness proceeds. D. EFFECTS OF DURATION ON WAGE ASPIRATIONS 9.19 Some indication of the role played by duration dependence in the persistence of Polish unemployment may come from data on the so-called reservation wage (the lowest wage at which individuals would take up job offers) of unemployed individuals. The LFS carried out in Poland since 1992 indeed contains a question on the lowest pay the interviewee is willing to accept when offered a job. The average reservation wage is in the various quarters about half of the actual average wage and nearly one-fourth larger than the minimum wage. The question on the reservation wage is also formulated in such a way as to find out whether the job seeker had in mind a post outside of the place of residence (which would therefore be likely to involve some compensation or premium for the costs of mobility), or whether reduced working time was sought (such as a part-time job). By checking all these factors, it is possible to receive some comparable information about the reservation wage of individuals. 9.20 Indications as to the reliability of such data come from matching observations on the same individual over time and comparing reservation wages stated when the person is still searching for a job with the actual accepted wages. Significantly, for those finding a job shortly after the interview in which they stated their wage aspirations (within two months), the ratio of the accepted wage to the reservation wage is very close. 9.21 According to job search theory, the optimal stopping rule of a rational jobseeker is to continue searching until receiving an offer which lies above a given threshold, capturing the opportunity costs of employment, that is, the reservation utility of the individual. If we believe that individuals are rational in their job search activities, then the stated reservation wage of workers should coincide with their reservation utility. 9.22 Table 9.3 shows results that were obtained by running a regression of the reservation wage of Polish unemployed individuals against LFS data from the second quarter of 2002. In particular, the following earning equation was estimated: ln(wi*)=+0MAi +1 Ai +2 Ai +1LEDi +2 HED+UBSi +JOBi +µFi +DURi + Xr +i 2 where w*i denotes the reservation wage of individual i, MA is a gender dummy variable assigning value 1 to male workers, A counts the years since the worker's birth (and is entered both linearly and with a quadratic term), LED identifies individuals with primary or lower levels of education and HED identifies individuals with tertiary education, and UBS is a dummy variable taking value 1 when the jobseeker is receiving unemployment benefits or long-term social assistance and 0 otherwise. We also included dummy variables (JOB) capturing the characteristics of the job being searched by the individual (e.g., part-time or full-time), the characteristics of the family (F) of the jobseeker (number of children, position in the household, number of relatives), and the region the individual belongs to (local unemployment rate, rural versus urban as well as a dummy capturing the fact of living in a rural area and receiving unemployment benefits). Clearly, the latter set of variables has only a cross-regional variation (we have information on 16 macro- 204 regions). Finally, we included a dummy variable (DUR), capturing the effect of long-term unemployment (defined as a jobless spell longer than 12 months) on the wage aspirations of the unemployed. 9.23 All variables are signed in line with the wage distribution, suggesting that wage aspirations are consistent with actual market wage premiums. For instance, the fact of having a primary or lower level of education yields a 10 percent wage discount with respect to individuals with secondary education. This is quite in line with estimates of mincer-type equations from actual earning data (e.g., Rutkowski, 1996). Significantly, the fact of living in a rural area involves a 5 percent reduction in wage aspirations, but not in cases where the individual is a recipient of unemployment benefits or social assistance (which does not affect reservation wages in urban areas). This can be grasped by looking at the variable interacting unemployment benefit or social assistance receipt with the fact of living in a rural area. Finally, the duration variable does not significantly affect the wage aspirations of jobseekers. Table 9.3: Estimates of the Reservation Wage of Polish Unemployed Workers, 2002 coefficient standard error significance Age 0.0190 0.0048 *** Square of age -0.0029 0.0001 *** Low edu. -0.0981 0.0157 *** High edu. 0.2346 0.0282 *** N° persons in 0.0094 0.0093 household Part-time seeker -0.1509 0.0257 *** First job seeker -0.0116 0.0196 UB and SA receipt -0.0152 0.0214 Spouse -0.0272 0.0210 Son or daughter -0.0887 0.0279 *** Relatives -0.0806 0.0426 * Un rate -0.3684 0.2876 Male -0.1850 0.0138 *** Rural area -0.0516 0.0226 ** Rural area* UB and 0.0551 0.0307 * SA LTU -0.0110 0.0138 Cons. 6.4286 0.1093 *** N° obs. 4.000 R-squared. 0.09 Source: Polish Labor Force Survey, 2002 205 9.24 There is, clearly, an endogeneity issue involved here, as duration is itself affected by the wage aspirations of individuals. Ideally, one would have to simultaneously estimate reservation wage functions and hazard rates. But the presence of negative duration dependence in job offer arrival rates should, in any event, result in declining wage aspirations with the length of unemployment spells. Thus, our results underplay the importance of negative duration dependence in the persistence of Polish unemployment. This fact may also be explained by the strong concentration of unemployment in rural areas in Poland. Agricultural jobs typically do not reward tenure and do not involve a serious risk of skill obsolescence, which means that low penalties are associated with longer duration unemployment. Boeri and Flinn (1999) found that tenure effects were negligible in job offer arrival rates in the private sector and that rural areas are dominated by private sector jobs. 9.25 Our estimates, on the other hand, suggest that unemployment benefit (or social assistance) receipt may undo the effects of local labor market conditions on the reservation wages of individuals. As unemployment benefits are paid at a flat rate, established by taking as reference the national average wage, and as in rural areas the cost of living may be as much as 30 percent lower than in urban areas, it is plausible that receipt of such transfers is more important in affecting optimal stopping rules in the rural region where Polish unemployment is concentrated. This does not necessarily mean that subsidies offered uniformly across the board are responsible for the spread of long-term unemployment in Poland, as it is important to also consider the regional profile of vacancy rates and, more broadly, labor demand. 9.26 Previous work evaluating the effects of unemployment benefits on job search incentives has, however, identified fairly negligible effects of unemployment benefit receipts on hazards from unemployment to employment in Poland (e.g., OECD, 1996; Puhani, 2000). Boeri and Steiner (1998) also found no evidence of residual unemployment benefit entitlement on exits from unemployment to jobs in Polish rural areas. Thus, we are rather inclined to believe that unemployment benefits and social assistance mainly play the role of an income support scheme to Polish long-term unemployed individuals which has little, if any, effect on job search incentives in the regions where the jobseekers are concentrated. Figure 9.3 is also consistent with this view. This suggests that the increasing duration of unemployment in Poland has always led to a decline in the coverage rate of unemployment benefits, rather than the opposite, as people came out of unemployment compensation rolls after exhausting the maximum duration of benefits. If labor supply incentives had been important in conditioning this process we would have expected a reverse order of causality and a positive association between the two series--that is, reduced coverage of benefits associated with the tightening of unemployment benefits should have induced a shorter duration of unemployment. 9.27 However, cash transfers may also play an obstructive role in rural areas insofar as social assistance provided at a relatively generous rate (in light of cross-regional differences in the cost of living) reduces incentives to relocate to where employment opportunities exist. While poverty reduction requires these transfers to be provided also in rural areas, rural-urban mobility can be encouraged with a combination of mobility incentives and tighter "passive policies." The next section assesses whether the current allocation of active policies corresponds to this policy mix and to the regional profile of Polish unemployment. E. THE SCOPE FOR ACTIVE POLICIES 9.28 The previous sections indicate that the spread of long-term unemployment in Poland is associated with the deterioration of the relative position of specific vulnerable groups, such as women re-entering the labor market, persons with low levels of education, and individuals living 206 in rural areas. We have also found that receipt of cash transfers significantly affected the opportunity cost of labor only for jobseekers living in rural areas. When unemployment risk is concentrated and the receipt of benefits discourages regional labor mobility, there may be a strong case for active labor market policies, on the grounds that they can hardly involve deadweight costs, while the substitution effects may be desirable. 9.29 Since the beginning of the transition to a market economy, the Polish government has applied a wide range of Active Labor Market Policies (ALMPs) to combat long-term unemployment. In terms of expenditures, three programs have been of particular importance: training, "intervention works" and public works. Training programs are meant to solve skill mismatches in the labor market. Workers with redundant skills or with no skills are trained in those occupations where there exists a strong demand by entrepreneurs in the expanding sectors of the economy. "Intervention works" is a program that in essence gives wage or job subsidies in the amount of the level of unemployment benefits. These wage subsidies are given to firms in the private or public sector if they hire an unemployed person, and they are larger the longer this person is kept on in the firm. Public works jobs are jobs directly created by the government, in particular by the municipalities, targeted mainly (but not exclusively) at the long-term unemployed. Many of these jobs are in construction and the cleaning of public buildings, parks, etc. (i.e., they have a low skills content). In principal, however, both intervention works and public works have been conceived to enhance or maintain the human capital of participants. 9.30 Several studies have investigated the effectiveness of ALMPs by having access to micro data on labor market transitions in Poland (Puhani and Steiner, 1996 and 1997; Boeri, 1997; Puhani, 2000) and by using a variety of techniques (e.g., estimates of augmented matching functions, simple duration models including program intakes, matching estimators substituting for randomization in labor market programs, etc.). The key question addressed by these studies is whether, after participating in an ALMP program, a person is better positioned in the labor market than if he or she had not taken part in the program. 9.31 The results of these studies are not altogether encouraging as to the effectiveness of these programs in fostering re-employment. In particular, training programs only mildly affect job finding rates. Better results are obtained in terms of wage subsidies, while public works schemes typically involve stigma effects which may actually reduce the likelihood of finding a job after the "treatment." While a proper evaluation of these schemes goes well beyond the scope of this chapter, we are interested here in assessing the rule implicitly followed by the Polish government in allocating resources for ALMPs. This is important, since greater allocations of resources imply larger active policy intakes for the beneficiary labor market policy administrations. Incentives to take up slots in active policy programs are particularly strong in Poland. Since the beginning of 1992, unemployment compensation has been limited to one year. If persons are unemployed for more than one year they have to rely on social assistance, which is not always paid or is paid in the form of material help. On the other hand, workers are entitled to unemployment benefits if they have worked at least 180 days in the preceding year. Thus, participating in a jobs program such as intervention works or public works for at least 6 months entitles a person to another round of 12 months of benefit payments. These "policy circles" create strong incentives to participate in ALMPs. 207 Figure 9.3: Share of Long-Term Unemployment and UB Coverage 85 ng-term 75 lo of e arhs 65 gna tne gearevoctfie my 55 45 unemplo bentn 35 yme mplo 25 une 15 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 year coverage of unemployed share of long term Source: Data supplied by the Ministry of Labor. 9.32 Figures 9.4A through 9.3D provide details as to the implicit rule followed in the allocation of ALMP resources in Poland. They plot active policy spending as a proportion of the labor force against various indicators of local labor market conditions. The regional allocation of ALMP expenditure has always been strongly responsive to the regional concentration of unemployment. The correlation coefficient has only mildly decreased over time. It was 0.93 in 1994 and 0.85 in 2002. Fewer resources have also been typically allocated to regions with relatively high vacancy rates. In this case the negative correlation has increased over time: it was ­0.36 in 1994 and ­0.51 in 2002. Moreover, regional of long-term unemployment are becoming less important in affecting the allocation of funds for ALMPs, as in 1994 the correlation was 0.65 and in 2002 it was insignificant at 1995 confidence levels (Figure 9.4C). 208 Figure 9.4A: ALMPS versus Unemployment Rates, 2002 70 60 50 R2 = 0.7186 Labor/ 40 30 Exp 20 10 0 ALMPs 10.0 15.0 20.0 25.0 30.0 Unemployment Rates Source: Data supplied by the Ministry of Labor. Figure 9.4B: ALMPS versus Vacancy Rates, 2002 70 Force 60 50 Labor/ 40 30 Exp 20 R2 = 0,2609 10 ALMPs 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Vacancy Rate Source: Data supplied by the Ministry of Labor. 209 Figure 9.4C: ALMPs versus LTU, 2002 70 60 R2 = 0,1571 50 Labor/ 40 30 Exp 20 10 ALMPs 0 0.92 0.925 0.93 0.935 0.94 0.945 LTU Share Source: Data supplied by the Ministry of Labor. Figure 9.4D: ALMPs versus Unskilled Unemployment, 2002 70 60 50 Labor/ 40 R 2 = 0,1278 30 Exp 20 10 ALMPs 0 10 15 20 25 30 35 40 Unskilled Unemployment Source: Data supplied by the Ministry of Labor. 9.33 Concerning the targeting of the vulnerable groups, that is, those groups more likely to experience long-term unemployment, Figures 9.4D suggests that currently ALPMs are not targeted to regions where women or the unskilled are relatively over-represented in the pool of jobseekers. 210 9.34 Thus, the above analysis suggests that ALMPs are being targeted increasingly to regions experiencing relatively large inflows into unemployment (e.g., those associated with large waves of labor shedding or with sizable cohorts of school leavers, rather than to regions experiencing a stronger incidence of long-term unemployment). This may be justified by the fact that the effectiveness of ALMPs is highly questionable in regions where there are few employment opportunities, but it also opens up the possibility of deadweight costs. The risk of providing employment opportunities to persons who would have found a job in any event is only partly mitigated by the fact of leaving aside the regions with the highest vacancy rates. In these regions with high unemployment but not particularly high (or not yet high) long-term unemployment, it may be advisable to devise programs for specific vulnerable groups. In other words, targeting should be better based on personal characteristics rather than on the presence of unemployment per se. 9.35 Resources spent in rural regions where long-term unemployment is particularly acute and where there are very few vacancies are likely to be wasted unless they are used to encourage mobility to regions with more buoyant labor market conditions. But such mobility incentives may be quite costly if they must offset larger differentials in living costs across rural and urban regions. F. CONCLUSIONS 9.36 Polish unemployment has become increasingly stagnant in the last five years. The spread of long-term unemployment is associated with the deterioration of labor market outcomes for women, persons with low levels of education, and persons living in rural areas. Policies aimed at reducing Polish unemployment will need to specifically target these groups and will need to use a combination of passive (income support) and activation measures. As long-term unemployment is regionally concentrated, ALMPs may also include mobility loans and other instruments that would encourage the regional mobility of jobseekers living in rural areas. All this activity requires careful scrutiny of the way in which the limited resources available for ALMPs are allocated to regions. There are indications that resources are currently being diverted away from the areas experiencing a higher incidence of long-term unemployment, and that targeting occurs mainly by looking at the local unemployment rate rather than at the presence of vulnerable groups in the population of jobseekers. 211 REFERENCES: Blanchard, Olivier, "Reform in Eastern Europe," Wider World Economy Group 1990 Report, WIDER and MIT Press, 1991. Blanchard, Olivier, Simon Commander, and Fabrizio Coricelli, eds., Unemployment, Restructuring and Labor Markets in East Europe and Russia, Washington, DC: The World Bank, 1995. Boeri, Tito, "Learning from Transition Economies: Assessing Labour Market Policies across Central and Eastern Europe," Journal of Comparative Economics, 25, 366-384, 1997. Boeri, Tito, "Transitional Unemployment," Economics of Transition, Vol. 2(1), 1-25, 1994. Boeri, Tito, and Christopher J. Flinn, "Returns to Mobility in the Transition to a Market Economy," Journal of Comparative Economics, 27, 4-32, 1999. Boeri, Tito, and Viktor Steiner, "Wait Unemployment in Transition Countries: Evidence from Poland," Konjunkturpolitik Applied Economics Quarterly, Berlin, 44-3, 1998. OECD, Lessons from Labour Market Policies in the Transition Countries, Paris: OECD. 1996. OECD, Unemployment in Transition Countries: Transient or Persistent? Paris: OECD. 1994. Pissarides, Christopher, "Search Theory at Twenty-One," CEP DP # 90, 1992. Puhani, Patrick, "Poland on the Dole: The Effect of Reducing the Unemployment Benefit Entitlement Period during Transition," Journal of Population Economics, February, 13:1, 35-44, 2000. Puhani, Patrick, and Viktor Steiner, "The Effectiveness and Efficiency of Active Labour Market Policies in Poland?" Empirica, 24, 3:209-231, 1997. Puhani, Patrick, and Viktor Steiner, "Public Works for Poland? Active Labor Market Policies during Transition," ZEW Discussion Papers, 96-01, Mannheim, 1996. Rutkowski, J., "High Skills Pay Off: the Changing Wage Structure during Economic Transition in Poland," Economics of Transition, Vol. 4, Nr. 1:89-112, 1996. Socha, Mieczyslaw W., and Jacob Weisberg, "Poland in Transition: Labor Market Data Collection," Monthly Labor Review, September 1999. Steiner, Viktor, and Eugeniusz Kwiatkowski, "The Polish Labor Market in Transition," ZEW Discussion Papers, n.3, Mannheim,1995. 212 10. THE EVOLUTION OF REGIONAL UNEMPLOYMENT IN POLAND, 1992-2002 Andrew Newell and Mieczyslaw W. Socha A. INTRODUCTION 10.1 This paper investigates the impact of the structural features of local labor markets on the geographical distribution of unemployment in Poland. We define structure primarily in terms of education and industrial composition. We interpret the educational composition of the potential labor force as reflecting the skill mix of the labor supply. The industrial, and sometimes the occupational, structures of employment are interpreted as being partly derived from the skill structure of the labor demand. 10.2 Why do we choose education as a key structural feature? There are three outstanding a priori reasons. First, most of the formerly centrally planned economies featured a much larger industrial sector and a correspondingly smaller service sector than their Western counterparts. Industry generally employs a much greater share of manual workers with vocational and primary education than services, in which secondary and tertiary educated workers predominate. A shift in the economy towards service production was one of the most predictable changes to occur through the transition. This shift set in train a change in the composition of the demand for labor. 10.3 Second, and reinforcing this change, experience in the United States and Western Europe shows that organizational change increases the demand for skilled white-collar labor. Privatization and the "big bang" hardening of the soft budget constraints facing state-owned firms in Poland are very likely responsible for similar organizational changes (see, for example, Newell and Socha, 2002). 10.4 Third, on the supply side there was a considerable rise through the 1990s in the completed educational levels among Poland's working age population. This is documented below. It is significant that there is a universal and reliable inverse relationship between broad educational rankings and unemployment rates. By and large, lower educated or less skilled workers have higher unemployment rates. This inverse relationship holds for Poland, and, again, this is documented below. 10.5 Theses three facts combine to suggest the hypothesis tested in this chapter, which can be restated as follows. If the regions of Poland differ markedly in their skill structures, and if they experience different size shifts in these structures, then these changes should account for part of the regional unemployment picture over the 1990s. 213 Box 10.1: Summary of Some Relevant Previous Studies Deichmann and Henderson (2000, DH), Faggio and Konings (2001, FK) and Newell and Pastore (2000, NP) find: · A great deal of variety exists across regions in restructuring and unemployment (FK) · Inflow rates to unemployment are strongly and inversely related to education levels (NP) · Regional unemployment rates are strongly persistent (NP) · Regional migration is low (DH) · The service sector is becoming more concentrated (DH). 10.6 Box 10.1 summarizes some relevant findings from three fairly recent papers, which motivate this study. For example, Poland's regions are diverse in terms of transition experience and unemployment. And relative unemployment rates are highly persistent. This suggests structural explanations. 10.7 The study covers 1994-2002 inclusive. It was a time of rapid economic growth in Poland. From 1994 to 1998, the Polish Labor Force Survey (PLFS) categorized industries, occupations and regions in a consistent way. After 1998, the 49 administrative regions of Poland, the wojewodstwa, were reduced to 16, with many new boundaries. The regions will be referred to by the anglicized expression voivodship. Because of this administrative change, researchers of Poland's regional evolution have concentrated on the period to 1998. We will use both pre-reform and post-reform data. 10.8 The next section (section B) provides a short macroeconomic and institutional context. Section C introduces the data, which are derived from various rounds of the Polish Labor Force Survey. We document the large shift in skills, the regional diversity of skills and industrial structures, and the changes in these structures across voivodships. Section D discusses theoretical issues in order to provide a foundation for the interpretation of the empirical work. We start with a standard model of regional labor markets, except that we incorporate two types of labor and skill-capital complementarity. We outline our natural rate and skill mismatch theories of regional unemployment in the medium term. We also discuss barriers to long-run wage and employment equalization. These barriers include: agglomeration effects, transportation advantages, differences in welfare payment systems, sorting effects where high skilled workers are reluctant to move to low skill regions, and regional differences in the quality of education. 10.9 Section E discusses the econometric results, and section F presents conclusions drawn from the findings. We find that higher skilled populations tend to generate lower unemployment rates and that controlling for population skill levels, and lower levels of the relative demand for unskilled workers, raises a region's unemployment rate. These results can be interpreted in terms of our natural rate and skill mismatch theories. We find that these theories contribute significantly to our understanding of the strong empirical relationship between regional unemployment and regional skill mix, but that they do not provide a full explanation. To arrive at a full explanation, some of the longer-run theories of regional aggregate labor markets need to be invoked. It is beyond the capacity of the data that we have to test their relative importance. 214 B. CONTEXT: POLAND'S MACROECONOMY AND LABOR MARKET Macroeconomic Context 10.10 Poland entered the 1990s with a deep recession, with hyperinflation, and with rapidly increasing unemployment. During the years 1990-92, GDP fell by almost 16 percent, total employment fell by 12 percent, consumer prices increased more than 15 times, and the unemployment rate rose to 14.3 percent. From 1992, the economy began to expand, with an annual average growth rate of 5.4 percent in the years 1993-2000. However, economic growth slowed sharply in the years 2001-2002, from 4 percent in 2000 to about 1 percent. In contrast to output, employment growth never emerged; the annual average employment growth rate for the period 1991-2000 is ­0.2 percent. Since 1998, the national economy experienced a net loss of 1.37 million jobs (about 8 percent). 10.11 Economic growth decreased, primarily as a result of a further decline in domestic demand growth (see Table A.10.1 in the Annex). However, despite the falling international competitiveness stemming from the real effective appreciation of the zloty (38 percent since 1995), exports continued to rise, and together with falling imports growth, the current account deficit fell from 8 percent in 1999 to about 3.3 percent of GDP in 2002. On the supply side, the basic facts of recent performance are: (i) a sharp increase in aggregate labor productivity (by 58.7 percent in 1993-2002); (ii) decreasing employment, as has been mentioned; and (iii) decreasing gross investment in fixed capital (the share of fixed investment in GDP dropped from about 25 percent in 1998 to 20 percent in 2001). Aggregate real wages were growing at a slower rate than labor productivity, by 30 percent and 55 percent, respectively, during 1993-2001. Since 1993, annual inflation declined substantially, falling to 1.2 percent in 2002. 10.12 A number of factors are likely to have influenced the Polish economy at the end of the last decade and to have caused the change in macro policy (see Newell, 2001). These factors include: (i) the Russian crisis in 1998; (ii) the increase in world energy prices; and (iii) the slowdown in the EU countries, combined with the depreciation of the euro. In 1998-99, the government adopted restrictive monetary policies when it was faced with growing external imbalances, an increased perceived risk of financial crises, and the need to meet the macroeconomic conditions required for EU membership. 10.13 Monetary policy was characterized by setting high real interest rates (two digit) while tolerating the appreciation of the domestic currency. (In April 2000, the government introduced a floating exchange rate regime.) On the other hand, fiscal policy was eased, and the public finance balance has deteriorated markedly during the last three years, largely as a result of persistent growing outlays on social transfers and of difficulties with the public sector reforms. The reforms of the social security system, the health sector, the education sector, and territorial administration, implemented in 1999, do not appear to have greatly improved public finance. 10.14 The tight disinflationary policy and the slowdown in economic activity contributed to the rapid increase in the registered unemployment rate. By the end of 2002, there were 3.2 million jobless representing 18.1 percent of the labor force. (In LFS data for the fourth quarter of 2002, the unemployment rate was 19.7 percent.) If we add the rapidly growing number of discouraged workers (their share in unemployment increased from 2.4 percent in 1992 to 11 percent in 2002), then broadly defined unemployment is much higher (about 22 percent in the third quarter of 2002). 215 10.15 There were three different periods of macroeconomic development in Poland. From 1990 to 1994, decreasing inflation (though still at very high levels) was evident, together with growing unemployment, which suggests shifts in aggregate demand. There was a benign decline in both variables in the period 1994-97. The classical shape of the Phillips curve in the period 1998-2002 reflected the impact of strong anti-inflationary monetary policy and possibly an emerging new equilibrium in the inflation-unemployment trade-off. The large scale of the increase in unemployment compared to the more modest fall in output growth suggests that structural and other changes led to a shifting out of the medium-term natural rate (see Newell, 2001). Certainly, increased unemployment at a stable level of vacancy rate suggests a decline in search efficiency (see Figure A.10.2 in the Annex). 10.16 The stagnating and recently declining level of employment associated with the increase in the number of jobless coincided with an unprecedented rise in the non-employment rate, especially for females. The data in Annex Table A.10.2 show monotonic declines in both participation rates (from 61.7 percent in November 1992 to 55.0 percent in the fourth quarter of 2002) and employment rates (from 53.3 percent to 44.1 percent, respectively). In the fourth quarter of 2002, 61.9 percent of women age 15 and over were not employed. One of the worrying consequences was the decrease in the share of households with at least one employed person (without unemployed), from 58.4 percent in 1993 to 52 percent in 2001. 10.17 A previous World Bank study (World Bank, 2001) highlighted the fact that the recent Polish labor market adjustment to the ongoing restructuring and privatization is characterized by: a high rate of job destruction; substantial changes in the structure of the labor demand; a slow rate of job creation; and barriers in the transition from old to new jobs. 10.18 Four basic changes on the demand side can be identified. First, there is a large shift in the labor demand away from low skilled employees and toward high skilled employees. Data from the Polish Labor Force Survey show that in 1992-2002, the number of employed with a tertiary level of education increased by 52.3 percent, while the number of workers with a primary and lesser education decreased by 53.3 percent. The employment rate for women at the latter level of education living in urban areas dropped to 14.4 percent. Previous studies of the wage distribution revealed increases in the wage premiums for higher education, skilled occupations and private sector employees (for details, see Rutkowski 1996, Newell 2001, Newell and Socha, 1998, and Newell and Socha, 2002). 10.19 Second, owing fundamentally to the industrial bias of the communist system, we are observing large changes in the industrial structure of employment. LFS data show that since 1994 employment in agriculture, hunting and forestry fell by 25 percent,1 while in mining the fall was 50 percent, and in manufacturing the fall has been about 19 percent since 1994. Employment in real estate and business activities increased by 180 percent, in hotels and restaurants by 47.8 percent and in trade by 10 percent.. However, since 1998 there has been a reversal of the trend with respect to rural activities, and the share of agriculture, forestry and hunting in total employment increased slightly from 18.1 to 18.5 percent. These changes are further discussed in section C. 10.20 Third, the driving force in job creation is the development of the private sector. The process of privatization has slowed down recently, but in 2002 the share of the private sector in 1Data for private agriculture estimated from other sources show a different trend (i.e., an increase in the number of employed in agriculture, forestry and hunting by 6.2 percent with a share of total employment of 29.2 percent. 216 total employment was 73 percent (according to the PLFS, 67.2 percent).2 Its share in industrial output was 75 percent, in investment it was 72 percent, in exports it was 88 percent and in imports it was 92 percent. The private sector is dominated by small and medium size businesses.3. Through the 1990s it attracted greater proportions of young, male and less educated workers than the public sector (Socha and Weisberg, 2002). Between 1992 and 2002, the public sector lost 3.1 million workers (41 percent), while employment in the private sector increased by 21 percent. It is interesting to note that after 1998 the decline in the number of employees is higher than other groups of workers in the private sector. This may reflect existing and possibly growing barriers to running old businesses and starting new businesses. 10.21 The macroeconomic stabilization of the Polish economy, as well as the recent high real interest rates, have coincided with inflows of foreign direct investment. The accumulated value of total inward FDI in December 2002 as estimated by the Polish Agency for Foreign Investment was US$65 billion. The fourfold increase in the number of workers employed by foreign companies since 1992 is impressive; however, the share of the foreign sector in total employment is still relatively small (4.2 percent at the end of 2001). 10.22 Fourth, there is evidence of the growing importance of atypical forms of employment. The share of part-time workers is very stable at a level of 10 percent, however; the share of temporary paid workers in the total number of paid employees increased between 1992 and 2000 from 2.9 percent to 6 percent. This may have been a response to the increased levels of statutory employment protection (see under section C, The Pattern of Regional Unemployment over the Period). Also, the Central Statistical Office's estimates of employment in the shadow economy suggest an increase from some 805,000 in 1995 to 895,000 in 2001. 10.23 To return to our discussion of the unemployed, it is interesting to note that the share of he unemployed who lost their jobs through plant closure has decreased since 1992 by about 30 percent (or 13 percentage points), which may suggest a less intensive restructuring during more recent years. Instead, the very sharp increase in the share of re-entrants and new entrants in he unemployment pool after 1998 can be seen as a sign that the baby boom of the early 1980s began to show up in the labor market. 10.24 The socio-demographic pattern of unemployment is very stable. The typical jobless person is young (55 percent are under 34), is low skilled (60 percent have no more than a basic vocational education), and lives in urban areas (67 percent). Among the unemployed, 48 percent have been searching for a job for more than 13 months. The gender structure of unemployment is balanced; however, the rate of unemployment is higher for women than for men in each year. Over 50 percent of women are long-term unemployed and the share of discouraged workers in unemployment is higher for women (12.8 percent) than for men (9 percent). A new phenomenon is the acceleration in the unemployment incidence for persons with university diplomas. This group has experienced an increase in unemployment incidence of 205 percent since 1998. 10.25 In summary, two primary conclusions can be drawn from this picture. First, the distribution of unemployment incidence is very uneven among various labor force groups, which indicates the structural character of the current slump in the labor market. Second, with the high and growing non-employment rate, the Polish labor market has not spread the benefits of the economic growth of the 1990s evenly across the population. 2Including private farmers. 3The number of individual, non-incorporated businesses increased after 1992-93 by 77 percent and in 2001 exceeded 2.66 million establishments. 217 Changes in Labor Institutions 10.26 This very high level of unemployment suggests the need to liberalize the labor market to increase the potential expansion of labor demand. Despite many attempts, the structural reform of the labor market has not been completed (for details, see Kwiatkowski et al., 2001). However, a system characterized by high levels of labor taxation, relatively generous welfare assistance and employment protection still exists. 10.27 Employment Protection. Changes in the legal regulations of employment contracts covered by the Labor Code were made in 1996 and 2002. Uniform rules were introduced for private and public sector employees. In 1996, the changes increased employment protection by extending the notice period for workers with various employment histories. For example, the qualification for three months' notice was cut from ten to three years of tenure. Similarly, the qualification from two weeks' notice was reduced from one year to less than six months. Additional burdens were imposed on the employers in small and medium-size firms. Any dismissal of more than 10 percent of the workforce in a firm with not more than 1,000 workers was defined as a group dismissal, with severance payments (of from one to three months' wages) and with compensatory pay for those who took a new job with a lower wage. The universal labor regulations (concerning working hours, overtime pay, the administration of workers' records, etc.) across ownership sectors, implemented by the 1996 changes in the Labor Code, increase the employment costs for private employers, especially in the SME sector. The 2002 revision of the labor regulations was to lower labor costs and increase labor market flexibility. The main changes are: · Very small firms (20 employees or fewer) do not have to establish wage rules and work codes. · Until Poland becomes a member of the EU, an unlimited number of successive fixed term contracts is permitted before a worker is automatically deemed a permanent employee. · The minimum overtime premium has been lowered. · Severance payments for collective dismissals in small firms have been abolished, and severance pay in other firms has been linked to current tenure rather than total to years of work experience. The level of severance pay varies from one monthly wage for workers with two years' tenure to three monthly wages for workers with eight years' tenure. 10.28 Riboud et al. (2002), in their study of the degree of labor market flexibility in transitional countries, show that Poland has (after Hungary) the second most flexible labor legislation, with the aggregate employment protection index value of 2 (with the average of 2 for all OECD countries and 2.4 for EU countries). The authors find that the highest level of protection in Poland is for collective dismissals (3.9) and the lowest is for temporary employment. 10.29 Labor Market Policies. Between 1992 and 2001, public expenditures on labor market policies varied between 0.7 percent and 1 percent of GDP. Over 75 percent of this sum was spent on paying unemployment benefits and paying early retirement schemes. Replacement ratios range from 37 percent in 1994 to 21 percent in recent years. Unemployment benefits expressed as a fraction of the minimum wage declined from the 80 percent in the first years of transition to 58 percent in 2001. Unemployment benefits are the area of the greatest progress in labor market policy reform. First, the eligibility criteria for benefits were tightened. At the beginning of the 1990s all claimants could obtain unemployment benefits, but since 1993 the unemployed must have worked for at least 180 days during the year preceding registration at the Labor Office. Since 1997, it is required to have had 365 days of working history during the last 18 months 218 before registration as unemployed. In addition, in 1997 school leavers lost the right to obtain unemployment benefits, and labor offices offered them training with a stipend instead. Second, the period of benefit duration was shortened and differentiated, across labor force groups and by the local level of unemployment. In 1990 the period of entitlement was unlimited; in 1992 it was shortened to 12 months, and in 1993 the duration was extended in regions with very high levels of unemployment. Since 1997, benefit duration is 6 months in regions where the unemployment rate is lower than the average national unemployment rate and 12 months in regions where the unemployment rate is higher than the average national rate. In addition, durations of 18 months are allowed in regions where the local unemployment rate is more than twice as high as the average, and also for the unemployed who have at least 20 years (females) or 30 years (males) of employment history.4 10.30 Third, the parameters of the unemployment benefit regime were changed several times. In 1990, the level of benefits depended on the previous earnings (with the regressive rate ranging from 0.7 to 0.4). From 1992, the level was unified for all unemployed to 36 percent of the average wage in six major sectors of the economy. In 1996, the level of the unemployment benefit was set at 260 zloty, quarterly adjusted to the inflation rate. Job losers aged 55 for women and 60 for men who have the minimal service entitlement to the pension could receive the unemployment benefit at the level of 75 percent of the previous wage, and the benefit has been indexed to the average wage increase. If the job losers were laid off and lived in the regions threatened with high structural unemployment, the level of unemployment benefit should be 160 percent of 260 zloty. In 2001, the basic rate was granted to those unemployed who worked at least 5 years (but no more than 20 years), while 80 percent of the basic rate was granted to those who worked less than 5 years, and 120 percent was granted to workers with at least 20 years of employment. Since 1997, older persons (women who reached age 58 with at least 20 years of employment, and men aged 63 with at least 30 years of employment) who are unemployed and eligible for unemployment benefits can receive pre-retirement benefits (120-160 percent of the basic benefits) or pre-retirement allowances (80 percent of the old-age pension, if the retirement age conditions were met). 10.31 Although active measures of labor market policies are similar in OECD countries, the number of unemployed covered by these programs is relatively low. The share of the unemployed in training, subsidized work, and public works, and on internships, in total unemployment decreased from 26 percent in 1997 to 4 percent in 2001. There were some attempts to focus ALMP on selected groups of unemployed: women, the long-term unemployed, jobless parents, school leavers, former prisoners, etc. However, the empirical studies on the efficiency of the ALMP point to the low effects on unemployment of these efforts, with the exception of training. 10.32 This brief overview of the main changes in labor market institutions clearly shows progress in removing barriers to job creation, higher labor market flexibility and labor supply adjustment. In addition, it would be difficult to evaluate the existing regulations concerning employment protection, wage negotiations, or labor policies as very different from those in other OECD countries. However, these positive changes have not created a sustainable increase in jobs and have not prevented the massive increase in unemployment. 4In some special cases the period of unemployment benefit payment can be extended to 24 months. 219 C. UNEMPLOYMENT AND STRUCTURAL CHANGE ACROSS POLAND'S REGIONS The Pattern of Regional Unemployment over the Period 10.33 Table 10.1 and Figures 10.1A and 10.1B establish the fact that although the pattern of unemployment across Poland's regions has been evolving, a strong correlation remains between the regional patterns of the early to mid-1990s and the early years of the twenty-first century. Some regions, such as those in which old, heavy industries or collective agriculture were important employers, seem to have persistently high unemployment (see Gorzelak, 2003 for a useful historical and structural review). Yet, as noted above, previous attempts to categorize regions by industrial characteristics have not worked very well. It is our contention that the deeper common feature of high unemployment regions is the low educational level of the population. We discuss the evidence for this hypothesis in the next section. Table 10.1: Unemployment in the New Voivodships, 1995, 2001, 2002 1995 May 2001 May 2002 q4 Poland 11.2 17.9 19.7 DOLNOLSKIE 14.4 22.9 27.2 KUJAWSKO-POMORSKIE 14.1 23.1 21.2 LUBELSKIE 7.7 14.6 16.6 LUBUSKIE 13.3 26.3 25.9 LÓDZKIE 11.8 21.1 19.5 MALOPOLSKIE 7.9 13.3 16.1 MAZOWIECKIE 10.0 12.9 16.9 OPOLSKIE 11.0 20.9 17.5 PODKARPACKIE 11.0 15.0 18.4 PODLASKIE 9.4 16.1 17.7 POMORSKIE 13.6 17.4 21.9 LSKIE 8.9 18.5 18.8 WITOKRZYSKIE 11.7 16.9 18.7 WARMISKO- MAZURSKIE 19.4 24.4 24.8 WIELKOPOLSKIE 11.2 18.7 18.2 ZACHODNIOPOMORSKIE 15.4 22.7 25.9 Source: PLFS. 220 Figure 10.1A: Unemployment rates (%) in the old voivodships, 1992 and 1998 25 20 1998 15 ber 10 Novem 5 0 0 5 10 15 20 25 30 November 1992 Source: Authors' calculations from PLFS data provided by the Central Statistical Office. Figure 10.1B: Unemployment rates (%) in the new voivodships, 1995 and 2002 30 25 2002 4 20 Quarter 15 10 5 7 9 11 13 15 17 19 May 1995 Source: Authors' calculations from PLFS data provided by the Central Statistical Office. The Inverse Relationship and the Big Shift in Skills 10.34 Table 10.2 demonstrates the shift in completed education levels. The middle section of the table shows a rise of 4.5 percentage points in the share of the working age population with a completed secondary or higher education for 1992-2002. It is clear that older, less-qualified people are retiring and being replaced by well-qualified young people. The population aged 15 and over is about 31 million in Poland, and thus this growth has resulted in about 1.4 million more adults with secondary or higher qualifications in 2002, compared to 1992. 221 10.35 This process was more dramatic among the employed. Over this ten-year period, there was a drop of just under 14 percentage points in the share of workers with a primary education in employment, and a corresponding rise of 11 percentage points in the share of workers with at least a secondary education. Table 10.2: The Big Shift in Skills: Changes in the Distributions of Employment, Population1 and Unemployment by Level of Completed Education, 1992-2002 Tertiary Secondary Lower Primary or lower Vocational Share in 13.6 27.7 31.4 27.3 employment in 1992 Share in 15.6 34.2 35.2 15.0 employment in 1999 Share in 18.3 34.0 34.0 13.6 employment in 2002 Change in 4.7 6.3 2.6 -13.7 employment share 92-02 Share in 10.6 26.6 29.5 33.3 population in 1992 Share in 10.5 29.7 31.0 28.8 population in 1999 Share in 11.8 29.8 31.0 27.1 population in 2002 Change in 1.2 3.2 1.5 -6.2 population share 92-02 Share in 5.2 28.9 42.9 23.0 unemployment in 1992 Share in 5.5 30.9 42.0 21.5 unemployment in 1999 Share in 6.6 30.2 42.9 20.2 unemployment in 2002 Change in 1.4 1.3 0.0 -2.8 unemployment share 92-02 1/ Population aged 15-60. Source: PLFS. 10.36 The top and bottom sections of the table show that the unemployed in Poland are typically less educated than those at work; in other words, the less educated workers have higher 222 unemployment rates. This is an important fact on which our analysis focuses, and, as has been mentioned, it is a well-documented phenomenon visible in many countries. It largely reflects the inverse relationship between job turnover rates and (therefore) inflow rates to unemployment on the one hand, and skills (or education) level on the other (see Layard, Nickell and Jackman, 1991, Chapter 6). The unemployment rate depends on both the rate of inflow to unemployment and the duration of unemployment spells. Layard et al. shows that differences by skills in unemployment durations also exist, but that these are less pronounced and thus less important in explaining the inverse relationship. We illustrate these relationships for Poland in 1997-98 in Tables 10.3, 10.4A and 10.4B. 10.37 Table 10.3 shows the inverse relationship very clearly. For example, workers with only the primary level of education experience an unemployment rate over three times that of tertiary level educated workers. Table 10.4A shows that uncompleted job tenures increase with completed education levels. This is consistent with lower turnover and lower inflow rates to unemployment for higher educated workers. Table 10.4B shows a milder, but still notable, inverse relation between length of unemployment spell and education level. Table 10.3: The Inverse Relationship, 1998 and 2002 Unemployment rate (%) Participation rate (%) Completed education 1998 2002 1998 2002 level Tertiary 4.1 9.3 90.3 89.0 Secondary 9.9 20.2 75.0 74.4 Lower Vocational 12.4 26.5 79.3 77.6 Primary 16.1 29.7 39.0 37.2 Source: PLFS, November 1998 and spring 2002. Table 10.4A: Labor Market Attachment by Completed Education Level, Spring 2002 Completed education level Percentage of employees not working one year earlier University 5.8 General Secondary 5.2 Lower Vocational 7.5 Primary 12.0 Notes: The sample was employees aged between 25 and 45. Source: PLFS, spring 2002. Table 10.4B: Average Uncompleted Unemployment Duration by Educational Attainment, 1998 and 2002 Average uncompleted unemployment duration (months) Completed education level November 1998 Spring 2002 Tertiary 9.2 12.0 Secondary 12.8 14.8 Lower vocational 12.2 15.5 Primary or less 13.5 17.3 Source: PLFS, November 1998 and spring 2002. 10.38 In Table 10.2, between 1992 and 2002 the gap in average levels of completed education between the employed and the unemployed grew significantly. The third and seventh lines of 223 Table 10.2 show a much larger shift towards higher levels of education among the employed than among the unemployed. If we construct a Duncan and Duncan (1955) segregation index5 to quantify the differences in education structure between the employed and unemployed, it rises from 12.7 for 1992 to 15.5 for 2002. 10.39 Turning to employment, the changing structure of jobs entailed a shift towards the employment of better-educated workers within industries and a shift in industrial structure towards education-intensive sectors. Table 10.5 illustrates this. If we compare the positions in 1992, 1998 and 2002, there is a within-industry shift away from the employment of lower vocational and primary educated workers for all the major industrial groupings. Between industries, there is a dramatic fall in employment in agriculture, an industry that employs mostly workers with lower levels of education. The main gains in employment come in trade and public services, both of which employ majorities of better educated workers. The relative importance of between and within industry changes can be gauged by a shift-share decomposition. Using one- digit industry level data, the aggregate change in the proportion of secondary or higher educated employment is roughly a one-third between-industry change and a two-thirds within-industry change. Table 10.5: The Structure of Employment by Major Industrial Sector and Education, 1992, 1998 and 2002 Tertiary and Lower Industry share secondary share in vocational and in total industry (%) primary share employment (%) in industry (%) Agriculture, 1992 14.8 85.2 26.8 Agriculture, 1998 16.4 83.5 21.3 Agriculture, 2002 21.4 78.6 23.3 Manufacturing, 1992 36.1 63.9 24.9 Manufacturing, 1998 37.5 62.5 20.5 Manufacturing, 2002 42.7 57.3 18.8 Trade and related, 1992 55.8 44.2 10.4 Trade and related, 1998 56.3 43.7 15.1 Trade and related, 2002 63.5 36.5 15.4 Public administration, health 76.1 23.9 16.8 and education 1992 Public administration, health 75.3 24.7 21.3 and education, 1998 Public administration, health 78.9 21.1 21.5 and education, 2002 Total, 1992 41.3 58.7 100 Total, 1998 46.1 54.9 100 Total, 2002 51.2 48.8 100 Source: PLFS, November 1992, November 1998 and spring 2002. 5If sne is the share of employment with education level e and sue is the share of unemployment with education level e, then the Duncan and Duncan (1955) segregation index, D, is D = 1 s - sue *100. 2 ne e 224 Regional Variations in Skill and Unemployment 10.40 Table 10.6 illustrates the variety of unemployment and skill structures that existed across the voivodships in 1994. It may be easiest to obtain an impression by looking at the maximum and minimum voivodship values for the variables. Unemployment rates ranged from moderate but substantial to extremely high. Table 10.6: Voivodship Variations in Changes in Unemployment, Industry and Skill, Percentage Point Changes, 1994-98 Levels Changes Max Min Mean Std. Dev. Max Min Unemployment rate 28 8 -3.6 3.1 2.0 -13.0 Completed education shares Tertiary 20 4 0.8 1.7 5.0 -3.0 Secondary 36 16 1.5 2.6 8.0 5.0 Lower voc. 35 14 1.2 2.8 8.0 -5.0 Industry shares Agriculture 68 21 -4.9 5.3 7.0 -18.0 Industry 56 12 1.0 4.8 12.0 -9.0 Services 66 3 4.0 4.2 14.0 -4.0 Non-agricultural occupational shares Non-manual 1 47 18 -0.3 5.9 14.0 -13.0 Non-manual 2 27 15 1.8 4.1 11.0 -7.0 Manual 61 32 -1.4 5.4 13.0 -20.0 Notes: Industry includes mining, manufacturing, utilities, construction and transport. Services include public administration, education, health, other services, financial and business services, and trade. Non-manual 1 includes professional, managerial and technical workers. Non-manual 2 includes clerical and sales workers. Source: PLFS. 10.41 There are wide variations in educational shares, which undoubtedly partly reflect the geographical distribution of economic activity generated in the communist period. The default category includes all of those workers with primary or lower levels of education. The voivodship with the highest share of workers with tertiary education (20 percent) is Warsaw. However, there are voivodships where the share of workers with tertiary education is as low as 4 percent. For a rough comparison, in the United Kingdom Labor Force Survey for spring 2002 the share of the population of working age with a university degree (which does not comprise all of the tertiary educated workers) was 15.4 percent. Across regions, this share varied from 11.7 percent in the West Midlands to 29.3 percent in Inner London. Clearly, there are regions in Poland where graduates are in very short supply compared to Western Europe. 10.42 There is a similar variety in occupational shares, which are calculated from non- agricultural employment. The industrial shares show the greatest spatial variation. D. THEORETICAL ISSUES 10.43 What factors give rise to skill-based unemployment differentials? The most obvious theory relies on lower rates of job turnover for skilled workers, which converts into lower frictional unemployment via smaller inflows to unemployment. The textbook answer to why skilled workers have lower turnover rates involves non-wage labor costs. For highly skilled workers, hiring, training and firing costs are usually greater (see, for example, Filer, Hamermesh 225 and Rees, 1996). This generates lower layoff probabilities, as firms seek to lower costs. In addition, in the face of these fixed costs, there is a greater incentive for firms to pay efficiency wages to reduce voluntary quits. 10.44 How much of Poland's regional unemployment variation could be due to such equilibrium or natural rate phenomena? To answer this, we develop first an equilibrium model of regional unemployment and then discuss possible disequilibria. What follows is an informal discussion, because the economic argument lacks sufficient ambiguity to warrant mathematics. 10.45 The working assumptions are that production of a single output takes place via a well- behaved production function of three inputs: capital, skilled labor and unskilled labor. We will assume skill-capital complimentarity, or, more formally, that the elasticity of substitution between capital and skill is lower than that between unskilled labor and the other two factors.6 We will further assume that workers cannot convert from one level of skill to the other. Lastly, we will assume that unskilled workers, because of lower hiring and firing costs, have higher rates of job turnover and higher rates of frictional unemployment.7 10.46 In a standard competitive model, with full factor mobility and regions identical in amenities, raw material endowments and technologies, regional unemployment differences cannot exist in equilibrium, as capital and labor will migrate to equalize wages across regions. If capital were immobile and unequally distributed across regions, but labor could migrate, then wages would still equalize, but there would be more skilled workers in high capital regions. This would leave equilibrium regional unemployment rate differences of the type discussed in below, under Regional Natural Rate Differences, with skill-intensive regions having lower unemployment rates. 10.47 We now disallow migration, but let capital be mobile. To add interest, we also assume that regions vary in the skill mix of their populations. Profit-seeking will ensure full employment (subject to turnover) of both types of labor, and so again regions with more highly skilled populations will enjoy lower unemployment rates. 10.48 Of course, a firm's location decision will depend on more than wage costs. The agglomeration effects emphasized by economic geographers (see Fujita et al., 1999), for instance, will weigh against full exploitation of wage differences, perhaps especially in manufacturing, but also in some services. Other features held constant in the thought exercises of the previous paragraphs, such as variations in material endowments and transportation advantages, will also generate a more varied and complex spatial distribution of skills and frictional unemployment. These features are discussed in this section under Aggregate Labor Demand and Endogenous Populations. Regional Natural Rate Differences 10.49 Regional labor immobility seems a reasonable assumption, given the evidence from Deichmann and Henderson 2000 (see Table 10.7). How much regional variation in unemployment can be generated by regional skill differences alone? We should first note that a 6 These assumptions could be replaced by a demand side where two goods are produced with different factor intensities. In that case, factor demand changes reflect changes in product demand. 7In an otherwise competitive framework, turnover and unemployment need motivation. The simplest mechanism we know of is due to Newell and Symons (1991), who assumed that each period the value of every worker-firm match was subject to a stochastic shock. A large enough negative shock leads to the breakup of the match. For unemployment, let there be a single fixed enforced period of unemployment between jobs for all workers. 226 region's unemployment rate can be written as a weighted average of the unemployment rates of workers of different types, with region-specific weights. We discuss below why regions might differ in terms of skill endowments. For the ith region and letting j index skill or education types, we can write the unemployment rate ui as: ui = wijui , (1) j j where wij = sij pi j. pi Table 10.7: The Slowdown of Migration, Gross Migration Flows (`000) Total Rural-urban Across voivodships: Average 1986-89 630 227 Average 1997-98 422 113 Within voivodships: 1989 369 150 1994 277 87 Source: Deichmann and Henderson (2000), Box 4. 10.50 Here sij is the share of education group j in region i and pij and pi are the regional overall and education group-specific participation rates. If educational group unemployment rates and participation rates do not vary much across regions, then we could calculate a region's unemployment rate as a weighted average of the economy-wide unemployment rates of the workers of different types, with region-specific weights as follows: ui = iju j , where N j ij = sij pj . Here p and pj are the aggregate overall and education group-specific participation p rates. Given that the w weights sum to unity, this formula can be modified. Let j be the group with the lowest level of education. Then the natural rate can be written: ui = sij N uj pj jJ - p uJ pJ + constant. p (2) 10.51 Equation (2) suggests to us a regression of unemployment rates on regional shares of skill. The estimated coefficients would have the interpretation of being participation-adjusted unemployment rate differentials. What magnitude might one expect of these coefficients? Table 10.3 offers data on education-specific unemployment and participation rates for Poland in 2002. Using these statistics for tertiary and primary educated workers, between whom there are the largest differences, the term in the square bracket of (2) can be calculated as (0.093*(.9/.75)- 0.297*(.39/.75)). The result is just under -0.05, or -5 percentage points. This can be interpreted as the fall in the so-defined natural rate if all primary-educated workers gained a degree. Section E shows that the econometrically estimated impacts of education shares on a region's 227 unemployment rate are much too large for this natural rate interpretation to be the full theory of Poland's regional inverse relationship. 10.52 Returning to equations (1) and (2), we can understand what might give us a larger impact of population education shares on unemployment rate than that predicted by the natural rate theory. To understand this, we need to consider other theories. Part of the explanation could be skill mismatch, to which we now turn. Regional Mismatch 10.53 Our theoretical discussion was about equilibrium outcomes. Can skill mismatch or disequilibria explain regional unemployment? Let us hypothesize that the regional skill distribution of the population at the outset of the transition reflected the pattern of labor demand. This is fairly likely under central planning, which put people where they were thought to be needed. Through the 1990s, we have seen that the relative employment of more highly educated workers grew faster than the relative supply. Even if average real wages adjust to fully employ all skilled workers, if relative wages do not adjust at the regional level in the medium term, then increases in the relative demand for educated workers over and above the increases in relative supply will create unskilled mismatch in unemployment8. The process is illustrated in Figure 10.2. With constant relative wages, or at least if relative wages do not fully adjust to compensate, increased relative demand for skilled workers results in excess supply and skill mismatch unemployment of unskilled workers. 10.54 We can summarize these two theories as follows: let the voivodship medium-term unemployment rate, uvt , be generated by the skill mix, S, of the workforce, where S increases with skill. But also, in the medium term, mismatch between S and the demand skill mix, D, can, via skill shortages,9 raise unemployment, so that we have uvt = u(Svt , Dvt - Svt ) , with uS < 0 and u(D-S)>0 (3) 8We do not test this regional relative wage rigidity proposition here. Poland's collective bargaining system, and our own initial empirical experimentation, suggest that relative wages by skill are not very sensitive to local labor market conditions. 9There is a technical condition required for this argument to hold. In a two-skill model, with profit-maximization the skill ratio will depend on relative wages. But the model also requires a full-employment condition for skilled workers to ensure that the skill shortage "bites." 228 Figure 10.2: Skill Mismatch with Rigid Wages Ws S1 Wu S2 Ws Wu D1 D2 Excess Demand Ns Nu Source: Newell and Socha (2002). Aggregate Labor Demand and Endogenous Populations 10.55 A final set of theories needs to be added. Although the housing market forms a formidable barrier to migration, it is hard to imagine it stopping people from moving over the long term. Similarly, though relative wages may not respond to local unemployment in the short run, it is difficult to imagine there being no response in the long run. What might act against these tendencies and sustain unemployment differences in the long run? We have already mentioned some possibilities: agglomeration effects will keep capital in the skill-rich cities, which probably also have transportation advantages. Therefore, there are good reasons to imagine the permanently limited regional equalization of physical capital. 10.56 The income support system is also important. With a nationally uniform unemployment benefit system, for example, replacement ratios will be higher in lower wage regions. As was outlined in section B, the benefit system is more generous in high unemployment regions, and thus the replacement rate will be much higher in those regions. The extent to which these benefits discourage search has been widely researched. The current consensus is a small but usually significant effect (see, for example, Arulampulam and Stewart, 1995, for high quality United Kingdom evidence). 10.57 A further possibility is regional variation in the education system. If the low-skilled regions also have less good education systems, then they will under-produce well-educated young workers. We might add that the sorting effects studied by Fernandez and Rogerson (2001) would exacerbate this. In that model, people of similar education have preferences for living close to 229 each other, so that, for example, the wage incentives required to make a highly skilled worker move to a low-skilled area might be very high. 10.58 These extra effects will create regions of physical and human capital shortages. It is hard to doubt that some combination of these effects is at work in many parts of the world. The policy message is that the skill base and the incentives to inward investment both need attention if pockets of high unemployment are to be successfully tackled. E. ECONOMETRIC RESULTS 10.59 Data on the 49 old voivodships (1994-98) were combined to make a panel of 245 observations. In addition, data from 16 new voivodships, combining 6 half-yearly PLFS rounds from autumn 1999 to spring 2002, were used to create a 96-observation data set. The shares of workers with different levels of completed education measure the skills of a region's workforce. The share of workers aged over 25 years was also included as a rough inverse proxy of the level of adaptability of the workforce. The characteristics of the unemployed, such as their previous work experience, were also added in initial experiments. Few were significant, perhaps surprisingly. The gap between demand and supply of skill, D ­ S, is measured by a variable we call skill mismatch. This is constructed by regressing the industrial structure of employment (in turn, the shares of industry and services in total employment) on the educational structure of the population. The fitted values from these regressions are interpreted as measures of the industrial structure supportable by the existing population. Skill mismatch is the sum of gaps between actual and fitted industrial shares. 10.60 We also include a set of variables reflecting other regional characteristics, such the existence of large conurbations, the share of the population living in rural areas and a variable Southeast which is a dummy variable suggested by the work of Gorzelak (2003). That paper demonstrates how the agricultural structure differs between the parts of Poland formerly under Austrian and Russian rule, with mostly small own-account farmers, and the formerly Prussian regions, where larger Junker farms were nationalized after the communist takeover. These farms have largely been privatized. Many, if not most, of the workers on these farms had little or no land of their own. Thus the ability to fall back on own-account agriculture in the absence of a job offer differs significantly between the two areas. 10.61 Some comments about the shock or residual in empirical versions of (3) are in order. These residuals are, by construction, all the influences on a region's unemployment rate not accounted for by skill supplies and demands. This could be quite a long list of variables, with the most obvious candidate being aggregate shocks. Herein lies a potential problem for interpreting and estimating empirical versions of (3). Shocks that are precisely orthogonal to the skill structures are unlikely. For example, the skill structure of employment may vary over the business cycle. In the past, researchers have more-or-less assumed that aggregate and structural effects on unemployment are orthogonal (see for example, the measures of structural unemployment in Layard (et al., Chapter 6). 10.62 Table 10.8 gives the results of estimating (2) using the panel of 49 old voivodships over five years (1994-98). Table 10.9 gives the estimates for the 16 voivodships over the period 1999- 2002. For both tables, in the first column, the estimation is by OLS. In the subsequent columns, fixed and random effects panel methods are applied. As expected, unemployment rates vary inversely with the educational level of the population. The missing educational category is tertiary education, thus, it should first be noted that the coefficients are all positive, implying that the tertiary educated population seems to generate lower unemployment rates. This is a very 230 consistent result. In the 1999-2002 data, a region's primary education share generates a larger rate of unemployment than the intermediate levels, although this is less true in the 1994-98 data. 10.63 It should be noted that, as we anticipated, the size of the effect is too large to be a pure natural rate phenomenon. Our calculations in section D suggested the coefficients of, at most, 0.05. All of our estimated coefficients are of magnitudes in the range 0.15-1.25. Clearly, regions with larger proportions of highly educated workers tend to have lower unemployment of all types of workers. This seems most likely to be a result of the long-term skill and capital shortages whose causes were discussed in section D under Aggregate Labor Demand and Endogenous Populations, but which, presently, defy empirical representation. Table 10.8: Estimates of the Influence of Structural Supply and Demand Features on Voivodship Unemployment Rates, Annual Panel of 49 Voivodships, 1994-98 (Dependent variable: unemployment rate) Method of estimation OLS Fixed Effects Random Effects Shares of population: Secondary educated 0.397 (2.2) 0.191 (1.2) 0.288 (1.9) Lower vocational educated 0.155 (1.2) 0.294 (2.1) 0.270 (2.1) Primary educated 0.324 (2.6) 0.172 (1.4) 0.285 (2.5) Aged 25+ 0.405 (4.0) 0.129 (1.3) 0.208 (2.3) Living in rural areas (RUR) -0.008 (0.4) dropped -0.019 (0.4) Large city in voivodship -0.028 (2.9) dropped -0.041 (2.4) Warsaw -0.022 (1.2) dropped -0.018 (0.8) Southeast 0.020 (1.3) dropped 0.030 (1.0) Southeast * RUR -0.046 (1.6) dropped -0.074 (1.2) Skill mismatch 0.162 (5.1) 0.065 (1.3) 0.086 (2.3) 1994 0.030 (4.7) 0.035 (7.4) 0.032 (5.3) 1995 0.039 (6.3) 0.041 (9.8) 0.039 (6.6) 1996 0.009 (1.6) 0.012 (3.0) 0.010 (1.9) 1997 0.007 (1.1) 0.006 (1.7) 0.006 (0.8) R2, Adj R2 0.54, 0.51 0.28 0.49 N 245 245 245 Notes: t-ratios in brackets. See text for discussion. Source: Authors' calculations from the PLFS data provided by the Central Statistical Office. 231 Table 10.9: Estimates of the Influence of Structural Supply and Demand Features on Voivodship Unemployment Rates, Half-yearly Panel of 16 New Voivodships, Autumn 1999-Spring 2002 (Dependent variable: unemployment rate) Method of estimation OLS Fixed effects Random effects shares of population: Secondary educated 0.72 (3.4) 0.49 (2.1) 0.65 (3.0) Lower vocational educated 0.45 (2.9) 0.47 (2.5) 0.39 (2.4) Primary educated 1.25 (8.9) 0.87 (4.3) 1.15(7.3) Living in rural areas (RUR) -0.082 (1.7) dropped -0.078 (1.1) Working in agriculture -0.46 (14.0) -0.44 (8.2) -0.45 (11.5) Southeast -0.67 (4.3) dropped -0.067 (2.8) Southeast * RUR 0.13 (3.0) dropped 0.12 (1.8) Skill mismatch 0.66 (7.0) 0.85 (7.9) 0.77 (7.8) Autumn 1999 -0.038 (6.3) -0.023 (3.8) -0.031 (5.3) Spring 2000 -0.032 (6.2) -0.019 (3.7) -0.026 (5.2) Autumn 2000 -0.033 (6.7) -0.023 (4.9) -0.028 (6.1) Spring 2001 -0.011 (2.7) -0.005 (1.3) -0.008 (2.0) Autumn 2001 -0.010 (2.5) -0.010 (2.8) -0.009 (2.5) R2, Adj R2 0.94, 0.93 0.73 0.97 N 96 96 96 Notes: t-ratios in brackets. See text for discussion. Source: Authors' calculations from the PLFS data provided by the Central Statistical Office. 10.64 Our attempt at constructing a mismatch variable, skill mismatch, is correctly signed and is usually significant. Other results are striking. Urban areas tend to have lower unemployment rates, as do areas where agriculture is still a major employer, at least in the 1999-2002 regression. Most agricultural workers are now self-employed own-account farmers. This activity is clearly a fallback for those without formal employment. The explanatory variables account for just over half of the voivodship unemployment rate variation in the earlier sample, and much more in the later sample. 10.65 An alternative way to demonstrate that regional educational levels affect unemployment though channels other than via individual characteristics is to model the probability of being unemployed as a function of other regional and individual characteristics. We report this using the November 1998 PLFS, with the old voivodship classification, in Table 10.10. 232 Table 10.10: Probit Analysis of Unemployment Status, November 1998 Explanatory variable coefficient Age -0.012** Aged 61 or more -0.35** Women 0.14** Head of household -0.37** Completed education level (default primary or less) university -0.85** post secondary -0.53** general secondary -0.38** vocational secondary -0.17** lower vocational -0.16** Regional variables proportion urban 0.059 Warsaw 0.015 proportion of pop. tertiary educated -3.48** proportion of pop. secondary educated -1.05 proportion of pop. lower vocational -3.07** proportion in of empl. in agriculture -1.32** Proportion of empl. in finance and real estate -3.48** Intercept 1.59** Pearson goodness of fit (p-value) 0.0 N 30568 % unemployed 11.8 **Indicates significance at the 1 percent level Source: PLFS. 10.66 In Table 10.10 we can see that age, marital status, gender and education all have important impacts on the probability of being unemployed rather than employed during the survey period. We add a set of voivodship-specific variables: proportion urban, a dummy for Warsaw, proportion of the population in three broad educational categories, and, after a little experimentation, two variables capturing industrial structure: the shares of employment in agriculture and the phases of employment in finance and real estate. On the last of these, as Deichmann and Henderson (2000) noted, services have become spatially concentrated. The presence of a significant financial and real estate service sector surely signals a region of economic expansion. As has been mentioned, agriculture sometimes acts as fallback employment. The regional education proportions attract uniformly large and negative coefficients, with two of the three being statistically significant. F. CONCLUSIONS 10.67 This chapter investigates the relationship between the evolving pattern of regional unemployment in Poland since the early 1990s and regional variations and imbalances in the supply and demand for skill. There have been massive changes in the demand for and supply of skill over the period, and these changes are by no means uniform across the Polish regions. We demonstrate that lower-skilled workers tend to generate greater inflows into unemployment, and, via that route, higher unemployment rates. Such unemployment differences by skill are more-or- 233 less universal and are usually attributed to skill differentials in non-wage labor costs, particularly adjustment costs. 10.68 We show that, with some regional immobility of labor, differences in regional skill mixes will generate different regional rates of unemployment. We also find that part of the regional skill-unemployment relationship can be attributed to the excess demand for highly educated labor. We use the labels natural rate and skill mismatch unemployment for these two aspects of the regional distribution. However, as our calculations show, the inverse relationship between education and unemployment is too large for these to be the full explanation. 10.69 There is an embarrassment of riches when it comes to theories of long-term regional unemployment and skill differences. We outline the possibilities in section D, Aggregate Labor Demand and Endogenous Populations: agglomeration effects, transportation advantages, differences in welfare payment systems, sorting effects where high-skilled workers are reluctant to move to low-skill regions, and regional differences in the quality of education. All warrant serious consideration, but are hard to test individually with the existing data. This is partly because most indicators of economically advantaged regions are strongly correlated with education levels, especially tertiary education. 10.70 On the brighter side, educational attainment has improved considerably in Poland-- although still not sufficiently--and an important part of the regional distribution of unemployment appears to be caused, as one authority (Tinbergen, 1975) put it, by "the race between technological development and education." 234 REFERENCES Arulampalam, W., and M. B. Stewart (1995), "The Determinants of Individual Unemployment Durations in an Era of High Unemployment," The Economic Journal, No. 105, March. Boeri, T., and P. Garibaldi (2003), "Dealing with an Increasingly Stagnant Unemployment Pool," mimeo, Universita Bocconi, July 2003, Paper prepared for the World Bank Living Standards Assessment. Carey, David, and Harry Tchilinguirian (2000), "Average Effective Tax Rates on Capital, Labor and Consumption," OECD Economics Department Working Papers No. 258, October. Deichmann, U., and V. Henderson (2000), "Urban and Regional Dynamics in Poland," World Bank Research Working Paper No. 2457, September, available at: http://econ.worldbank.org. Duncan, O. D., and B. Duncan (1955), "A Methodological Analysis of Segregation Indices," American Sociological Review, 20: 210-217. Faggio, G., and J. Konings (2001), "Job Creation, Job Destruction and Employment Growth in Transition Countries in the 90s," IZA Discussion Paper No. 242, IZA, Bonn. Fernandez, R., and R. Rogerson (2001), "Sorting and Long-run Inequality," Quarterly Journal of Economics, CXVI, November 2001, 1305-1342. Filer, R., D. Hamermesh and A. Rees (1996), The Economics of Work and Pay (6th ed.), New York: HarperCollins. Fujita, M., P. Krugman, and A. Venables (1999), The Spatial Economy: Cities Regions and International Trade, Cambridge MA: MIT Press. Gorzelak, G., with M. Smetkowski (2003), "The Regional Differentiation in Poland," mimeo, Center for European Regional and Local Studies, University of Warsaw, July 2003, Paper prepared for the World Bank Living Standards Assessment. GUS (2002), "Labor Force Survey in Poland in the Years 1992-2001," Central Statistical Office, Warsaw. GUS (2003), "Labor Force Survey in Poland. IV Quarter 2002," Central Statistical Office, Warsaw. Kwiatkowski, E., M. W. Socha, and U. Sztanderska (2001), "Labor Market Flexibility and Employment Security, Poland," ILO Employment Paper 2001/28, ILO, Geneva. Layard, R., S. Nickell, and R. Jackman (1991), Unemployment: Macroeconomic Performance and the Labor Market, Oxford: Oxford University Press. MLSP (2000), "Basic Statistical Data on Social Policy," Ministry of Labor and Social Policy, Warsaw, November. 235 Newell, A. (2001), "Why Have a Million More Polish Workers Become Unemployed in the Midst of an Economic Boom?" mimeo, University of Sussex, July 2001. Paper prepared for the ILO/Ministry of Labour conference on Poland's unemployment, July 2001. Newell, A., and F. Pastore (2000), "Regional Unemployment and Industrial Unemployment in Poland," IZA Discussion Paper No. 194, August, IZA, Bonn. Newell, A., and M. Socha (1998), "Wage Distribution in Poland: The Roles of Privatization and Changes in International Trade, 1992-1996," Economics of Transition, 6(1). Newell, A., and M. Socha, (2002), "The Rising Non-manual Wage Premium in Poland," mimeo, University of Sussex Discussion Papers in Economics, No 89, October, 2002. Newell, A., and M. Socha, (2003), "The Distribution of Wages in Poland, 1992-2002," mimeo, University of Sussex, July 2003, Paper prepared for the World Bank Living Standards Assessment. Newell, A., and J. Symons (1991), "Endogenous Separations in a Matching Model," LSE Center for Economics Performance, Discussion Paper no. 35, June. OECD (2001), "Taxing Wages," OECD Paris. Overman, H., and D. Puga (2002), "Regional Unemployment Cluster," Economic Policy, 34, Spring, pp.116-147. Riboud, M., C. Sainchez-Piramo, and C. Silva-Jauregui (2002), "Does Eurosclerosis Matter? Institutional Reform and Labor Market Performance in Central and Eastern European Countries in the 1990s," World Bank SP Discussion Paper No. 0202, March 2002. Rutkowski, J. (1996), "High Skills Pay Off: The Changing Wage Structure During the Transition in Poland," Economics of Transition, 4, pp.89­111. Socha, M. W., and J. Weisberg (2002), "Labor Market Transition in Poland: Changes in the Public and Private Sectors," International Journal of Manpower, Vol. 23 No. 6. Tinbergen, J. (1975), Income Distribution, Analysis and Policies, North-Holland, Amsterdam. World Bank (2001), Poland's Labor Market: The Challenge of Job Creation, World Bank, Washington, October. 236 Annex Table A.10.1: Macroeconomic Data 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 percent change in real terms GDP -11.6 -7 2.6 3.8 5.2 7 6 6.8 4.8 4.1 4.0 1.0 1.4 Domestic demand n.a n.a n.a 6.1 4.8 4.5 12.2 9.4 6.4 4.8 2.8 -1.7 0.8 Private consumption -15.3 6.3 2.3 5.2 4.4 3.3 8.3 6.9 4.8 5.2 2.8 2.0 3.3 Government consumption 0.5 10.2 6.4 3.8 2.2 2.9 3.4 3.2 1.4 1.0 1.1 0.4 1.4 Gross capital formation -10.6 -4.4 2.3 12.8 9.1 10.5 34 20.9 14,2 6.8 2.7 -8.8 -7.2 Fixed investment -10.1 -4.1 0.4 2.9 9.2 16.5 19.7 21.7 14.2 6.8 2.7 -8.8 -7.2 External demand (net exports; contribution to GDP growth) n.a n.a n.a -1.6 0.6 0.2 -3.4 -2.7 -1.8 -1.2 1.3 3.1 0.6 Exports of goods and nonfactor services n.a -1.7 10.8 3.2 13.1 22.8 12 12.2 11 -2.6 23.2 10.2 5.0 Imports of goods and nonfactor services n.a 29.6 1.7 13.2 11.3 24.2 28 21.4 15.8 1.0 15.6 -0.1 3.4 Industrial production 2.8 6.4 12.1 9.7 8.3 11.5 4.6 4.8 6.7 0.6 1.4 Average total employment 16280 15326 14677 14330 14475 14735 15017 15439 15799 15373 15017.5 14923.6 14784 - - 237 Aggregate labor productivity (GDP/employment) 7.7 1.20 7.1 6.3 4.1 5.1 4.0 3.9 2.4 7.0 6.5 1.6 N2.4 GDP deflator 480.1 55.3 38.5 30.5 28.4 27.9 18.7 14 11.8 6.9 7.1 4.2 n.a CPI (percentage change from previous period) 585.8 70.3 43 35.3 32.2 27.8 19.9 14.9 11.8 7.3 10.1 5.5 1.9 Compensation rate of the business sector n. a. n. a. n. a. n. a. 45.1 30.8 29.4 20.5 15.3 14.1 9.7 8.5 3.5 Labor productivity in business sector n. a. n. a. n. a. n. a. 8.8 7.1 5.5 6.1 4 9.3 6.4 3.8 4.8 Current account - as a percentage of GDP n. a. n. a. n. a. -5.2 1 0.7 -2.3 -4 -4.4 -8.1 -6.3 -3 -3.3 Short-term interest rate n. a. n. a. n. a. 34.9 31.8 27.7 21.3 23.1 19.9 14.7 18.9 15.7 8.8 Real IRS (GDP deflator) n. a. n. a. n. a. 1.03353 1.0261 0.998 1.022 1.0799 1.0725 1.0728 1.1099 1.1106 n.a. Real IRS (CPI deflator) n. a. n. a. n. a. 0.99686 0.99665 0.999 1.012 1.0715 1.0725 1.0688 1.0797 1.0969 1.068106 Real Effective Exchange Rate BASED ON REL.CP 51.3 80.26 85.37 91.63 92.4 100 108.78 111.39 116.96 112.29 121.62 138.3 n.a. Government net lending - as a percentage of GDP n.a. n.a. n.a. -4.5 -3.5 -2.5 -2.9 -2.8 -2.3 -2.0 -3.1 -5.5 -6.0 468 1.656 3.340 5.083,0 8.036,0 14.027,7 20.587,7 30.651,2 38.912,6 49.392,5 56.833,5 Cumulative value of inward FDI (million USD) n. a. 61.600a Notes to Table A.10.1 as of June 2002. Source: IFS; IMF Staff Country Report No 00/61, Republic of Poland: Statistical Appendix, April 2000; IMF Country Report No. 02/128, Republic of Poland: Selected Issues and Statistical Appendix, June 2002; IMF Country Report No. 03/188, Republic of Poland: Selected Issues, June 2003; Poland. Quarterly Statistics, Central Statistical Office, Warsaw, various years; Statistical yearbook of the Republic of Poland; various years; OECD Economic Outlook No. 72, OECD Paris 2002 and authors' calculations. Figure A.10.1: Poland's Phillips Curve 80 60 rate 40 inflation20 238 0 0 5 10 15 20 unemployment rate Source: See Table A.10.1. Figure A.10.2: Poland's Beveridge Curve 0.0035 0.003 0.0025 rate 0.002 0.0015 vacancy 0.001 0.0005 0 239 0 5 10 15 20 unemployment rate Source: See A.10.1 Table A.10.2: Polish Labor Market 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Population in thousands persons 38183 38309 38418 38505 38581 38609 38639 38660 38667 38654 38644 38632.00 38300.00 Employed (at the end of December) in thousand persons 16485 15772 15357 15118 15282 15486 15842 16229 16174 15919 15480 14988.00 14858.00 of which private sector in thousands persons 7902 8390.5 8404.5 8700.9 9045.6 9506 10075 10868 11249 11352.9 11170.7 10967.90n.a. Foreign ownership in thousands persons 167.4 89.0 119.9 150.0 228.10 269.4 337.7 399.7 484.2 566.8 602.0 621.40n.a. Share of private sector 0.48 0.53 0.55 0.58 0.59 0.61 0.64 0.67 0.70 0.71 0.72 0.73 Registered unemployment (at the end of December) in thousand persons 1126 2156 2509 2890 2838 2629 2360 1826 1831 2350 2703 3115.00 3217.00 Unemployment rate (at the end of December) in % 6.50 12.20 14.30 16.40 16.00 14.90 13.20 10.30 10.40 13.10 15.10 17.50 18.10 Average gross nominal earnings 289.73 390.43 525.02 690.92 874.30 1065.8 1232.7 1697.1 1893.7 2045.12133.0 Average gross real earnings (percentage change from previous period) 24.40 -0.30 -2.70 -0.30 1.70 2.80 5.50 5.90 3.30 4.70 1.00 2.50 Number of reported job offers (annual averages) in thousands n.a. 483.80 527.80 801.00 914.20 878.90 915.80 761.10 680.70 607.90 465.70 555.60 Reported vacancies (the end of December) in thousands 54.10 29.10 22.90 21.70 25.20 20.50 13.80 11.90 7.30 6.10 5.70 5.30 8.30 Registered unemployed benefited from selected ALMP (in thousands persons) 106.90 54.00 210.10 279.10 402.30 378.90 324.50 485.10 414.40 367.90 296.90 141.40 153.40 240 Unemployed in training in thousands persons 18.00 70.40 72.20 96.10 81.80 85.60 143.50 139.00 128.40 98.70 47.60 68.60 Unemployed in subsidized jobs in thousands persons 106.90 36.00 104.50 132.90 195.50 184.00 139.10 166.20 142.90 125.90 99.40 39.30 51.10 Unemployed public works 35.20 74.00 110.70 113.10 99.80 149.90 104.10 68.70 50.30 29.00 33.70 Unemployed on internship with the employer 25.50 28.40 44.90 48.50 25.50n.a. Share of Unemployed without unemployment benefits in % 20.80 21.00 47.70 51.80 49.80 41.10 48.20 69.50 77.10 76.40 79.70 80.00 83.30 Ratio of unemployment benefit to average wage 19.60 34.20 37.90 36.00 37.00 36.70 33.40 32.00 30.30 23.70 22.7 N22.7 n.a. Ratio of unemployment benefit to minimum wage 81.20 118.20 82.40 82.30 82.30 84.60 78.80 75.50 58.20 60.60 58.06 58.05n.a. Ratio of minimum monthly net wage to average wage 37.50 41.00 41.00 41.20 40.50 40.00 40.20 38.30 37.00 37.20n.a. Ratio of pre-retirement benefit to average wage 0.31 0.31n.a. ratio of pre-retirement allowance to average wage 0.52 0.53n.a. Ratio of pre-retirement benefit to minimal wage 0.85 0.84n.a. ratio of pre-retirement allowance to minimal wage 1.41 1.44n.a. Source: Statistical Yearbook of the Republic of Poland, Central Statistical Office, Warsaw, various years; and Basic Statistical Data on Social Policy, Ministry of Labour and Social Policy, Warsaw, various issues. 11. THE DISTRIBUTION OF WAGES IN POLAND, 1992-2002 Andrew Newell and Mieczyslaw W. Socha A. BACKGROUND 11.1 This paper describes and analyzes the changes that the transition has brought about in the distribution of wages in Poland in the decade 1992-2002. Our analysis thus complements previous work, such as Keane and Prasad (2002), who cover the early transition period up to 1996. We find little change in overall wage inequality from 1992 to 2000, and then an increase in 2001 and 2002. The lack of change in wage inequality up to 2000 was perhaps unexpected. Our analysis, together with the work of Newell (2001a), suggests that two main forces were at work, pulling wage distribution in different directions. First, the increasing share of employment in the private sector has been a force in raising wage inequality. Occupational wage premiums are higher and are growing further apart in Poland's private sector. Alternatively, we demonstrate how the private sector tends to pay less at the low wage end of the labor market and more at the high wage end. 11.2 Second, the main force contracting wage inequality through the 1990s was the reduction in lower skilled workers in employment. This came about because first, young people were staying on longer in education, and second, a less educated older generation was going into retirement. These participation changes slowed down at the turn of the century, and we speculate that the rise in wage inequality that took place in 2001 and 2002 was a result of this slowdown. 11.3 Section B provides the macroeconomic and institutional context for our study. Section C introduces the Polish Labor Force Survey (PLFS) data sets and illustrates some of the trends discussed above, and section D reports the results of wage equation estimation, detailing initially how the responsiveness of wages to individual characteristics has changed through Poland's transition. We then report how quantile regression analysis and correcting for participation bias alter and enrich the picture of an evolving wage distribution. Section E draws conclusions on the basis of the discussion in the chapter. B. MACROECONOMIC AND INSTITUTIONAL CONTEXT Macroeconomic Overview of the Labor Market 11.4 The highlights to take into account are the period of rapid growth during 1993-98, which was characterized by steady employment levels and falling unemployment, and the ensuing slowdown of growth, which was accompanied by falling employment and rising unemployment. The shape of wage distribution may have been affected by this cycle of economic activity, but more profound effects would have been generated by structural changes in the labor market. (A 241 detailed discussion of macroeconomic developments in Poland since the early 1990s is contained in Newell and Socha, 2003.) 11.5 Four main structural changes on the demand side can be identified. First, there has been a large shift in labor demand away from lower skilled and toward higher skilled employees. In 1992-2002, the number of employed with a tertiary level of education increased by 52.3 percent, while workers with primary and less education decreased by 53 percent. The employment rate for women at the latter level of education living in urban areas dropped to 14.4 percent. Previous studies of wage distribution revealed increases in wage premiums to higher education, to skilled occupations, and to private sector employees (for details see Rutkowski, 1996; Newell 2001b; and Newell and Socha, 1998 and 2002). 11.6 Second, owing fundamentally to the industrial bias of the communist system, large changes are taking place in the industrial structure of employment. Labor Force Survey data show that since 1994 employment in agriculture, hunting, and forestry fell by 25 percent.10 In mining, the fall was 50 percent, and in manufacturing the fall has been about 19 percent since 1994. Employment in real estate and business activities increased by 180 percent, in hotels and restaurants by 47 percent, and in trade by 10 percent. However, since 1998, there has been a reversal of the trend with respect to rural activities, and the share of agriculture, forestry, and hunting in total employment has increased to 18.5 percent. 11.7 Third, the driving force in job creation is the development of the private sector. The process of privatization has slowed down recently, but in 2002 the share of the private sector in total employment was 73 percent,11 in industrial output it was 75 percent, in investment, 72 percent, in exports, 88 percent, and in imports, 92 percent. The private sector is dominated by small and medium-size businesses.12 Through the 1990s, the private sector attracted a greater proportion of young, male, and less educated workers than the public sector (Socha and Weisberg, 2002). Between 1992 and 2002, the public sector lost 3.1 million workers (41 percent), while employment in the private sector increased by 21 percent. 11.8 Fourth, there is evidence of the growing importance of atypical forms of employment. The share of part-time workers is very stable at the level of 10 percent; however, the share of temporary paid workers in the total number of paid employees increased between 1992 and 2001 from 2.9 percent to 6 percent. This may have been a response to the increased levels of statutory employment protection. In addition, the Central Statistical Office's estimates of employment in the shadow economy suggest an increase from 805,000 in 1995 to 895,000 in 2001. Changes in Labor Institutions 11.9 Despite considerable effort, the structural reform of the labor market has not been completed (for details see Kwiatkowski et al., 2001). Instead, a system characterized by high levels of labor taxation, relatively generous welfare assistance, and employment protection still exists. 10Data for private agriculture estimated from other sources show a different trend (i.e., an increase in the number of employed in agriculture, forestry, and hunting of 6.2 percent with a share in total employment of 29.2 percent. 11Including private farmers. 12The number of individual non-incorporated businesses has increased since 1992-93 by 77 percent and in 2001 exceeded 2.66 million establishments. 242 11.10 Labor Market Regulation. The main regulation changes that are likely to affect relative wages are in the 2002 revision of the labor regulations, which was intended to lower labor costs and increase labor market flexibility (for a detailed discussion, see Newell and Socha, 2003). The main changes are the following: · Very small firms (20 or fewer employees) do not have to establish wage rules and work codes. · Until Poland becomes a member of the EU, an unlimited number of successive fixed term contracts are permitted before a worker is automatically deemed a permanent employee. · The minimum overtime premium has been lowered. · Severance payments for collective dismissals in small firms have been abolished and severance pay in other firms has been linked to current tenure rather than to total years of work experience. The level of severance pay varies from one monthly wage for workers with two years' tenure to three monthly wages for workers with eight years' tenure. 11.11 Wage Taxation. Compared to other transitional and OECD countries, Poland has one of the highest payroll taxes (48 percent) with a total tax rate (income and consumption taxes included) of about 80 percent (Riboud et al., 2002).13 This high tax burden is likely to discourage both labor supply and job creation. To illustrate the scale of taxation, employer contributions calculated as a percentage of gross wage in Poland are the second highest in the OECD, after the Netherlands (OECD, 2001, chart 2, p. 105). 11.12 The major change in the system of labor taxation is linked to the social security and health care systems reforms that came into force in 1999. The social insurance reform has not changed the rate of the social contribution, but it has introduced a division between employer and employee contributions. The employer's obligatory social contribution is 20.43 percent of gross wages. This includes a social security contribution of 17.9 percent (9.76 percent for pension, 6.5 percent for disability, and 2.45 percent for industrial injuries insurance); a 2.45 percent contribution to the Labor Fund, and a 0.08 percent contribution to the Benefits Guarantee Fund. The employee pays 18.71 percent of gross wages in the form of pension insurance (9.76 percent), disability insurance (6.5 percent) and sickness insurance (1.62 percent). The contribution to health insurance amounts to 7.75 percent of individual income. Additionally, some measures were taken to reduce the abuse of disability and sickness benefits. The perspectives for lowering this fiscal burden on employment are rather gloomy, as reductions in payroll taxes (without cuts in budgetary expenditures) would increase the budget deficit. 11.13 The Wage Bargaining System. After restrictive anti-inflation wage taxes were abolished in 1995, the wage bargaining system was organized through the Tripartite Commission. Wage negotiations between employers and workers can take place at the industry level and the firm level. Local governments, employers, and trade union organizations are not involved in wage bargaining. Since the collective agreement can be signed only in enterprises with trade unions, this form of industrial relations regulates the wages in state enterprises (or recently privatized large firms) and in public non-profit entities. There are no reliable data on unionization. However (according to Riboud et al., 2002), Poland has the lowest union density of the accession countries (33.8 percent of salaried employees, compared to 39.6 percent in OECD countries and 44.4 percent in EU countries). In the private sector, which is populated by small and medium-size companies without trade unions, wages are set by employers. There are also a few supra- 13Carey and Tchilinguirian (2000) estimated the average effective tax rate for labor in the period 1991-95 as 42.06 percent. 243 enterprise agreements. The Ministry of Labor, the unions, and employers' organizations negotiate an annual wage increase in the public sector and revise the national minimum wage. 11.14 There are two main problems concerning the present wage bargaining system. The first problem is that there are a large number of trade unions and employers' organizations within the enterprises and at the national level. This makes it difficult to establish a clear representation of both workers' and employers' interests, and impedes coordination during wage bargaining. Riboud et al. (2002) construct a 3-point scale for the degree of union and employer coordination. For Poland, union coordination is estimated at 2 points and coordination among employers at 1 point. 11.15 The second problem concerning the wage bargaining system is that owing to the political rivalry between competing union confederations and the low level of confidence between the social partners, the credibility of the Tripartite Commission is very low. Attempts to implement the government's Social Pact Agreement with the employers and trade unions have failed so far. Even agreements about the changes in the average wages in the public sector have not always been reached in every year, but have had at times to be settled by government decree. 11.16 Minimum Wage. The minimum wage is based on monthly remuneration and covers all sectors and all types of workers. It has been revised three to five times a year in line with inflation. The ratio of the minimum wage to the average wage decreased slightly during recent years from 40 to 37 percent. The differentiation of minimum wages by locality and/or age had been proposed by the government for some time, and in January 2003, workers in their first year of work should have received at least 80 percent of the standard minimum wage, while those in the second year of work were protected by a minimum wage at 90 percent of the standard. C. PLFS DATA 11.17 Table 11.1 gives some wage inequality statistics from 12 rounds of the Polish Labor Force Survey (PLFS) covering the last decade. These 12 rounds will be employed throughout this section and the next one. There was little change in wage inequality through the 1990s. It seems, however, that there was a moderate increase in wage inequality after 2000. For example, the 90/10 decile ratio moves very little until it jumps in the spring of 2001; thereafter it remains high. Similarly, the Gini coefficient and the coefficient of variation rise (respectively, from about 24 percent to about 26 percent and from about 6 percent to about 7 percent) at the beginning of 2001. Two questions arise. First, what caused this increase in wage inequality? Second, is this rise likely to be permanent? A detailed look at the wages in the surveys shows that the increased inequality derives mostly from a moderate increase in the numbers of workers with low wages. We have mentioned that there has been a drop in the minimum wage to average wage ratio in the last few years, and this might be one cause of the increase in inequality. Other possible causes include the rise of temporary contracts and also the rise in the share of employment in small firms. In our regression work reported below we show that as the variance of wages increases in the later rounds of the survey, so does the coefficient of determination (R2). As a consequence, the regression work might be able to point to some of the systematic processes at work. We will return to this issue in our conclusions. 244 Table 11.1: Monthly Wage Distribution, 1994-2002 90/10 90/50 50/10 Gini Coefficient of variation decile decile decile of log wages ratio ratio ratio Autumn 1994 2.81 1.84 1.52 0.238 0.051 Autumn 1995 2.62 1.81 1.45 0.236 0.066 Autumn 1996 2.71 1.83 1.49 0.238 0.064 Autumn 1997 2.62 1.69 1.55 0.236 0.062 Spring 1998 2.58 1.71 1.63 0.235 0.061 Autumn 1998 2.50 1.79 1.56 0.231 0.060 Autumn 1999 2.88 1.87 1.54 0.238 0.061 Spring 2000 2.68 1.76 1.70 0.240 0.060 Autumn 2000 2.50 1.67 1.50 0.240 0.061 Spring 2001 2.91 1.78 1.64 0.270 0.072 Autumn 2001 2.76 1.78 1.55 0.257 0.070 Spring 2002 2.86 1.78 1.61 0.264 0.072 Source: Polish Labor Force Survey. 11.18 Table 11.2 gives summary statistics on the population of working age in the PLFS. Some clear trends emerge. First, there is a steady but inevitably slow increase in the preponderance of more highly educated workers. This has been noted many times by previous researchers. Second, there is a fall in participation. By adding the unemployed and employed proportions we can see that there has been a significant reduction in the overall participation rate among PLFS respondents, from around 70 percent in the first half of the 1990s to between 66 and 67 percent since 1998. This fall partly reflects the greater time spent in education by younger people, but it also reflects earlier retirement. Table 11.2: Summary Statistics for the Population of Working Age, 1992-2002 Proportion who are: 1992r4 1993r4 1994r4 1995r4 1996r4 1997r4 1998r2 Women 0.4876 0.4826 0.4914 0.4946 0.4917 0.4889 0.4875 Household head 0.3958 0.3779 0.3835 0.3774 0.3711 0.3697 0.3610 Married 0.6724 0.6652 0.6675 0.6553 0.6510 0.6430 0.6296 Ed-University 0.0729 0.0564 0.0729 0.0727 0.0746 0.0759 0.0760 Ed-Post Sec. 0.0326 0.0217 0.0274 0.0279 0.0266 0.0260 0.0266 Ed-General Sec. 0.1956 0.1727 0.1959 0.1998 0.2054 0.2083 0.2064 Ed-Vocational Sec. 0.0683 0.0634 0.0699 0.0695 0.0707 0.0724 0.0745 Ed-Lower Vocational 0.2941 0.3112 0.3020 0.3070 0.3088 0.3116 0.3058 Ed-Primary 0.3188 0.3564 0.3179 0.3103 0.3027 0.2956 0.2831 Employee 0.4365 0.3770 0.4135 0.4179 0.4243 0.4315 0.4308 Unemployed 0.0920 0.1105 0.1044 0.1024 0.0892 0.0766 0.0783 Employed 0.6102 0.6097 0.5861 0.5833 0.5881 0.5935 0.5903 Aged 15 to 20 0.1504 0.1592 0.1623 0.1660 0.1644 0.1620 0.1631 Aged 21 to 25 0.0971 0.1037 0.0993 0.1058 0.1115 0.1127 0.1144 Aged 26 to30 0.1005 .0985 0.0908 0.0900 0.0918 0.0923 0.0919 Aged 31 to35 0.1250 0.1178 0.1103 0.1058 0.0987 0.0971 0.0926 Aged 36 to 40 0.1397 0.1318 0.1370 0.1328 0.1242 0.1208 0.1111 Aged 41 to 45 0.1257 0.1222 0.1318 0.1301 0.1294 0.1288 0.1346 Aged 46 to 50 0.0777 0.0851 0.0943 0.1040 0.1151 0.1177 0.1220 Aged 51 to 55 0.0799 0.0779 0.0730 0.0696 0.0687 0.0736 0.0823 Aged 56 to 60 0.0741 0.0734 0.0721 0.0676 0.0687 0.0670 0.0633 Sample Size 37486 66835 43666 44374 44631 44326 44120 Proportion who are: 1998r4 1999 2000r2 2000r4 2001r2 2001r4 2002r2 Women 0.4891 0.4893 0.4886 0.4916 0.4897 0.4914 0.4907 Household head 0.3662 0.3632 0.3556 0.3577 0.3583 0.3564 0.3544 245 Married 0.6331 0.6299 0.6158 0.6182 0.6120 0.6130 0.6060 Ed-University 0.0803 0.0792 0.0798 0.0828 0.0833 0.0902 0.0883 Ed-Post Sec. 0.0255 0.0257 0.0247 0.0263 0.0290 0.0285 0.0296 Ed-General Sec. 0.2092 0.2170 0.2104 0.2144 0.2066 0.2103 0.2092 Ed-Vocational Sec. 0.0719 0.0774 0.0778 0.0793 0.0800 0.0815 0.0863 Ed-Lower Vocational 0.3136 0.3106 0.3132 0.3098 0.3137 0.3173 0.3102 Ed_Primary 0.2904 0.2825 0.2761 0.2873 0.2858 0.2686 0.2726 Employee 0.4331 0.3994 0.3918 0.3926 0.3809 0.3744 0.3649 Unemployed 0.0816 0.1178 0.1233 0.1194 0.1383 0.1350 0.1463 Employed 0.5887 0.5519 0.5452 0.5446 0.5340 0.5236 0.5154 Aged 15 to 20 0.1666 0.1663 0.1628 0.1692 0.1565 0.1626 0.1496 Aged 21 to 25 0.1145 0.1129 0.1167 0.1142 0.1161 0.1167 0.1201 Aged 26 to30 0.0935 0.0938 0.0987 0.0945 0.1003 0.0988 0.1013 Aged 31 to35 0.0933 0.0872 0.0870 0.0870 0.0863 0.0884 0.0884 Aged 36 to 40 0.1165 0.1084 0.1010 0.1021 0.0952 0.0953 0.0882 Aged 41 to 45 0.1297 0.1307 0.1283 0.1305 0.1240 0.1216 0.1187 Aged 46 to 50 0.1177 0.1208 0.1264 0.1195 0.1272 0.1254 0.1287 Aged 51 to 55 0.0785 0.0907 0.0966 0.0992 0.1093 0.1057 0.1144 Aged 56 to 60 0.0621 0.0626 0.0568 0.0583 0.0601 0.0599 0.0655 Sample Size 43799 37039 36894 38398 37553 37776 38132 Source: Authors' calculations from the PLFS. 11.19 Third, there are falls in the proportions of heads of household and married people in the survey. Both of these phenomena are likely to be due in part to the longer time young people are spending in education. Finally, two demographic effects are clear. First, the post-World War II baby boomers passed from around 40 years old in 1992 to around 50 years old in 2002. Second, the early 1980s baby boomers reached working age. 11.20 Table 11.3 gives descriptive statistics of the samples of employees in the 12 rounds of the PLFS that we used for wage equation estimation. Starting at the top of the table and moving down, there are a number of trends that deserve mention. The share of women among employees has risen, at the same time that, following the population trend, the share of married people in employment has fallen. The proportion of employees with more advanced education has increased rapidly; indeed, it has increased much more rapidly than in the population of working age in general. In industry, mining has contracted and other manual-intensive industries such as manufacturing and the power utilities show mildly falling shares. The share of workers in construction has followed a cyclical path: booming in the late 1990s and falling after 1999. The growth sectors are trade and repair, and financial services. Table 11.3: Descriptive Statistics for Employees, 1994-2002 Proportion who are: 1994 r4 1995 r4 1996 r4 1997 r4 1998 r2 1998r4 Women 0.4664 0.4680 0.4597 0.4607 0.4589 0.4688 Household head 0.4957 0.4891 0.4812 0.4743 0.4710 0.4652 Married 0.7833 0.7727 0.7641 0.7546 0.7543 0.7489 Ed-University 0.1264 0.1275 0.1293 0.1309 0.1313 0.1341 Ed-Post Sec. 0.0472 0.0474 0.0439 0.0428 0.0424 0.0423 Ed-General Sec. 0.2663 0.2679 0.2710 0.2746 0.2816 0.2790 Ed-Vocational Sec. 0.0688 0.0640 0.0639 0.0639 0.0651 0.0623 Ed-Lower Vocational 0.3491 0.3589 0.3656 0.3680 0.3672 0.3679 Ed-Primary 0.1414 0.1337 0.1250 0.1191 0.1116 0.1142 Agriculture 0.0351 0.0321 0.0296 0.0298 0.0319 0.0287 Mining 0.0484 0.0499 0.0459 0.0393 0.0350 0.0328 Manufacturing 0.2914 0.2903 0.2843 0.2805 0.2822 0.2774 246 Utilities 0.0256 0.0247 0.0257 0.0259 0.0242 0.0230 Construction 0.0706 0.0703 0.0745 0.0778 0.0778 0.0805 Trade and repair 0.1079 0.1175 0.1195 0.1269 0.1310 0.1404 Transport 0.0699 0.0721 0.0741 0.0737 0.0771 0.0719 Finance and real estate 0.0489 0.0497 0.0515 0.0556 0.0546 0.0574 Public services and defense 0.3006 0.2920 0.2947 0.2903 0.2859 0.2876 Other services 0.0017 0.0013 0.0002 0.0002 0.0002 0.0003 Hold temporary jobs 0.0295 0.0285 0.0300 0.0340 0.0287 0.0357 Private Sector 0.2506 0.2963 0.3386 0.3793 0.3959 0.4079 Firm size 1 to 5 0.0780 0.0836 0.0860 0.0938 0.0956 0.1029 Firm size 6 to 20 0.1746 0.1753 0.1766 0.1808 0.1856 0.1883 Firm size 21 to 50 0.1671 0.1673 0.1717 0.1681 0.1729 0.1644 Firm size 51 to 100 0.1290 0.1258 0.1293 0.1219 0.1145 0.1204 Firm Size over 100 0.4513 0.4480 0.4364 0.4064 0.3959 0.3839 Professional 0.1262 0.1214 0.1193 0.1200 0.1204 0.1194 Managerial 0.0486 0.0498 0.0445 0.0424 0.0416 0.0412 Technical 0.1594 0.1531 0.1569 0.1517 0.1527 0.1517 Clerical 0.0934 0.0976 0.0987 0.1025 0.1047 0.1035 Sales 0.0811 0.0871 0.0936 0.0941 0.0968 0.1009 Farm workers 0.100 0.0085 0.0071 0.0074 0.0095 0.0082 Skilled manual 0.2628 0.2654 0.2659 0.2617 0.2530 0.2557 Semi-skilled manual 0.1053 0.1061 0.1087 0.1142 0.1160 0.1136 Unskilled 0.1133 0.1111 0.1053 0.1061 0.1054 0.1059 Has a recent spell unemployed 0.1638 0.1834 0.1926 0.2136 0.2155 0.2230 Full-time student 0.0012 0.0031 0.0036 0.0033 0.0033 0.0048 Under 5 years experience 0.1171 0.1231 0.1310 0.1402 0.1283 0.1461 5 to 10 years experience 0.1267 0.1288 0.1330 0.1291 0.1310 0.1287 10 to 20 years experience 0.3255 0.3146 0.2968 0.2895 0.2742 0.2779 Over 20 years experience 0.4307 0.4335 0.4392 0.4412 0.4664 0.4473 1 to 5 years tenure 0.2273 0.2459 0.2501 0.2621 0.2759 0.2808 5 to 10 years tenure 0.1801 0.1776 0.1785 0.1803 0.1748 0.1761 Over 10 years tenure 0.4431 0.4224 0.4162 0.3986 0.3860 0.3909 Proportion who are: 1999 2000 r2 2000 r4 2001 r2 2001 r4 2002 r2 Women 0.4611 0.4679 0.4652 0.4761 0.4730 0.4844 Household head 0.4861 0.4951 0.4900 0.4978 0.4955 0.4984 Married 0.7537 0.7525 0.7512 0.7434 0.7457 0.7382 Ed-University 0.1422 0.1495 0.1462 0.1513 0.1587 0.1535 Ed-Post Sec. 0.0420 0.0402 0.0435 0.0455 0.0443 0.0508 Ed-General Sec. 0.2894 0.2815 0.2850 0.2738 0.2732 0.2813 Ed-Vocational Sec. 0.0723 0.0662 0.0694 0.0720 0.0697 0.0774 Ed-Lower Vocational 0.3509 0.3595 0.3555 0.3515 0.3570 0.3368 Ed-Primary 0.1030 0.1028 0.1004 0.1058 0.0967 0.0998 Agriculture 0.0300 0.0274 0.0270 0.0243 0.0244 0.0272 Mining 0.0209 0.0232 0.0229 0.0206 0.0203 0.0207 Manufacturing 0.2762 0.2739 0.2779 0.2793 0.2697 0.2669 Utilities 0.0243 0.0253 0.0249 0.0262 0.0252 0.0229 Construction 0.0790 0.0815 0.0782 0.0722 0.0743 0.0631 Trade and repair 0.1344 0.1400 0.1447 0.1536 0.1538 0.1607 Transport 0.0738 0.0712 0.0701 0.0723 0.0675 0.0705 Finance and real estate 0.0566 0.0578 0.0603 0.0636 0.0718 0.0693 Public services and defense 0.3045 0.2991 0.2937 0.2870 0.2919 0.2974 Other services 0.0004 0.0006 0.0003 0.0008 0.0011 0.0013 Hold temporary contracts 0.0360 0.0342 0.0390 0.0846 0.0820 0.1329 Private Sector 0.4332 0.4524 0.4651 0.4959 0.5062 0.5568 Firm size under 10 0.0991 0.1452 0.1404 0.1721 0.1695 0.1811 247 Firm size 10 to 20 0.1794 0.1350 0.1488 0.1226 0.1215 0.1215 Firm size 21 to 50 0.1685 0.1649 0.1705 0.1694 0.1758 0.1740 Firm size 51 to 100 0.1248 0.1227 0.1193 0.1416 0.1409 0.1478 Firm size over 100 0.3728 0.3725 0.3559 0.3487 0.3377 0.3228 Professional 0.1230 0.1275 0.1220 0.1239 0.1229 0.1188 Managerial 0.0418 0.0408 0.0427 0.0392 0.0396 0.0379 Technical 0.1658 0.1568 0.1586 0.1522 0.1511 0.1587 Clerical 0.1007 0.0959 0.1026 0.0977 0.1021 0.1023 Sales 0.0996 0.1082 0.1081 0.1174 0.1186 0.1265 Farm workers 0.0072 0.0070 0.0072 0.0086 0.0079 0.0088 Skilled manual 0.2371 0.2331 0.2373 0.2267 0.2266 0.2159 Semi-skilled manual 0.1242 0.1243 0.1214 0.1186 0.1160 0.1168 Unskilled 0.1006 0.1064 0.1000 0.1156 0.1151 0.1143 Has a recent spell unemployed 0.2303 0.2162 0.2379 0.0500 0.0491 0.0504 Full-time student 0.0086 0.0072 0.0087 0.0568 0.0633 0.0637 Under 5 years experience 0.1434 0.1219 0.1363 0.1183 0.1366 0.1252 5 to 10 years experience 0.1256 0.1300 0.1224 0.1292 0.1247 0.1289 10 to 20 years experience 0.2697 0.2654 0.2725 0.2512 0.2578 0.2495 Over 20 years experience 0.4614 0.4827 0.4688 0.5013 0.4809 0.4964 1 to 5 years tenure 0.2634 0.2825 0.2745 0.2953 0.2952 0.3003 5 to 10 years tenure 0.1703 0.1678 0.1835 0.1799 0.1940 0.1946 Over 10 years tenure 0.4014 0.3965 0.3995 0.3825 0.3844 0.3746 Apprentice 0.0311 0.0378 0.0194 Note: For 1998, the smallest firm size category is 1-5, and the next is 6-20. For 2001 and 2002 the firm size categories change as follows: under 10 becomes under 11; 10 to 20 becomes 11 to 19; 21 to 50 becomes 20 to 49; 51 to 100 becomes 50 to 100; over 100 is unchanged. Source: PLFS. 11.21 The private sector has grown in importance. There is a related decline in the proportion of workers employed in large firms. Among occupations, the share of skilled manual workers has declined, while the share of sales workers has increased; otherwise the changes are relatively small. Average potential experience increases as the post-War baby boomers become older. Here and throughout, potential experience is measured as equal to age ­ 7 ­ years in education. On the other hand, job tenure declines, and more workers report relatively short job durations. D. WAGE EQUATION ESTIMATION The Evolution of Pay Determination over the Transition Period 11.22 Tables 11.4 and 11.5 report wage equations for the private sector and the public sector respectively. We employ 12 rounds of the PLFS, covering the nine years 1994-2002 inclusive in order to search for effects that are shifting over time. We estimate by Ordinary Least Squares. In this section, under Quantile Regression Analysis, we investigate whether quantile regression offers more subtle insights. Also in this section, under Controlling for Participation Bias, we attempt to control for the various selection biases. 248 Table 11.4: OLS Modeling of (log) Monthly Earnings in the Private Sector, 1994-2002 1994r4 1995r4 1996r4 1997r4 1998r2 1998r4 Women -0.155** -0.160** -0.139** -0.138** -0.162** -0.145** Household head 0.094** 0.078** 0.076** 0.077** 0.074** 0.086** Married 0.017 0.023 0.033** 0.037** 0.033** 0.033** Ed-university 0.406** 0.493** 0.457** 0.439** 0.521** 0.449** Ed-post secondary 0.184** 0.176** 0.194** 0.172** 0.213** 0.238** Ed-general secondary 0.108** 0.140** 0.114** 0.108** 0.140** 0.152** Ed-vocational secondary 0.138** 0.158** 0.134** 0.141** 0.177** 0.223** Ed-lower vocational 0.070** 0.072** 0.060** 0.057** 0.066** 0.094** Ind-agriculture -0.044 -0.111** -0.068 -0.098** -0.109** -0.125** Ind-mining -0.042 0.068 0.097 0.019 0.070 0.133* Ind-manufacturing -0.012 0.012 -0.017 -0.029 0.004 -0.018 Ind ­utilities 0.069 -0.041 0.077 0.092* -0.003 0.021 Ind ­construction 0.029 0.056* 0.043 0.084** 0.066** 0.073** Ind-trade and repair -0.018 -0.009 0.010 0.003 0.042* 0.000 Ind-transport 0.062 0.093** 0.049 0.097** 0.126** 0.067* Ind-finance and real estate 0.048 0.054 0.017 0.058* 0.061* 0.062* Hold temporary contracts -0.087** -0.074** -0.093** -0.087** -0.057** -0.106** Firm size 1 to 5 -0.138** -0.103** -0.081** -0.113** -0.104** -0.104** Firm size 6 to 20 -0.079** -0.036* -0.024 -0.028* -0.036** -0.060** Firm size 51 to 100 -0.014** 0.048* 0.067** 0.059** 0.060** 0.028 Firm size over 100 0.077** 0.109** 0.114** 0.117** 0.137** 0.071** Occ-professional 0.236** 0.155** 0.241** 0.236** 0.239** 0.310** Occ-managerial 0.380** 0.339** 0.378** 0.377** 0.391** 0.413** Occ-technical 0.106** 0.156** 0.139** 0.149** 0.165** 0.176** Occ-clerical 0.055* 0.043* 0.059** 0.052** 0.066** 0.084** Occ-sales -0.046 -0.029 -0.066** -0.049** -0.058** -0.055** Occ-farm workers -0.085 0.083 0.036 -0.003 0.023 0.022 Occ-semi-skilled 0.033 0.020 0.037* 0.035* 0.037** 0.042** Occ-unskilled -0.092** -0.068** -0.067** -0.105** -0.061** -0.072** Full-time student -0.054 0.301** 0.029 -0.239** -0.265** -0.235** 5 to 10 years experience 0.029 0.066** 0.058** 0.047** 0.050** 0.081** 10 to 20 year experience 0.069** 0.077** 0.045** 0.023 0.048** 0.061** Over 20 year experience 0.070** 0.079** 0.061** 0.018 0.057** 0.051** 1 to 5 years tenure 0.052** 0.049** 0.034** 0.035** 0.039** 0.039** 5 to 10 years tenure 0.004 0.042* 0.073** 0.080** 0.076** 0.085** Over 10 years tenure 0.012 0.046** 0.058** 0.069** 0.067** 0.105** Log hours 0.339** 0.256** 0.345** 0.327** 0.364** 0.256** Urban voivodship 0.003** 0.003** 0.002** 0.003** 0.002** 0.003** Voivodship unemployment rate -0.798** -0.338** -0.944** -0.818** -1.021** -1.039** Adj. R-sq 0.372 0.427 0.408 0.422 0.475 0.447 See 0.353 0.314 0.319 0.391 0.305 0.312 N 3932 4715 5380 6016 6077 6138 Notes: 1. * and ** indicate significance at the 1 percent and 5 percent levels, respectively. 2. The default educational group comprises those with no more than a completed primary education. The default industry is public and other service. The default firm size is 21-50. The default occupational group is skilled manual. The default level of potential experience is less than 5 years and the default job tenure is less than 1 year. Source: Authors' calculations from the Polish Labor Force Survey. 249 Table 11.4 (continued): OLS Modeling of (log) Monthly Earnings in the Private Sector, 1994-2002 1999r4 2000r2 2000r4 2001r2 2001r4 2002r2 Women -0.160** -0.150** -0.156** -0.102** -0.131** -0.125** Household head 0.097** 0.097** 0.102** 0.104** 0.108** 0.105** Married 0.036** 0.026* 0.056** 0.049** 0.064** 0.053** Ed-university 0.432** 0.450** 0.447** 0.513** 0.518** 0.527** Ed-post secondary 0.177** 0.169** 0.111** 0.273** 0.326** 0.283** Ed-general secondary 0.171** 0.154** 0.180** 0.258** 0.276** 0.262** Ed-vocational secondary 0.220** 0.183** 0.197** 0.321** 0.271** 0.324** Ed-lower vocational 0.098** 0.092** 0.096** 0.159** 0.205** 0.172** Ind-agriculture -0.065 -0.112** -0.112* 0.023 0.035 -0.004 Ind-mining 0.099 0.072 0.082 0.162 0.196** 0.196** Ind-manufacturing 0.031 -0.058* -0.026 0.049 0.023 0.031 Ind ­utilities 0.174** 0.077 0.101 0.104 0.140* 0.111 Ind ­construction 0.143** 0.034 0.071* 0.134** 0.101** 0.133** Ind-trade and repair 0.063* -0.001 0.021 0.070* 0.067* 0.032 Ind-transport 0.136** 0.060 0.135** 0.155** 0.161** 0.151** Ind-finance and real estate 0.101** -0.001 0.090** 0.136** 0.109** 0.087** Holds a temporary contract -0.106** -0.094** -0.091** -0.103** -0.098** -0.086** Firm size 1 to 10 -0.093** -0.081** -0.102** -0.091** -0.123** -0.078** Firm size 10 to 20 -0.034* -0.024 -0.053** -0.016 -0.038* -0.006 Firm size 51 to 100 0.028 0.040* 0.016 0.014 0.059** 0.047* Firm size over 100 0.100** 0.079** 0.068** 0.128** 0.112** 0.130** Occ-professional 0.336** 0.382** 0.294** 0.442** 0.365** 0.376** Occ-managerial 0.510** 0.478** 0.393** 0.505** 0.441** 0.509** Occ-technical 0.198** 0.246** 0.201** 0.200** 0.245** 0.215** Occ-clerical 0.113** 0.072** 0.096** 0.091** 0.084** 0.126** Occ-sales -0.026 -0.072** -0.077** -0.085** -0.068** -0.052* Occ-farm workers 0.021 0.030 0.107 -0.025 -0.146* -0.042 Occ-semi-skilled 0.060** 0.020 0.047** 0.040* 0.051** 0.059** Occ-unskilled -0.043* -0.081** -0.086** -0.063** -0.074** -0.075** Full-time student -0.287** -0.287** -0.300** -0.167** -0.128** -0.168** 5 to 10 years experience 0.081** 0.052** 0.038* 0.099** 0.070** 0.098** 10 to 20 year experience 0.070** 0.055** 0.056** 0.110** 0.055** 0.092** Over 20 year experience 0.063** 0.049** 0.032 0.072** 0.020 0.052* 1 to 5 years tenure 0.039** 0.078** 0.063** 0.027 0.049** 0.040* 5 to 10 years tenure 0.103** 0.102** 0.088** 0.074** 0.097** 0.099** Over 10 years tenure 0.097** 0.105** 0.101** 0.096** 0.121** 0.119** Log hours 0.274** 0.212** 0.205** 0.523** 0.489** 0.518** Urban voivodship 0.003** 0.004** 0.004** 0.006** 0.003** 0.005** Voivodship unemployment rate -0.082 -0.789** -1.023** -1.668** -0.710** -1.050** Adj. R-sq 0.404 0.446 0.412 0.478 0.490 0.485 See 0.338 0.320 0.341 0.383 0.370 0.375 N 4623 4639 4834 5281 5147 5540 Notes: 1. * and ** indicate significance at the 1 percent and 5 percent levels, respectively. 2. The default educational group comprises those with no more than a completed primary education. The default industry is public and other service. The default firm size is 21-50. The default occupational group is skilled manual. The default level of potential experience is less than 5 years and the default job tenure is less than 1 year. 3. For 2001 and 2002, the firm size categories change as follows: under 10 becomes under 11; 10 to 20 becomes 11 to 19; 21 to 50 becomes 20 to 49; 51 to 100 becomes 50 to 100; over 100 is unchanged. Source: Authors' calculations from the PLFS. 250 Table 11.5: OLS Modeling of (log) Monthly Earnings in the State Sector, 1994-2002 1994r4 1995r4 1996r4 1997r4 1998r2 1998r4 Women -0.166**-0.150**-0.185**-0.181**-0.173**-0.173** Household head 0.070** 0.093** 0.085** 0.075** 0.082** 0.071** Married 0.039** 0.059** 0.058** 0.043** 0.041** 0.031** Ed-university 0.418** 0.431** 0.446** 0.440** 0.425** 0.401** Ed-post secondary 0.176** 0.201** 0.189** 0.178** 0.180** 0.146** Ed-general secondary 0.137** 0.157** 0.168** 0.171** 0.155** 0.147** Ed-vocational secondary 0.146** 0.187** 0.194** 0.169** 0.168** 0.144** Ed-lower vocational 0.054** 0.071** 0.054** 0.049** 0.054** 0.043** Ind-agriculture -0.087**-0.089**-0.053** -0.036* 0.002 -0.009 Ind-mining 0.377** 0.341** 0.333** 0.308** 0.302** 0.319** Ind-manufacturing 0.021* 0.020* 0.062** 0.061** 0.057** 0.058** Ind ­utilities 0.256** 0.164 0.169** 0.178** 0.198** 0.207** Ind ­construction 0.001 -0.008 0.058** 0.046** 0.092** 0.084** Ind-trade and repair -0.039** -0.032* -0.012 -0.017 0.017 -0.018 Ind-transport 0.045** 0.010 0.041** 0.048** 0.064** 0.086** Ind-finance and real estate 0.089** 0.041** 0.076** 0.099** 0.125** 0.095** Holds a temporary contract -0.126**-0.088**-0.101**-0.120**-0.161**-0.118** Firm size 1 to 5 -0.062**-0.055** -0.021 -0.034 -0.044* -0.026 Firm size 6 to 20 -0.018 -0.019 -0.012 -0.037** -0.027* -0.023* Firm size 51 to 100 0.009 0.015 0.009 -0.006 0.024* 0.006 Firm size over 100 0.075** 0.065** 0.069** 0.060** 0.066** 0.051** Occ-professional 0.094** 0.092** 0.085** 0.093** 0.128** 0.142** Occ-managerial 0.270** 0.275** 0.273** 0.286** 0.355** 0.330** Occ-technical 0.122** 0.094** 0.084** 0.101** 0.118** 0.120** Occ-clerical 0.012 0.009 0.033* 0.039** 0.051** 0.052** Occ-sales -0.027* -0.053** -0.015 -0.028 0.000 -0.013 Occ-farm workers 0.034 -0.035 -0.043 -0.047 0.043 0.017 Occ-semi-skilled 0.031** 0.011 0.024* 0.039** 0.020 0.041** Occ-unskilled -0.142**-0.146**-0.124**-0.116**-0.116**-0.111** Full-time student 0.153 -0.053 0.024 -0.040 -0.024 0.018 5 to 10 years experience 0.060** 0.057** 0.059** 0.074** 0.060** 0.099** 10 to 20 year experience 0.111** 0.101** 0.107** 0.119** 0.117** 0.123** Over 20 year experience 0.142** 0.128** 0.134** 0.153** 0.146** 0.165** 1 to 5 years tenure 0.055** 0.076** 0.050** 0.057** 0.052** 0.050** 5 to 10 years tenure 0.044** 0.073** 0.083** 0.079** 0.074** 0.069** Over 10 years tenure 0.067** 0.093** 0.095** 0.086** 0.090** 0.077** Log hours 0.214** 0.215** 0.184** 0.180** 0.226** 0.232** Urban voivodship 0.002** 0.002** 0.002** 0.002** 0.002** 0.002** Voivodship unemployment rate 0.046 -0.085 -0.263**-0.400** -0.184* -0.414** Adj. R-sq 0.459 0.460 0.459 0.466 0.463 0.462 see 0.293 0.289 0.290 0.284 0.282 0.277 N 11683 11107 10434 9782 9201 8889 Notes: 1. * and ** indicate significance at the 1 percent and 5 percent levels, respectively. 2. The default educational group comprises those with no more than a completed primary education. The default industry is public and other service. The default firm size is 21-50. The default occupational group is skilled manual. The default level of potential experience is less than 5 years and the default job tenure is less than 1 year. Source: Authors' calculations from the PLFS. 251 Table 11.5 (continued): OLS Modeling of (log) Monthly Earnings in the State Sector, 1994-2002 1999r4 2000r2 2000r4 2001r2 2001r4 2002r2 Women -0.173** -0.180** -0.184** -0.144** -0.145** -0.131** Household head 0.083** 0.087** 0.084** 0.094** 0.113** 0.103** Married 0.049** 0.036** 0.045** 0.047** 0.049** 0.055** Ed-university 0.435** 0.414** 0.436** 0.470** 0.424** 0.444** Ed-post secondary 0.186** 0.149** 0.138** 0.213** 0.179** 0.194** Ed-general secondary 0.154** 0.144** 0.147** 0.185** 0.183** 0.185** Ed-vocational secondary 0.195** 0.159** 0.203** 0.162** 0.244** 0.187** Ed-lower vocational 0.046** 0.063** 0.055** 0.078** 0.096** 0.076** Ind-agriculture 0.046 -0.030 -0.018 0.028 -0.019 0.076* Ind-mining 0.275** 0.307** 0.287** 0.291** 0.279** 0.327** Ind-manufacturing 0.062** 0.060** 0.058** 0.035* 0.025 0.031 Ind ­utilities 0.214** 0.190** 0.186** 0.163** 0.116** 0.154** Ind ­construction 0.138** 0.118** 0.147** 0.105** 0.134** 0.036 Ind-trade and repair 0.000 0.012 -0.013 0.001 -0.039 0.036 Ind-transport 0.088** 0.139** 0.082** 0.104** 0.093** 0.075** Ind-finance and real estate 0.126** 0.123** 0.157** 0.105** 0.135** 0.070** Holds a temporary contract -0.200** -0.138** -0.137** -0.145** -0.199** -0.177** Firm size 1 to 10 -0.083** -0.068** -0.041* -0.085** -0.110** -0.054** Firm size 10 to 20 -0.024 -0.016 -0.018 -0.027 -0.045** -0.015 Firm size 51 to 100 0.038** 0.022 0.029* 0.008 0.010 0.015 Firm size over 100 0.069** 0.076** 0.061** 0.052** 0.043** 0.044** Occ-professional 0.134** 0.161** 0.157** 0.162** 0.267** 0.219** Occ-managerial 0.382** 0.381** 0.382** 0.420** 0.401** 0.423** Occ-technical 0.139** 0.177** 0.171** 0.130** 0.176** 0.158** Occ-clerical 0.056** 0.102** 0.081** 0.063** 0.057** 0.075** Occ-sales 0.024 -0.006 0.025 -0.045 -0.010 -0.026 Occ-farm workers 0.036 -0.049 -0.049 -0.066 -0.005 -0.094 Occ-semi-skilled 0.053** 0.022 0.052** 0.036 0.026 0.047* Occ-unskilled -0.109** -0.108** -0.096** -0.142** -0.126** -0.145** Full-time student -0.017 -0.139* -0.118* -0.021 -0.057** 0.005 5 to 10 years experience 0.085** 0.115** 0.099** 0.140** 0.129** 0.156** 10 to 20 year experience 0.141** 0.172** 0.129** 0.145** 0.118** 0.190** Over 20 year experience 0.183** 0.229** 0.166** 0.169** 0.127** 0.194** 1 to 5 years tenure 0.047** 0.077** 0.064** 0.053* 0.028 0.044 5 to 10 years tenure 0.057** 0.098** 0.106** 0.107** 0.039 0.070** Over 10 years tenure 0.050** 0.070** 0.089** 0.104** 0.080** 0.077** Log hours 0.157** 0.172** 0.163** 0.297** 0.376** 0.409** Urban voivodship 0.003** 0.003** 0.004** 0.004** 0.002** 0.002** Voivodship unemployment rate -0.128 -0.304* -0.533** -0.500** -0.202 0.020 Adj. R-sq 0.439 0.454 0.449 0.463 0.479 0.486 See 0.290 0.290 0.291 0.323 0.313 0.319 N 6084 5619 5538 5372 5033 4371 Notes: 1. * and ** indicate significance at the 1 percent and 5 percent levels, respectively. 2. The default educational group comprises those with no more than a completed primary education. The default industry is public and other service. The default firm size is 21-50. The default occupational group is skilled manual. The default level of potential experience is less than 5 years and the default job tenure is less than 1 year. 3. For 2001 and 2002, the firm size categories change as follows: under 10 becomes under 11; 10 to 20 becomes 11 to 19; 21 to 50 becomes 20 to 49; 51 to 100 becomes 50 to 100; over 100 is unchanged. Source: Authors' calculations from the PLFS. 252 11.23 Starting at the top of the list of variables, we see that the ceteris paribus gender wage gap is slightly smaller in the private sector and that it narrowed sharply in both sectors after 2000. The wage return to heads of households shows a small but steady increase over time. Being the head of household often gives a positive wage return in studies such as this one. Most researchers rationalize the result by arguing that taking domestic responsibility reflects, or is thought by employers to reflect, an ability that enhances workplace productivity. This must be right, and so any increase in the wage return might reflect the fall in the supply of household heads noted in section C. 11.24 The next three results suggest that the private sector exhibits greater and faster rising ceteris paribus wage dispersion than the public sector. First, there are large increases in the returns to all forms of post-primary education in the private sector. In the public sector there are some increases, but they are much smaller. 11.25 Second, the positive relationship between firm size and wages is consistently more pronounced in the private sector. Additionally, there appear to be increasing wage premiums to skilled white-collar occupations, especially to professional, managerial, and technical workers. These premiums are larger in the private sector than in the public sector throughout the period. 11.26 One factor that reverses this tendency for inequality to be higher in the private sector is that wage returns to potential experience are higher in the public sector, which probably reflects a greater prevalence of seniority pay scales.14 However, the returns to longer tenures in the private sector grow over the period. It is tempting to hypothesize that those who started work after 1990 are beginning to be rewarded for loyalty. 11.27 Third, the impact of local unemployment on wagesthe Blanchflower-Oswald wage curve effectappears to have a reliable impact only on private sector wages. This result may reflect the greater importance of nationally negotiated pay rates in the public sector. Our two regional variables are the proportion urbanized, which varies from roughly 40 to 70 percent in 2002, and the voivodship unemployment rate, which has a range of 14 percentage points in 2002. Taking the results from the final column of Table 11.4, this means that the urbanization effect on wages is at most 20 percent, while the difference in ceteris paribus wages between the lowest and the highest unemployment regions is approximately15 percent. Disaggregating by Gender 11.28 In Tables A.11.1 to A.11.3 in Annex 1, we report some results from a re-estimation of this specification on data disaggregated by gender as well as by sector. Three sets of results are worthy of note. First, the wage impact of professional status (relative to skilled manuals) is lower for men than for women, and much lower for men in the public sector (see Table A.11.1). Second, in Table A.11.2, the effect of tenure on women's wages is low and declining in the public sector, whereas, as noted above, it rises for both genders in the private sector. Third, the return for potential experience is lower in the private sector, especially for women; the return to experience for women in the public sector is usually significantly higher than for other groups (see Table A.11.3). These results for tenure and experience suggest that women in the public sector are paid a little differently from other groups; for these women, tenure affects wages less than potential experience. Such an arrangement would suit workers with an interrupted pattern of participation over the lifespan. It is not obvious that the emerging private sector is following this pattern of pay. 14See below for more on this issue. 253 Quantile Regression Analysis 11.29 In Table 11.6 we investigate whether these estimated effects vary across the wage spectrum, by estimating quantile regressions15 for the spring 2002 data set. Estimates of the LAD estimator at the 10th, 25th, 50th, 75th, and 90th quantiles are given in the first five columns, with OLS estimates given in the sixth column, for reference. The main results are as follows. First, the female wage disadvantage widens as we move up through the wage distribution. This result is familiar in transition countries (see, for instance, Newell and Reilly, 2001). Second, the returns to the university, post-secondary, and vocational-secondary levels of education rise as we move up through the wage distribution. Third, the returns to working in the private sector swing from significantly negative at the bottom end of the wage distribution to significantly positive at the top end. Thus, as promised above, the private sector seems to generate greater ceteris paribus wage inequality. The returns to white collar occupations all increase across the wage spectrum, as do the returns to semi-skilled work. Last, the returns to long experience (over 20 years) are larger at the high end of the wage spectrum. Thus, in summary, many wage determinants have larger proportional impacts on wages in the upper parts of the wage distribution. Table 11.6: Quantile Wage Regressions for Spring 2002 q10 q25 q50 q75 q90 OLS Women -0.084** -0.112** -0.135** -0.155** -0.194** -0.127** Household head 0.075** 0.090** 0.102** 0.111** 0.114** 0.103** Married 0.032** 0.047** 0.045** 0.057** 0.057** 0.049** Ed-university 0.365** 0.402** 0.417** 0.441** 0.478** 0.470** Ed-post secondary 0.183** 0.183** 0.156** 0.183** 0.231** 0.221** Ed-general secondary 0.188** 0.169** 0.152** 0.172** 0.184** 0.211** Ed-vocational secondary 0.164** 0.183** 0.177** 0.237** 0.255** 0.254** Ed-lower vocational 0.148** 0.139** 0.141** 0.179** 0.195** 0.213** Ind-agriculture -0.119* -0.006 0.006 0.024 0.023 -0.006 Ind-mining 0.299** 0.293** 0.311** 0.303** 0.244** 0.292** Ind-manufacturing 0.003 0.014 0.027 0.036** 0.021 0.019 Ind ­utilities 0.146** 0.154** 0.148** 0.167** 0.146** 0.135** Ind ­construction 0.085** 0.099** 0.114** 0.101** 0.114** 0.097** Ind-trade and repair 0.031 0.005 0.018 0.033 0.049* 0.015 Ind-transport 0.088** 0.098** 0.103** 0.111** 0.115** 0.100** Ind-finance and real estate 0.025 0.037* 0.085** 0.086** 0.097** 0.073** Holds a temporary contract -0.125** -0.133** -0.118** -0.124** -0.130** -0.135** Apprentice -1.246** -0.830** -0.255** -0.229** -0.258** -0.510** Private Sector -0.047** -0.029** 0.003 0.032* 0.067** 0.010 Firm size 1 to 10 -0.080** -0.074** -0.084** -0.081** -0.087** -0.091** Firm size 10 to 19 -0.009 -0.006 -0.006 -0.018 -0.015 -0.010 Firm size 50 to 100 0.050** 0.031* 0.026** 0.019 0.015 0.026* Firm size over 100 0.042** 0.050** 0.063** 0.085** 0.106** 0.071** Occ-professional 0.261** 0.260** 0.311** 0.387** 0.393** 0.333** Occ-managerial 0.325** 0.372** 0.456** 0.536** 0.654** 0.474** Occ-technical 0.072** 0.108** 0.175** 0.249** 0.291** 0.184** 15See Koenker and Hallock (2001) for an excellent introduction to quantile regression. 254 Occ-clerical 0.066* 0.067** 0.077** 0.119** 0.143** 0.099** Occ-sales -0.050* -0.048** -0.045** -0.037 -0.034 -0.044** Occ-farm workers -0.027 -0.027 -0.027 -0.060 -0.051 -0.047 Occ-semi-skilled 0.033 0.018 0.020 0.056** 0.066** 0.044** Occ-unskilled -0.094** -0.106** -0.116** -0.104** -0.104** -0.091** Full-time student -0.032 -0.038* 0.007 0.014 0.016 -0.029 5 to 10 years experience 0.080** 0.074** 0.059** 0.073** 0.069 0.094** 10 to 20 year experience 0.071* 0.065** 0.075** 0.104** 0.113** 0.112** Over 20 year experience 0.059* 0.066** 0.083** 0.118** 0.141** 0.111** 1 to 5 years tenure 0.024 0.014 0.025 0.026 0.015 0.016 5 to 10 years tenure 0.064** 0.046* 0.064** 0.056** 0.040 0.047** Over 10 years tenure 0.106** 0.080** 0.082** 0.064** 0.041 0.063** Log normal hours 0.678** 0.625** 0.552** 0.535** 0.528** 0.632** Urban voivodship 0.002** 0.003** 0.003** 0.003** 0.003** 0.003** Voivodship unemployment rate -0.004** -0.003* -0.004** -0.007** -0.009** -0.006** Psuedo R-sq 0.363 0.307 0.304 0.316 0.331 Adj. R-sq 0.5345 See 0.33755 N 10319 Note: 1. For 2001 and 2002 the firm size groupings are: 1 to 10 employees, 11 to 19, 20 to 49, 50 to 100 and over 100. 2. * and ** indicate significance at the 1 percent and 5 percent levels, respectively. Source: Authors' calculations from the PLFS. Controlling for Participation Bias 11.30 In Table 11.7 we report wage equation estimates where we have attempted to control for biases. There are several possible sources of bias. First, participation in the labor market depends significantly upon personal, household, and regional characteristics; thus, the sample of participants is not an unbiased sample from the population. Second, only a fraction of participants are wage-earning employees, the others being the unemployed, the self-employed, and unpaid family workers. None of these groups reports wages and none is an unbiased selection from participants, nor from the population of working age. For example, unemployment is always more concentrated among lower skilled workers, and in the PLFS data the self-employed and unpaid family workers are predominantly own-account farmers and their families. Table 11.7: ML Heckman Wage Equation Estimates, 1994, 1995, 2001, 2002 November November Autumn 2001 1994 1995 Spring 2002 Women -0.163** -0.156** -0.144** -0.130** Household head 0.079** 0.088** 0.086** 0.079** Married 0.034** 0.047** 0.040** 0.033** Ed-university 0.417** 0.437** 0.409** 0.396** Ed-post secondary 0.178** 0.197** 0.191** 0.145** Ed-general secondary 0.133** 0.151** 0.180** 0.157** Ed-vocational secondary 0.145** 0.178** 0.219** 0.216** Ed-lower vocational 0.062** 0.071** 0.111** 0.075** Ind-agriculture -0.077** -0.089** -0.039 -0.009 255 Ind-mining 0.362** 0.329** 0.254** 0.287** Ind-manufacturing 0.016* 0.019* 0.007 0.018 Ind ­utilities 0.241** 0.148** 0.112** 0.133** Ind ­construction 0.029* 0.035** 0.074** 0.096** Ind-trade and repair -0.018 -0.014 0.021 0.014 Ind-transport 0.051** 0.027* 0.096** 0.096** Ind-finance and real estate 0.086** 0.053** 0.106** 0.072** Holds a temporary contract -0.111** -0.079** -0.156** -0.133** Apprentice -0.257** -0.504** Private Sector 0.121** 0.088** 0.048** 0.012 Firm size 1 to 5 -0.097** -0.088** -0.12** -0.081** Firm size 6 to 20 -0.035** -0.026** -0.037** 0.001 Firm size 51 to 100 0.001 0.021* 0.027* 0.036** Firm size over 100 0.070** 0.072** 0.069** 0.080** Occ-professional 0.128** 0.118** 0.351** 0.339** Occ-managerial 0.303** 0.309** 0.425** 0.489** Occ-technical 0.129** 0.118** 0.204** 0.190** Occ-clerical 0.028* 0.028** 0.081** 0.101** Occ-sales -0.032* -0.038** -0.037* -0.042** Occ-farm workers 0.009 -0.000 -0.031 -0.051 Occ-semi-skilled 0.035** 0.020* 0.041** 0.042** Occ-unskilled -0.126** -0.112** -0.093** -0.091** Full-time student 0.059 0.097* -0.056** -0.038* 5 to 10 years experience 0.048** 0.054** 0.070** 0.071** 10 to 20 year experience 0.092** 0.082** 0.055** 0.077** Over 20 year experience 0.113** 0.101** 0.061** 0.078** 1 to 5 years tenure 0.049** 0.059** -0.01 0.018 5 to 10 years tenure 0.031** 0.058** 0.016 0.051** Over 10 years tenure 0.056** 0.085** 0.055** 0.072** Log hours 0.251** 0.229** Log normal hours 0.591** 0.616** Urban voivodship 0.002** 0.002** 0.003** 0.004** Voivodship unemployment rate -0.196** -0.180** -0.489** -0.599** Table 11.7 (continued): Heckman Wage Selection Equations November November Autumn Spring 1994 1995 2001 2002 Women -0.111** -0.105** -0.012 0.012 Head of Household 0.302** 0.294** 0.361** 0.391** Married 0.075** 0.059** 0.157** 0.162** Ed-university 1.035** 1.059** 0.898** 0.811** Ed-post secondary 1.049** 1.008** 0.801** 0.852** Ed-general secondary 0.697** 0.677** 0.605** 0.585** Ed-vocational secondary 0.480** 0.436** 0.405** 0.377** Ed-lower vocational 0.512** 0.527** 0.464** 0.387** Aged 21 to 25 0.806** 0.805** 0.742** 0.824** Aged 26 to30 1.029** 1.042** 0.911** 1.035** Aged 31 to35 0.965** 1.008** 0.948** 0.987** 256 Aged 36 to 40 1.048** 1.068** 0.848** 0.978** Aged 41 to 45 1.054** 1.045** 0.849** 0.957** Aged 46 to 50 0.953** 0.938** 0.763** 0.867** Aged 51 to 55 0.588** 0.612** 0.544** 0.668** Aged 56 to 60 -0.107* -0.156** 0.102* 0.187** Aged over 61 -1.308** -1.178** 0.830** -0.790** Vague 0.644** 0.691** Urban voivodship 0.003** 0.003** 0.003** 0.004** Voivodship Unemployment -0.048 -0.185 0.344 -0.404 rate 1 1+ ^ 0.021 -0.009 -0.208** -0.281** 2 ln1- ^ ln( ) -1.170** -1.209** -1.082** -1.062** LR test ( = 0), (1) 2 0.23 0.04 23.16** 36.68** Wald chi2(41) 8529.03 8851.92 8168.56 8638.81 Log likelihood -27253.67 -27183 -23283.95 -23076.63 N 53043 53761 47094 48009 Uncensored obs 15615 15822 10634 10319 Notes: 1. For 2001 and 2002, the firm size groupings are: 1 to 10 employees, 11 to 19, 20 to 49, 50 to 100 and over 100. 2. * and ** indicate significance at the 1 percent and 5 percent levels, respectively. 3. Estimation is by Heckman's (1976) maximum likelihood method as programmed in Stata 7. Source: Authors' calculations from the PLFS. 11.31 There is a third potential bias stemming from the fact that not all employees reveal their earnings to the PLFS interviewers. In particular, white-collar workers, such as managers and clerical workers, are significantly less likely to report wages. 11.32 In principle, these three sources of bias, which we might call the participation, labor force status, and non-reporting biases, could be dealt with separately if suitable identifying variables were available. We have only one identifying variable (see below) and so we collapse the three steps into a single step and we model what makes someone report their wage, as distinct from what makes someone a non-reporter. We estimate these joint participation and wage equation systems for the first two years in our sample (1994 and 1995) and for the last two rounds of our sample (Autumn 2001 and Spring 2002) using Heckman's (1976) full maximum likelihood estimation procedure. We model participation as depending on individual characteristics, such as education, age, gender, and household and marital status, as well as on two potentially relevant regional characteristics--the degree of urbanization and the unemployment rate. 11.33 In the latter two rounds of the survey we have a variable that helps identify employees who do not report wages. This variable stems from a revision in the design of the question about firm size. In later rounds of the survey the respondent is allowed to record an uncertain response, such as "Don't know." In earlier rounds only definite answers were allowed. We create an indicator variable called vague that takes the value 1 when a respondent is uncertain about firm size, and zero otherwise. Our hypothesis is that if respondents are uncertain about the size of the firm, they also may not recall their wage, either out of genuine ignorance or because of an unwillingness to fully engage with the survey. 11.34 The specifications vary in two other important ways between the two mid-1990s samples and the more recent samples. First, in the later surveys the questionnaire made the distinction 257 between actual hours worked and normal hours. Experimentation revealed that monthly wages are much more strongly associated with normal hours than with actual hours. As a consequence, where we have normal hours, we use them. Second, the later questionnaires also introduced a question on apprenticeship. The response to this question is added to the specifications where it is available. 11.35 The results are given in Table 11.7. The estimated wage equation coefficients are in general very similar to those found in Tables 11.4 and 11.5. There are differences, however. Most notably, the broad and large upward trend in most of the education coefficients visible in Tables 11.4 and 11.5 is more-or-less absent here. The education coefficients estimated for 1994 and 1995 using the Heckman procedure are very close to those estimated by OLS. However, for the Autumn 2001 and Spring 2002 data sets, these education coefficients are uniformly lower than those obtained by OLS. The result is that there appears to be no change over time in these education effects. This is not true for any of the other coefficients in Table 11.7. Other than the education coefficients, the Heckman procedure generates a reassuringly similar picture to that of OLS. Experimentation with the hours and apprenticeship variables revealed that these specification changes played no role in changing the education coefficients. 11.36 What causes the difference between the Heckman and OLS estimated education coefficients in the 2001 and 2002 samples? The standard answer would involve returning to basic theory in labor supply. We know that better educated workers are more likely to participate. In the jargon, this means their market wages are more likely to be above their reservation wages. We also know, from the quantile regression analysis, that the wage return to education is higher at higher wages. Thus, it is likely that non-participants experience, on average, low wage returns to education. Thus, controlling for the exclusion should lead to higher estimated wage returns to education. However, we find that the opposite is the case: controlling for participation seems to lower the return to education. 11.37 The reason for this seems to be a large shift in the propensity to reveal wages, especially among graduates. In the November 1994 survey, 90 percent of employee respondents supplied an estimate of their monthly earnings. The proportion was a little lower among graduates, at 87 percent. In the spring 2002 survey, only 73 percent of employees supplied wage information, and only 63 percent of graduates responded. Thus, there has been a marked reduction in the amount of responses to wage questions. It is not clear why this has happened. It is not simply the result of privatization. Private sector workers are more reluctant to reveal wages, but this accounts only for a small fraction of the drop in wage responses. 11.38 If the probability of refusing to estimate wages is inversely related to unmeasured productivity characteristics, which seems likely given the results, then the bias in this shift in the propensity to report wages accounts for the lower returns to education in the Heckman-corrected wage equation estimates. 11.39 This correction does not apply in the 1994 and 1995 samples, for two reasons. First, there is much less of a bias in the reporting of wages in the earlier samples. Second, since the early 1990s, there has been a large fall in participation in paid employment by workers with primary or lower education, so that the potential for bias is much greater now than it was formerly. For example, in 1994 the employee-to-population ratio of people with primary or lower education was about 45 percent of all workers. By 2002 this had dropped to about 37 percent; thus, these people are a declining fraction of the labor force. This suggests that these changes in the pattern of participation might be behind these results. The message is that the rises in estimated returns to 258 education, visible in the tables 11.4 and 11.5, may not be robust to biases due to changes in the pattern of participation. What Generates Low Pay? 11.40 In Table 11.8 we report an attempt to account for the gap in wages between workers in the lowest decile of the wage distribution and workers at the mean of the wage distribution, using data from 1994 and 2002. In 2002 the gap between the mean log wage and the mean log wage in the lowest decile is 0.6126, so that workers at the mean wage earn on average about 185 percent of the mean wage of those in the lowest decile. Just over half of both gaps is in the residual (i.e., it is unexplained). Some researchers designate this as unobserved productivity, and surely this is partly true, but other explanations are possible. Just under a quarter of each gap (13 out of 61 log points in 2002, 12 out of 61 points in 1994) is due to education and occupation. The final quarter of the gap is distributed among personal characteristics (gender, marital status, and status in the household), tenure and experience, and various job characteristics. The lion's share of the explained component of the wage gap is therefore due to educational and occupational choices. The results for the two samples are quite similar, except for the much larger concentration of temporary workers among the low paid in the 2002 sample. It should be noted from Table 11.3 and from the discussion in section B, Macroeconomic Overview of the Labor Market, that temporary workers increased from 3 percent of the sample of wage earners in 1994 to 13 percent in 2002. Table 11.8: Accounting for the Gap between Wages in the Lowest Decile and the Mean Wage, Full-time Workers, Autumn 1994 and Spring 2002 Autumn Spring 1994 2002 Gap between the mean log wage and the mean log wage 0.6054 0.6126 in the lowest decile Of which: unexplained 0.3541 0.3250 explained, of which: 0.2513 0.2876 due to education/occupation 0.1213 0.1315 due to personal and household factors 0.0541 0.0524 due to firm size 0.0193 0.0255 due to the share of temporary contract workers 0.0074 0.0256 due to tenure/experience 0.0152 0.0208 due to industrial differences 0.0261 0.0124 due to minor systematic influences 0.0080 0.0194 Notes: These statistics derive from an estimated wage model specified almost exactly as in Table 11.6. The unexplained component is the (negative) of the average residual from this regression among lowest decile workers. The explained components are calculated by multiplying mean characteristic differences by the OLS coefficients. For brevity, the regression results are not reported, but they are available from the authors on request. Source: Authors' calculations from the PLFS. E. CONCLUSIONS 11.41 First, until 2001, the distribution of wages in Poland had not become decisively more unequal through the transition. Some forces have been pushing toward greater inequality, while others have been pulling in the opposite direction. Among the former, there has been a considerable widening of occupational differentials. For example, the average wage mark-up of 259 professional and managerial workers over manual workers has increased since the mid-1990s by about 20 percentage points. Newell and Socha (2002) study this rise across manufacturing industries and find it strongly associated with TFP growth. They also find that both TFP and the white-collar premium are strongly associated with privatization and, to a lesser extent, with the levels of import penetration and R&D expenditures. But even controlling for its role in widening occupational differentials, privatization is tending to increase wage inequality. Our quantile regression analysis demonstrates how private firms tend to pay less at the bottom end of the wage distribution and more at the top end. 11.42 Participation changes pulled in the opposite direction, toward greater wage equality, at least until the end of the 1990s (Newell, 2001). Newell (2001) (using a Gini decomposition given in Annex 2) shows that all of the increase in household labor income inequality in 1994-98 was due to falling participation. Over the period from 1992 to 1998, participation fell, largely because young people were staying on longer in education. Together with the retirement of an older, less well-educated, generation, this has resulted in a contraction in the share of people with no more than primary education in the working age population, from 33.7 percent in 1992 to 27.6 percent in 2002. There has been a similar fall in the share of these people among employees, and this reduction in the share of potentially low wage workers has probably been a contractionary force on wage inequality. 11.43 The end of this fall in participation around the turn of the century suggests a neat theory of why wage inequality increased somewhat from 2001 onward. We have argued that this fall in participation has tended to reduce inequality in wages, while privatization works in the opposite direction. It is tempting to argue that as privatization continued and the fall in participation ended, increased wage inequality emerged. 11.44 Rising wage inequality is a difficult phenomenon for policymakers. There is little doubt that high income inequality leads to lower levels of social cohesion, and that it may cause political unrest and may retard the rate of economic growth. It is also true that some policies aimed at reducing inequality, such as redistributive taxes and social security systems, can have similar adverse effects. The worst situation occurs when redistributive policies have little effect on inequality and merely reinforce the diminution of social cohesion. 11.45 If a population chooses to increase its level of education, it is possible that the level of cross-section income inequality could riseif, for instance, students earned little in education, but earned higher wages than otherwise once they were in the labor market. 11.46 It seems to us that this policy is misdirected if it is deliberately aimed at re-distributing wages. Evidence from around the world suggests that equality of educational opportunities and an effective social safety net for the poor are the best ways of coping with the adverse side effects of wage inequality. It is most important that the largest number of people take up educational and training opportunities, in order to maximize the supply of voluntarily well-educated workers. 260 REFERENCES Carey, David, and Tchilinguirian, Harry (2000), "Average Effective Tax Rates on Capital, Labor and Consumption," OECD Economics Department Working Papers No. 258. GUS (2002), "Labor Force Survey in Poland in the Years 1992-2001," Central Statistical Office, Warsaw. GUS (2003), "Labor Force Survey in Poland: IIV Quarter 2002," Central Statistical Office, Warsaw. Heckman, J. (1976), "The Common Structure of Statistical Models of Truncation, Sample Selection, and Limited Dependent Variables and a Simple Estimator for Such Models," Annals of Economic and Social Measurement, 5, 1976, 475-492. Keane, M. P., and E. Prasad (2002), "Changes in the Structure of Earnings During the Polish Transition," IZA Discussion Paper No 496, May. Koenker, R., and K. F. Hallock (2001), "Quantile Regressions," Journal of Economic Perspectives, 15: 4, 143-156. Kwiatkowski, E., M. W. Socha, and U. Sztanderska (2001), "Labor Market Flexibility and Employment Security. Poland," ILO Employment Paper 2001/28, ILO, Geneva. MLSP (2000), "Basic Statistical Data on Social Policy," Ministry of Labor and Social Policy, Warsaw, November. Newell, A. (2001a), "The Distribution of Wages in the Transition Countries," IZA Bonn Discussion Paper No 267, March 2001 and University of Sussex, Discussion Papers in Economics No 70, February 2001. Newell, A. (2001b), "Why Have a Million More Polish Workers Become Unemployed in the Midst of an Economic Boom?" mimeo, University of Sussex, July 2001. Paper prepared for the ILO/Ministry of Labour conference on Poland's unemployment, July 2001. Newell, A., and B. Reilly (2001), "The Gender Wage Gap in the Transition from Communism: Some Empirical Evidence," Economic Systems, 25, 287-304. Newell, A., and M. Socha (1998), "The Roles of Privatization and Changes in International Trade in Changes in the Distribution of Wages in Poland, 1992-1996," Economics of Transition, 6(1). Newell, A., and M. Socha (2002), "The Rising Non-manual Wage Premium in Poland," University of Sussex, Discussion Papers in Economics, No 89, October, 2002. Newell, A., and M. Socha (2003), `The Evolution of Regional Unemployment in Poland," mimeo, University of Sussex, June 2003. Paper prepared for the World Bank Living Standards Assessment of Poland, 2003. OECD (2001), "Taxing Wages," OECD, Paris. 261 Riboud, M., C. Sainchez-Piramo, and C. Silva-Jauregui (2002), "Does Eurosclerosis Matter? Institutional Reform and Labor Market Performance in Central and Eastern European Countries in the 1990s," World Bank SP Discussion Paper No. 0202, March 2002. Rutkowski, J. (1996), "High Skills Pay Off: The Changing Wage Structure During the Transition in Poland," Economics of Transition, 4, pp.89­111. Socha, M. W., and J. Weisberg (2002), "Labor Market Transition in Poland: Changes in the Public and Private Sectors," International Journal of Manpower, Vol. 23 No 6. 262 Annex 1: Some Results for Data Decomposed by Sector and Gender In the following tables we report coefficients estimated by OLS from separate wage equations for men and women in the private and public sectors. We use specifications otherwise identical to those in Tables 11.4 and 11.5. Only a few of the results showed systematic differences by gender. These are presented below and discussed in Section D in the text. Table A.11.1: Estimated Wage Impact of Professional Status (relative to skilled manuals), 1994-2002 Men Women Public sector Private Sector Public Sector Private Sector 1994 0.043 0.228** 0.158** 0.265** 1995 0.014 0.167** 0.138** 0.175** 1996 0.012 0.185** 0.139** 0.342** 1997 0.028 0.194** 0.180** 0.318** 1998 0.099** 0.290** 0.173** 0.367** 1999 0.076* 0.296** 0.242** 0.415** 2000 0.077* 0.243** 0.218** 0.384** 2001 0.175** 0.296** 0.366** 0.496** 2002 0.141** 0.368** 0.228** 0.451** Source: Authors' calculations from the PLFS. Table A.11.2: Estimated Wage Impact of Tenure over 10 Years (relative to tenure of 1 year or less), 1994-2002 Men Women Public sector Private Sector Public Sector Private Sector 1994 0.063** -0.008 0.062** 0.038 1995 0.077** 0.047* 0.070** 0.051* 1996 0.096** 0.060** 0.075** 0.053* 1997 0.110** 0.075** 0.036* 0.060** 1998 0.096** 0.100** 0.046* 0.108** 1999 0.072** 0.092** 0.014 0.102** 2000 0.124** 0.096** 0.047* 0.100** 2001 0.095** 0.126** 0.048 0.114** 2002 0.086* 0.103** 0.051 0.128** Source: Authors' calculations from the PLFS. Table A.11.3: Estimated Wage Impact of Experience over 20 Years (relative to less than 5 years experience), 1994-2002 Men Women Public sector Private Sector Public Sector Private Sector 1994 0.100** 0.072** 0.190** 0.064* 1995 0.083** 0.076** 0.186** 0.085** 1996 0.053** 0.040 0.226** 0.089** 1997 0.123** 0.031 0.200** 0.039 1998 0.122** 0.061** 0.217** 0.036 1999 0.106** 0.081** 0.262** 0.028 2000 0.080** 0.064** 0.246** -0.013 2001 0.133** 0.035 0.148** -0.000 2002 0.201** 0.052 0.212** 0.061* Notes: Cells contain the estimated impact on log monthly wages of dummy variables indicating the relevant characteristic. As elsewhere, * and ** indicate significance at the 1 percent and 5 percent levels, respectively. Source: Authors' calculations from the PLFS. 263 Annex 2: Gini Coefficients In the text we assert that where there are households with zero earnings, changes in the Gini coefficient can be decomposed into two components. The first component is proportional to changes in Gini inequality among earners and the second component is proportional to changes in the preponderance of non-earners. Define the overall Gini coefficient as Go, where 1 1 Go = - l(s)ds. 2 x Here, x is the proportion of zero-income households, s takes values between x and unity, and l(s) is the Lorenz function. The Gini coefficient of inequality among income-receiving households, Gw , is 1 1 Gw = - l((1- x)y + x)dy, where y = s 2 + 1 . 1- x 1- x 0 1 1 Now, l(s)ds = l((1- x)y + x)(1- x)dy , so x 0 G0 = + x 1- x Gw. Totally differentiating we have, 2 2 dGo = 1- Gw dx + 1 - x dGw, as required. 2 2 wb13696 P:\POLAND\PREM\Living Standard Assessment\4RED\Part 3\Papers\FINALPLS Vol 2 Chap8-11 0301.doc March 24, 2004 11:03 AM 264 PART IV: VULNERABILITY, SAFETY NETS AND INCENTIVES 12. INFORMAL NETWORKS OF SUPPORT OF POOR PEOPLE IN POLAND Wielislawa Warzywoda-Kruszyska and Jolanta Grotowska-Leder A. INTRODUCTION 12.1 For the day-to-day subsistence of a human being, that person's individual resources are important: his or her professional qualifications, family life, and health, as well as his or her social capital, understood as social relationships that a person can lean on and resort to when carrying out both current and future life goals (Giza-Poleszczuk, Marody, Rychard [1999]). The social capital remaining at the disposal of an individual is usually a derivative of his or her place in the social structure. Research has shown that the higher the social position of an individual, the more numerous and diverse are that individual's social relations. Social categories at the bottom of the social ladder are characterized by fewer social contacts and are more homogeneous. Social capital is an important resource in acquiring the following types of support: · instrumental (goods and services), · emotional (acceptance for actions or mobilization to take action), · valuation (affirmation that one is an important person for the interaction partner), · information-related (obtaining advice, guidance, instructions) · spiritual (having the sense that one is not alone in a difficult situation) (Kawula 2002). 12.2 Informal support networks as an important component of social capital have become more significant today in Poland, because the restitution of capitalist relations has led to unemployment and has resulted in decreased access to income from work. This deficiency is not compensated by social transfers, which are limited by the scarcity of public finances. Private support therefore replaces formal institutions and lowers subsistence costs, making it possible for a few people to survive. This is also an important means of alleviating the state budget. In the system of informal social security a special role is played by the family, which in Polish society for historical and religious reasons has always been an important institution of support. Today, however, remaining in stable and extensive family associations has become an instrumental necessity according to recent research. With the help of relatives existing social-living conditions are maintained (Giza-Poleszczuk [1993[,urek [1996, 1997, 2001]). 12.3 Such a strategy exists even in big cities, where the nuclear family model prevails. Here, families function as extended dispersed families, consisting of several nuclear structures living separately, but remaining in continuous cooperation (urek [2001]). The cooperation consists of mutually provided support in the form of material and financial assistance, services (called the flow of time or the gift of time) and provision of the use of a home (the gift of space) (Szukalski [2002]). A large, dispersed family supports its members regularly, replacing professional care institutions (usually a kindergarten), and provide support in crisis situations when a sudden 266 unfavorable event occurs (urek [2001]). The needs of small, nuclear families are satisfied primarily by the parents of both spouses, on a voluntary basis, and less often by more distant relatives and kinfolk (Giza-Poleszczuk, Marody, Rychard [2000]; urek [2001]). 12.4 The significance of informal networks of support should not be underestimated when a person finds himself or herself in poverty. However, the range of this support among poor people and the applied strategies of survival with the use of informal contacts continue to be little investigated in Poland. 12.5 Only quantitative research, where direct questions were asked about informal support networks functioning among poor people, was carried out by the authors of this chapter in the big city environment1. Lód and Katowice will serve as examples. B. INFORMAL NETWORKS OF SUPPORT OF THE CITY POOR (LÓD AND KATOWICE) Potential Networks of Support 12.6 The social contacts an individual/group has are resources of potential support that may be used when such a need arises. The following have been adopted as indicators of potential support: · social contacts maintained with relatives, neighbors, friends and colleagues, and · subjective reports of happiness and loneliness. 12.7 Research carried out in Lód and Katowice has shown that poor people2 for the most part have potential support groups and are within networks of informal social relations. Their social contacts are above all determined by the family relations (83 percent meet with their relatives), less often by relations with friends (62 percent) and neighbors (61 percent), and least often by career contacts (23 percent) ­ which is understandable because of their low occupational activity. However, the frequency of meetings, even with family members, is not high. Only every fifth respondent sees those relatives who live separately, and every fourth respondent his neighbors once a week or more often. Such a pattern of social contacts for poor people exists in both Lód and Katowice. 12.8 If one assumes that the family is of primary importance in the system of informal support, then those who do not maintain contacts with relatives or who do not have any relatives are in the worst situation, as they have to depend entirely on formal institutions. In Lód as well as Katowice the risk group of people remaining beyond the system of family support consists mainly of single men, without children, with low educational attainment. In Katowice the scope of exclusion from the family in this group of men is clearly higher than in Lód. In the former city nearly every other single poor man has no contacts with relatives, while in Lód this applies to somewhat more than every fourth man. This difference may result from the fact that in socialist times men were recruited for work in Silesia from all over Poland. Members of their family of origin probably lived far away and if only for this reason were unable to form their support network. It is also possible that single men in Katowice do not contact their relatives out of fear of revealing their failure. During socialist times, work in Silesia produced high incomes and gave prestige to the employee, particularly in rural circles, from which most of these men were 1Detailed information on the research on which this report is based can be found in Annex 1. 2Defined as social assistance clients. 267 recruited. Admitting to the loss of such a position might be harder for some than turning to public institutions or charitable organizations for help and therefore these men avoid contacts with relatives. However, in this group there are cases where the families do not want to associate with the respondents. This is because there are often alcoholics in this group who have been rejected by their families. 12.9 Total social isolation, expressed in a complete absence of informal social contacts, is very rare in the population of poor people. Only 6 percent of respondents from Lód and 4 percent from Katowice have no potential informal support whatsoever. A solitary life is led in Lód mostly by poor people with low educational attainment (accounting for 52 percent of all those isolated socially), living alone (62 percent), without any steady income (62 percent).1 12.10 A bigger group consists of persons who maintain relations with the family, with neighbors, with colleagues, and with friends. Such fully diverse social capital is available for 9 percent of social assistance clients of Lód and 16 percent of the social assistance clients of Katowice. This means that there are relatives, neighbors, friends, and colleagues in the potential support network of every fifth impoverished inhabitant of Katowice and every eleventh impoverished inhabitant of Lód. Diversity of the composition of the informal network appears to be particularly significant, since various groups can provide different forms of support. 12.11 Apart from contacts with other people, an important indicator of belonging to an informal network can be a subjective sense of happiness or loneliness. With a few exceptions, people associate happiness and the absence of a sense of loneliness with the fact of having people who are close to them that they can count on. If this assumption is correct, the decided majority of poor inhabitants of Lód and Katowice feel they have social roots. However, 15 percent of the Lód inhabitants and 13 percent of the Katowice inhabitants described themselves as very unhappy, and 15 percent and 12 percent, respectively, as very lonely. It appears then that the group of people who feel unhappy and very lonely is larger than the group who isolate themselves or are isolated from family, neighbors and friends. Thus, it is not only the fact of being outside the network of informal support that gives rise to a feeling of the absence of social roots. This may mean that the sense of security arises not only from the existence or absence of a potential support network, but also from any possible actions that relatives, friends, or neighbors may take (i.e., the "quality" of this network). Perceiving a Network as a Support Group 12.12 Transforming a potential support network into an actual one depends on various factors: · The diversity and "quality" of the participants in this network, i.e., their age, state of health, and economic, cultural, social (etc.) resources; · The degree of readiness of the recipient to ask for help; · The degree of readiness of the donor to provide support; · The concurrence of needs reported by a person in need of support with the resources that the possible donor has available. 12.13 The authors of research conducted on representative samples of Polish population have determined that poor people have "relatively small family capital, measured by the number of influential or affluent persons," but that in their lives an important role is played by family contacts that consist of mutual support of family members in small matters, through small material assistance serving to equalize or minimize the scarcity and deficiencies that they have to cope with (Giza-Poleszczuk, Marody, Rychard [1999]). Research carried out among the poor in 268 Lód confirms this state of things. The history of the life of 90 people shows that the circle of their relatives and friends consists of people of the same social status as themselves,(i.e., with low education, poor living conditions, an unstable career or unemployment). Although they are also working people, they usually hold low career positions and have low incomes. Therefore, they can mitigate the financial needs of their needy relatives, friends or neighbors only to a small extent. The sense of being "immersed in poverty," that is, of living in a circle of poor people, is felt by more than every other respondent in Lód (59 percent) and more than every third in Katowice (37 percent). 12.14 Relatives, neighbors, and friends nonetheless are perceived by poor people as real support groups that they can turn to in difficult moments. The indicator of perceiving a given category as an actual support group was the reply to the question of whom among persons not living with the respondent would the respondent ask for help if: · he or she were in danger, · he or she needed to borrow money for several weeks, · he or she had an important problem to be solved and needed advice, support. 12.15 The great majority of respondents stated that they had someone to turn to for such support. However, more than every fourth (27 percent) inhabitant of Lód and Katowice had no one who could lend him or her money, more than every fifth (22 percent) had no one to ask for advice, and 13 percent (Lód) and 16 percent (Katowice) had no one who would help in case of danger. In the matters referred to most frequently, support groups were neighbors and parents. The inhabitants of both cities were more inclined to turn to neighbors than to parents in the first two cases listed above, but could count on the support of both in the same way when they had a problem to solve. The Need for Support and the Received Support 12.16 The population of poor people is diverse in its perceived need of help from other people. The proportion of those who would like to obtain help depends on the kind of need. Everyday activities ( childcare, shopping, household repairs on chores) are not so arduous for poor people as to require the help of others (with the exception of home renovation and repairs). However, the need to obtain moral support and advice from others is urgent. As much as 80 percent of respondents from each city reported this need. However, the group of people who actually receive support is considerably narrowed down, which means that the practical usefulness of a support network is decidedly smaller than its potential usefulness. It turns out that more than every other respondent in need of help in the form of services (childcare, household repairs or chores, shopping) has no one to count on, apart from members of his or her own household. While such a pattern of real instrumental support is characteristic of the poor inhabitants of both Lód and Katowice, in Lód the perception of the lack of real support in the matters referred to is more frequent, and persons outside the group of relatives form a support group less often. 12.17 A resource that the poor resort to widely is psychological support (emotional and/or, spiritual support, information). Such support is obtained by nearly three-quarters of the respondents in each city. However, certain categories complain about the absence of such support more often than others. In Katowice this category consists primarily of middle-aged single mothers, and in Lód it consists of people with low educational attainment. 269 12.18 The real usefulness of support groups depends on the kind of help provided. In general, support is provided within the family. With the exception of repairs and renovation, where outside acquaintances are more likely to help, poor people can mainly count on family members, especially the closest family members. The family provides psychological support. Nearly every other respondent in each city asks for advice and receives moral support from relatives. 12.19 Functioning in a social environment requires social exchange and mutual support among people. The exchanges between poor people and members of their family, neighbors, or colleagues are asymmetrical. Poor people take more from others than they give back, in the form of goods and services and in the form of emotional support: · Nearly 40 percent of those in need of help obtain help in caring for a child, but only 18 percent of the people give this kind of help; · Nearly 50 percent of those in need obtain help with home renovation, and 18 percent give this kind of help; · Nearly 40 percent obtain help with household chores, and 25 percent give this kind of help; · Nearly 36 percent obtain help with shopping, and 25 percent give this kind of help; · Nearly 73 percent make use of psychological support, while 50 percent give advice to other people. 12.20 Persons remaining outside the system of information-emotional-spiritual support consist mainly of single men, of whom nearly half are not asked for information or advice. 12.21 Although in general the patterns of obtaining/providing support are the same among poor people in Lód and in Katowice, there are small differences in the extent of use of the individual resources. Analysis of data points to the following: · Help with child care provided by parents/in-laws occurs nearly twice as often in Lód as in Katowice (28 percent and about 15 percent respectively). In Lód this kind of support more often involves siblings and in Katowice more distant relatives (siblings: 3.9 percent in Lodz and 1.6 percent in Katowice; more distant relatives: 1.9 percent in Lodz and 3.6 percent in Katowice) · A greater portion of poor people in Katowice than in Lodz are helped by friends, colleagues, and neighbors with home renovation (34 percent and 25 percent receive, and 11.2 percent and 7.6 percent give, respectively) and with daily household chores (21.5 percent and 10.6 percent receive, and 9.9 percent and 6.1 percent give, respectively). · Inhabitants of Lód help their parents in daily household chores more often than inhabitants of Katowice (14.9 percent and 7.2 percent respectively). · The poor inhabitants of Katowice receive psychological support from friends, colleagues and neighbors somewhat more often than do their counterparts in Lód (30 percent and 25 percent respectively). The poor people of Lód give such support to their friends, colleagues and neighbors somewhat more often than do their counterparts in Katowice (30 percent and 25 percent respectively). · In all the analyzed forms of help, the poor people of Katowice called on the resources of more distant relatives, more often than did the poor of Lodz. The Family as the Basic Group of Real Support 12.22 It is clear that in Lód as well as in Katowice help and services for poor people are provided mainly within the family. Two essential patterns of family support can be distinguished: 270 · help for descendants · help for ascendants. 12.23 The first pattern is definitely more common and consists of services, emotional help, and support provided by parents to adult children directly or indirectly through support for grandchildren. 12.24 The forms of support for descendants vary. Usually they involve satisfying the most elementary needs of adult children deprived of income and even saving them from hunger and homelessness, and less often they involve improving the living standard of the younger generation. A young unemployed 30-year-old woman raising two children alone said: "My parents don't have much, only a small pension and [unemployment] benefit, but they save us with a grosz or two for bread and butter ...If it weren't for them, I don't know if I could manage, we would go hungry, or I would have to start begging" (Grotowska-Leder [2002], 254-55). 12.25 Relatively often parents ensure a home for their children by living with them. This is the case with a large portion of Lód families who are clients of social assistance. Poor people usually cannot afford to buy or rent an apartment for their children who set up their own families. Even if they do succeed in acquiring an apartment, the young generation is unable to cover the costs of maintaining the home. When the young generation lives separately, parents/in-laws often partly or wholly finance the housing fees of their children. Another common practice is changing an apartment in a housing settlement into two sub-standard housing units, of which one is meant for the son/daughter and the other for the parents (more often for a single mother). In such a situation the mother/father sometimes pays the costs of maintaining both apartments. In some situations, the mother/father looks for a partner for himself/herself and moves in with him/her, leaving his/her apartment for the children. 12.26 There are also quite a few cases of adult children with their families returning to their parents' home in the country. During socialist times young men and women left the villages to work in towns and cities, where they set up their families and rented an apartment. When they lost their jobs and were no longer able to keep their apartments, they returned to their parents to live with them and often also to live off their old age pensions. 12.27 A very common form of help in poor families is providing care for the grandchildren. At times this is 24-hour care when the small grandchild stays at the grandparents' home. More often, however, grandparents pick up their grandchildren from school, prepare meals for them, and buy clothing for them. One of the respondents said: "My grandson comes to me every Friday after school, he's here right after school. He comes on Friday and I take him back home on Sunday evening. That's always one less mouth to feed at my daughter's ... He got his school sack from his grandparents, his jackets from his grandparents. Yesterday he also came running and said, Grandma, you got your pension, you promised to buy me a flannel shirt, or a sweatshirt... Of course, I'll buy it for you..." (Golczyska-Grondas et al [2003] 175-95). When the grandparents live in the country, their grandchildren spend their vacations with them and are supported by them during this time. When the children go back to their parents, the whole family is provided with food by the grandparents. 12.28 Grandparents take in adult grandchildren as well in order to lower their costs of living, or they register them at their own home fictitiously (when the grandchildren in fact live with their parents); in this way after the grandparents' deaths the grandchildren have priority for being allocated this apartment. Family help also includes loaning small amounts of money and buying food. A young woman who shares a household with three daughters and her husband regularly 271 borrows money from her parents to buy home furnishings. She said: "My parents lent us money [for furniture], we agreed that we'll give it back in three months, but they always say that ... when we finally give it back, we borrow again for something else and so it goes on..." (Grotowska- Leder [2002] 255). A 50-year old man said: "My in-laws (the wife is deceased) help my younger son, they contribute to the costs of living since they're raising him... sometimes I see them almost every day... they don't help me financially, rather only by giving food, they lend some things, but not money" (Golczyska-Grondas et al [2003], 175-95). 12.29 Much less often parents or grandparents are helped by their grandchildren, and if so, this help is in the form of services rather than money. Adult children help parents/in-laws with household chores such as cleaning, cooking, and shopping and with renovation or repairs. Usually this is done when the parent is sick or infirm. However, older people say that when grandchildren undertake services for them, they expect to be paid for it. One woman said that her granddaughter visits her regularly on the day "when I get my allowance or pension, she won't stay long, she asks what she can do, but when I give her money for shopping, she brings it quickly and off she goes, saying she has no time. If I don't give her something from my pension she asks me to" (Grotowska-Leder [2002], 224-26). 12.30 Research carried out in Lód and Katowice shows that in poor families support is rarely provided by lateral relatives (siblings). However, social assistance clients receive help from siblings more often than they provide help to siblings. The most frequent form of support consists of handing down clothing to the children of the brother/sister that the donor's children have grown out of; less often this support is help with home renovation or repairs, or the lending of money. Brothers and sisters who are better off help their siblings materially, but there are times when services are provided the other way round. In our empirical material there is much evidence that siblings are a poor resource. Respondents say that their siblings are not well off, that they are unemployed, and that they are often addicted to alcohol. 12.31 Family support takes one-way or two-way direction, and sometimes (although less often) multilateral forms. Sometimes the parent (usually the mother) helps adult children (daughters more often than sons) materially, who in turn help their adult children. Family usefulness depends on gender. Mothers live longer than fathers and are able to support their children's families longer in various ways. They care for the grandchildren and run the homes of their children, and they search out less expensive products in supermarkets and also adjust clothing. The family usefulness of poor men is definitely less; in addition, poor men have not always fulfilled their parental duties properly. For this reason daughters and sons are not very willing to provide services for them, which can rarely be returned. Single men addicted to alcohol are the group most threatened with extreme social exclusion. Social exclusion is more probable in a poor neighborhood because spatial segregation is accompanied by progressing social isolation C. INFORMAL SUPPORT NETWORKS OF POOR PEOPLE FROM POST-STATE FARM (PGR) AREAS AND SMALL TOWNS 12.32 The daily life of poor people living in post-state-farm areas also includes contacts they maintain within the structures of formal and informal ties (Tarkowska [2000], Palska [2002]). 12.33 In the picture of poor inhabitants in post-state-farm villages, the widespread model is that members of poor families are excluded significantly from the closer social environment as well as from the more distant, supra-local environment. In the rural environment, which is undergoing 272 transformation, a characteristic feature of the lifestyle of people living in poverty is limited contacts with other people and with institutions (Tarkowska [2000]). The poorer contacts of poor rural inhabitants with institutions are the result of a poorer infrastructure and the prevalence of social ties based on direct contacts. 12.34 In satisfying daily needs, three formal institutions appear to play an important role in the life of poor former state-farm workers: · The first of these is the gmina3 center for social assistance. Poor rural inhabitants report to the gmina social assistance center usually as a last resort, owing to the related embarrassment and humiliation, which is much worse in the rural environment where the inhabitants know nearly everybody. · The school is another institution also involved in actions to help the poor through supplementary feeding of poor children. A special institution for the poor is the local shop which, through informal credit in the form of buying "in the notebook," resolves the daily problems of buying staple products. · Poor former state-farm workers also contact the employment office in the town or city; by registering as unemployed persons they have access to free health services, and to the Agricultural Property Agency, which is involved in organizing vacations for children (financing camps) and which supports their education (financial aid). However, some of the poor cannot afford even the trip to the seat of the powiat office to register, whereas the children of the poorest people often resign from the vacation and aid offers because a trip to camp or the continuation of education requires additional expenses for the family. 12.35 The main support for poor former state-farm workers is the general family ­ siblings, parents, grandparents. Many respondents say that when in need they can count on their family only. To manage their lives an extended family is needed and if for some reason there is no such family, (because of divorce, illness, or long distance), poverty becomes particularly difficult. 12.36 The poor rural families are similar to poor families in cities. What appears to be different is that rural families have fewer social contacts outside the family. 12.37 People living in poverty can count on their families, but they cannot expect to receive the help of neighbors. In many responses of poor inhabitants of post-state-farm areas, neighbors are presented in a negative way. Several reasons are given for this. First, the inhabitants of these areas are very poor. "Here everyone is as poor as me" (Tarkowska [2000]: 152). Therefore, there is no point in turning to anyone outside the family for help. Second, it is emphasized that people are unkind to each other, that they are even spiteful, and are not interested in helping another person. The poor are not ready to help others just as poor, while the "winners" in villages stay away from the poor. Relatively often rural inhabitants complain in their descriptions about the neighbors: "Here no one wants to lend money to anyone," "Nowadays people could drown you in a spoon" (Tarkowska [2000]: 153). 12.38 A low intensity of neighborly contacts is also characteristic of poor people in towns in the post-state-farm areas. People in these areas describe the feeling of being a stranger, the lack of kindness, and at best the indifference of their neighbors (Tarkowska [2000]. 3Gmina means the lowest level of administration in Polish government. 273 12.39 In the condition of poverty the social world is an alien and hostile world, with the exception of an expanded network of family ties. Examples of neighborly cooperation occur rarely, the prevalent picture being one of broken ties and a hostile environment against which problems are concealed - the best solution is to stay at home by oneself. Interactions with neighbors are limited to request for help in emergencies, when institutional forms of support and the family fail or are not enough. 12.40 In describing contacts with neighbors, a girl from a poor village family said that when "there was no money, my mother usually went [to the neighbors] to borrow money, to take something and buy something to eat" (Palska [2002]: 197). When describing poor rural environments, Palska noted that the social surroundings are "distant" and "close" at the same time. Because maintaining social ties requires outlays, such relations are limited, but at the same time it is acknowledged that breaking all ties is dangerous, as it can lead to being "cut off" from potential sources of help in case of emergency; therefore, people feel that they are forced to maintain a certain kind of closeness with their social surroundings (Palska [2002]: 196-97). 12.41 Familiarity which is coerced, being rooted in the closest family ties, is another feature of the poor who are undergoing pauperization because of unemployment. Rossa described the dimensions of marginalization of this category of poor people living in Gorzów Wielkopolski in one of the housing settlements located around industrial plants where employment was significantly reduced.4 Rossa showed that the poor people living there are enclosed in a small circle made up of their family and other people who are like family. According to Rossa, this is so because these people do not attract others as donors, because they are unable to offer much. 12.42 Cutting poor people off from circles where ties are based on what they have to offer, makes poor people intensify contacts within the family, but there are also times when they restrict these contacts. What they can offer includes: economic capital (money and other material goods), cultural capital (education, interests), and social capital (contacts, possibilities). 12.43 Family members become closer to one another when, in Rossa's words, there is a "mechanical" or "limited" integration of the family system, (i.e., when family members have a common fate, a similarity in this situation). A situation made up of limitations makes it necessary to seek support in daily activities within the family. Such relations, particularly when family members have a common system of values and there is good communication among them, make the family an "organic" whole. In the opinion of Rossa, marginalization may also lead to a limitation of family contacts and may even contribute to pathological phenomena. Scarcity gives rise to "emotional divorce," "privatization" of the lives of family members, and conflicts in which there is also violence (Rossa [1997]: 158). 4The research was carried out in 1998 within the Ordered Research Project "Social policy of the state in the process of reconstruction of the political and economic system." The research material on poor people was gathered by way of an unstructured interview, in part tape recorded, with social workers. 274 12.44 The narrowing down of the social world that is characteristic of poor inhabitants of villages and small urban settlements leads to greater marginalization. Social marginalization combined with self isolation, is an important problem for social policy. If one considers the inadequacy of public state and nongovernmental institutions of support it is easier to understand why a large portion of poor people (especially in rural areas) appear to be left to their own devices to resolve everyday problems. This situation is conducive to intensification of the new Polish poverty, that is, poverty among young adults rather than among the aged and disabled as was the case in the past. D. FINAL REMARKS 12.45 According to our research, it is misleading to assume that poor people are passive and devoid of any support. Like those who are not poor, they take advantage of all possible resources. The use of informal support is shown in various surveys.5 However, the scope of real support varies because the availability of a family support network and its "quality" varies. In areas that during socialist times were destinations for people seeking work (Silesia, post-state-farm areas), the usefulness of a dispersed extensive family as a support group is smaller because of its spatial distance. As Domaska (2000) pointed out, the inability to move closer to the family was made worse when the people from the former state farms were, in a way, tied to their place of residence as a result of being offered the option of buying their occupied housing unit at a low price. Although the price was affordable, it often exhausted all of their savings. People bought their apartments not realizing that in a degraded locality with difficult road access and with no work opportunities, they would not find buyers for their apartments. Thus, the apartment became a prison. 12.46 Another factor limiting the effectiveness of the family network is the frequent occurrence of unstable forms of family life primarily found among people living in poverty enclaves and showing pathological behavior. Frequent changes of partners, resulting in children living in various, often unstable relationships, makes family identification difficult and impairs the sense of belonging to a circle of relatives and the sense of responsibility toward them. 12.47 However, changing partners and having successive children with the new partner is often done with the intention of obtaining support for existing children, according to research carried out in Lód. If the former relationship turns out to be ineffective in providing financial support, a woman will look for another partner in the hope that the new partner will support not only his child, but also the woman's children from previous relationships. Thus, the general interpretation of such behavior as irresponsible is oversimplified. 12.48 In rural and small-town communities, where people know each other, there is a particular kind of non-family form of informal support, rarely or never practiced in large urban areas. This is the institution of shopping "in the notebook." In this system the grocery shop owner sells goods on credit, writing down the debt in a notebook. When the debtor receives a benefit, he/she repays the interest-free credit and has a new open "debit account." Such a transaction is based on trust and is totally informal. It is advantageous for both the shopkeeper and the poor people. The shopkeeper obtains his/her dues, even if with a delay, while the customers are able to satisfy their basic needs. The institution of buying "in the notebook" is a replacement of a banking service that 5Carried out in post-state-farm villages (headed by E. Tarkowska ­ Tarkowska, 2000; Tarkowska, 1998; Tarkowska and Korzeniewska, 2002, and headed by Z. Kawczyska-Butrym ­ Kawczyska-Butrym, 2001), in towns (Rossa, 1999) and in cities (headed by W. Warzywoda-Kruszyska - Warzywoda-Kruszyska and Grotowska-Leder, 1996, Grotowska-Leder, 2002). 275 is inaccessible to poor people. In this system, credit is granted according to the shopkeeper's knowledge of a person's ability to repay. E. CONCLUSIONS 12.49 The informal redistribution of income and the provision of services inside a dispersed extended family makes it possible for poor people to function and prevents the outbreak of uncontrolled social discontent. However, when state or nongovernmental agencies provide insufficient support, that burden is shifted to the shoulders of the oldest generation, which supports its adult children and grandchildren in various ways. In Poland today, the direction of money flow between generations has been reversed. In the old days, according to Sikorska (Sikorska 1998), even in the 1980s, adult children financially supported their parents, whose old age pensions were relatively low. Today old age pensions are still not high, but in many families they are the only source of regular income. The middle and younger generations of people at the lower end of the social ladder have no work; thus, the burden of their support is now borne by the oldest generation. In the social security reform and social policy strategy currently under preparation, this function of old age and sickness pensions is not taken into account; it has become the source of subsistence for a much greater number of people than the insured person alone. The strict verification of the right to a sickness pension being introduced by the government, along with the less generous amount of old age pensions, correct in its macro-social financial aspect, may have effects opposite to those intended: instead of decreasing the burdens of the state budget, these new policies may increase the burden on state budgets. People who until now have been supported by relatives will have to receive much more generous support from social assistance if social calm is to be preserved. However, this would mean that the responsibility for social support would be lifted from the family and placed explicitly on the state. · In the near future, when the oldest generation dies, the situation will become worse. Among poor people from the middle generation only a few will be entitled to a full old age pension because of the inadequate old age contributions paid during the years of unemployment. This means that these people will be unable to help their children and grandchildren in the way that their parents do now. · Informal networks of support mitigate the current difficulties of poor people but cannot draw them out of poverty or prevent the inter-generational transmission of poverty. Grandparents can buy clothing for their grandchildren or can give them food from time to time, but they are unable to finance expenses that will increase their human capital through improvement of intellectual development. The educational model practiced in Poland is based on the private financing of the acquisition of the knowledge and skills that will increase chances for social mobility. Expenditures for this purpose are beyond the financial capacity of even extended families. It is necessary to have a government program of support for children from poor families and regions, aimed at improving their human capital and not merely at supplementary feeding. 276 REFERENCES Beskid, L., and L. Zarzycka-Skrzypek (1993), ,,Zachowania przystosowawcze gospodarstw domowych" (Adjustment measures in households), in: L. Kolarska-Bobiska (ed.), Ekonomiczny wymiar ycia codziennego. Raport z badania `92 (Economic dimension of everyday life. Report from 1992 survey), Warsaw: CBOS. Czapka, E. (2001), ,,Radzenie sobie pracowników bylych PGR-ów i ich rodzin w codziennym yciu" (How former state-farm workers and their families manage in everyday life), in: W. Warzywoda-Kruszynska, E. Komicki, and H. Januszek (eds.), Bieda na wsi na tle globalnych przemian spoleczno-gospodarczych w Polsce (Poverty in rural areas and global socio-economic changes in Poland), Pozna. Daszyska, M. (1998), ,,Korzystanie gospodarstw domowych z pomocy" (How households obtain help), in: Zrónicowanie warunków ycia ludnoci w Polsce w 1997r. (Differentiation of living conditions of people in Poland), Warsaw: GUS (Central Statistical Office). Domaska, L. (2002), ,,Deprywacja potrzeb bylych pracowników pastwowych przedsibiorstw gospodarki rolnej" (Deprivation of the needs of former workers of state owned enterprises of agricultural economy), doctoral dissertation defended at the Faculty of Humanities of the Nicolaus Copernicus University in Toru, typescript. Giza-Poleszczuk, A., M. Marody, and A. Rychard (1999), ,,Strategie Polaków wobec zmiany systemowej" (The strategies of Poles and systemic transformation), in: J. Kurczewska (ed.), Zmiana spoleczna. Teorie i dowiadczenia polskie (Social change. Theories and the Polish experience), Warsaw: IFIS PAN (Institute of Philosophy and Sociology of the Polish Academy of Sciences). Giza-Poleszczuk, A. (1993), Rodzina i system spoleczny" (The family and the social system), Studia Socjologiczne No. 2. Giza-Poleszczuk, A., M. Marody, and A. Rychard (2000), Strategie i system (Strategies and the system), Warsaw. Golczyska-Grondas, A., M. Potoczna, and W. Warzywoda-Kruszyska (2003), ,,Wielkomiejska bieda" (Big city poverty), in: E. Tarkowska, W. Warzywoda-Kruszyska, K. Wódz (ed.), Biedni o sobie i o swoim yciu (The poor about themselves and their lives), Katowice: Publ. lsk (in print). Grotowska-Leder, J. (2002), Fenomen wielkomiejskiej biedy. Od epizodu do underclass (The phenomenon of city poverty. From episode to underclass), Lód: Lód University Publishers. Kawczyska-Butrym, Z. (2001), Mieszkacy osiedli bylych PGR-ów o swojej sytuacji yciowej" (Inhabitants of former state farms on their lives), in: Z. Kawczyska-Butrym (ed.), Report on survey, Olsztyn: Studio Poligrafii Komputerowej SQL. Kawula, S. (2002), Pomocnoiczo i wsparcie. Kategorie pedagogiki spolecznej (Help and support. Categories of social pedagogy), Olsztyn. 277 Koralewicz, J. (1987), Autorytaryzm, lk, konformizm (Authoritarianism, fear, conformity), Warsaw. Milic-Czerniak, R. (1999), ,,Zachowania przystosowawcze do nowych warunków ekonomicznych" (Behavior adapting to the new economic conditions), in: L. Beskid (ed.), Zmiany w yciu Polaków w gospodarce rynkowej (Changes in the lives of Poles in a market economy), Warsaw: IFiS PAN (Institute of Philosophy and Sociology of the Polish Academy of Sciences). Palska, H. (2002), Bieda i dostatek. O nowych stylach ycia w Polsce koca lat dziewidziesitych (Poverty and affluence. On new lifestyles in Poland of the late nineties), Warsaw: Publishers: IFiS PAN. Rossa, J. (1999), ,,Wymiary marginalizacji" (The dimensions of marginalization), Problemy Polityki Spolecznej No. 1. Sikorska, J. (1998), Konsumpcja. Warunki, zrónicowania, strategie (Consumption. The conditions, differences, strategies), Warsaw: IFiS PAN. Szukalski, P. (2002), Przeplywy midzypokoleniowe i ich kontekst demograficzny (Inter- generation flows and their demographic context), Lód: Lód University Publishers. Tarkowska, E. (1998), ,,Ubóstwo w dawnych PGR-ach: w poszukiwaniu dawnych ródel nowej biedy" (Poverty in former state farms: the quest for old sources of the new poverty), Kultura i Spoleczestwo No. 2. Tarkowska, E. (ed.) (2000), Zrozumie biednego (How to understand a poor person), Warsaw: IFiS PAN. Tarkowska, E, and K. Korzeniewska (2002), Mlodzie z bylych PGR-ów. Raport z bada (Young people from former state farms. Report on survey), Warsaw: ISP. Warzywoda-Kruszyska, W. (ed.) (1998), y i pracowa w enklawach biedy (Living and working in poverty enclaves), Lód: Institute of Sociology of Lód University. W. Warzywoda-Kruszyska (ed.) (1999), (y) Na marginesie wielkiego miasta (Living on the margin of a big city), Lód: Institute of Sociology of Lód University. W. Warzywoda-Kruszyska (2000), ,,Koncentracja i gettyzacja ludnoci biednej w Lodzi - porównania midzydzielnicowe" (Concentration and ghettoes of poor people in Lód ­ comparisons between districts), in: W. Warzywoda-Kruszyska, and J. Grotowska-Leder (ed.), Ryzyka transformacji systemowej (na przykladzie Lodzi) (The risks of systemic transformations /on the example of Lód/), Lód: Institute of Sociology of Lód University. Warzywoda-Kruszynska, W. and J. Grotowska-Leder (1996), Wielkomiejska bieda w okresie transformacji (Big city poverty during transformation), Lód: Institute of Sociology of Lód University. Wódz, K. and Lcki K. (2003), ,,Ubóstwo w regionie pogranicza kulturowego" (Indigence in the region of cultural borderline), in: E. Tarkowska, W. Warzywoda-Kruszyska, and K. 278 Wódz (eds.), Biedni o sobie i o swoim yciu (Poor people about themselves and their lives), Katowice: Publ. lsk (in print). urek, A. (2001), ,,Relacje zachodzce pomidzy miejska rodzina mal a rodzina du" (Relations occurring between a small and big urban family), Roczniki Socjologii Rodziny (Family Sociology Annals), Vol. 13. urek, A. (1997), ,,Rodzinny kontekst ycia jednostki a jej osadzenie w mikrostrukturze spolecznej" (The family context of an individual and his place in the social micro- structure), Roczniki Socjologii Rodziny (Family Sociology Annals), Vol. 9. urek, A. (1996), ,,Orientowanie si na rodzin a orientacja indywidualistyczna we wspólczesnym spoleczestwie polskim" (Orientation on the family and individualistic orientation in today's society in Poland), Roczniki Socjologii Rodziny (Family Sociology Annals), Vol. 8. 279 Annex 1: Empirical Foundations of the Report The report is mainly based on the findings of research carried out by the authors themselves, in Lód, in the years 1993-1999, with the application of quantitative as well as qualitative techniques. These techniques consist of: · Questionnaires filled out by a random sample of social assistance clients in Lód and in Katowice, within the project of KBN (State Committee for Scientific Research); The Urban Poor. A New Social Stratum? Every fiftieth household on the list of clients in Lód in 1993 and in Katowice in 1994 was drawn. The Lód sample consisted of 1,000 households (with 2,841 persons); the Katowice sample ­ consisted of 500 households (1,265 persons). Additionally were conducted in the homes of the respondents, with persons collecting a benefit.6 · Quantitative research within the project commissioned by KBN: Forms of Poverty and Social Hazards and Their Spatial Layout in Lód, carried out in 1997-1999 for the purpose of disclosing poverty enclaves in Lód. The poverty enclaves were designated on the basis of the poverty rate indicator (the percentage share of household members supported by social assistance among all inhabitants of the given area), counted for the individual quarters of streets. A poverty enclave was defined as at least two quarters of streets adjacent to each other, where at least 30 percent of the inhabitants live in households supported by social assistance. · Qualitative research within the projects Forms of Poverty and Social Hazards and Their Spatial Layout in Lód and Social History of Poverty in Central Europe. The study Social History of Poverty in Central Europe: The Polish Case (No. 97-1-114) was carried out in 1998-1999 within the project The Social Costs of Economic Transformation in Central Europe, under a grant received from the Institute for Human Sciences in Vienna. Family life stories of 30 Lód families were gathered, in which the middle generation (40-50 year olds) was on the list of social assistance clients and lived in one of the 17 poverty enclaves of Lód. In the 30 families, two generations forming two separate households were studied (in descending or ascending order), and in 10 families, three generations were studied, also forming separate households. The families were selected at random. Among the families indicated by social workers operating in the individual enclaves, the sample included families in which at least two generations consented to take part in an interview. · The changes that took place over five years in financial situation and social standing were traced in the case of 300 families of Lód registered in 1993 as collecting social assistance benefits. These families were in the Lód sample drawn for research within the Urban Poor project. For longitudinal research, within The Phenomenon of City Poverty project, every third household from this sample was drawn. In the project both a questionnaire interview and a family life story were used.7 6The findings were published in: Warzywoda-Kruszyska, Grotowska-Leder 1996. 7Focusing solely on social assistance benefit recipients we are aware that not all, but only some types of poor families are included within the range of interest, namely those which are also described as belonging to "official" or "administrative" poverty, i.e. fulfilling the criteria established in the Act on Social Assistance. However, it is obvious that social assistance beneficiaries are a population relatively poor in comparison with the "main portion," the "majority" of the population. The decided majority of them has basic needs that are satisfied, and sometimes the standard of their life, measured by a criterion other than the income criterion that does not always differ from the 280 In the report, attention has been focused on big-city poor defined as household members supported by social assistance, living in two large Polish cities: mainly in Lód and to a lesser extent in Katowice. In the interpretations of the obtained results we refer to the findings of other research ­ carried out mainly in the post-state-farm villages, since it also included social assistance clients ­ which widens the explanatory range of the formulated conclusions. remaining, "normal" members of society. We have included them in our research because they have fallen into indigence during the transformation period and because it is from this group that the permanently poor will be recruited, forming the big-city underclass. The point of departure of our research was the assumptions of the structural paradigm, maintaining that the poor constitute a group, or social category, distinguished by the very fact of being supported by social assistance (collecting a benefit). At the same time, however, we have assumed that this is an internally diverse category, also in the ways it functions. 281 Annex 2: Life Strategies ­ Case Studies of Two Lód Families Case 1: Family R (Grotowska-Leder 2002: 231-234) In family R, the generations of parents, children and adult grandchildren collect benefits. The family description has been formulated on the basis of the statements of a representative of the oldest generation (68 years) and her daughter (49 years). The extended family currently consists of the mother, two daughters and a son, three adult granddaughters, an adult grandson and two great-grandchildren. The oldest person is a widow. The educational attainment in each of the generations is very low ­ one of the daughters has completed a basic vocational school, the other has ended her education at the primary school level, and the son has completed one grade of a basic vocational school. Their personal lives are not regulated either ­ one daughter is a widow, the other is divorced, and the son is in separation; the grandchildren are living in informal relationships. The son has been punished by imprisonment many times for not paying alimony. The entire family lives together in a "social" housing unit in an old tenement house in a city-center poverty enclave. It consists of ten people, four generations in three rooms, officially occupied by separate households. Only the oldest daughter, a widow, lives separately with her boyfriend, but officially is registered in the social apartment of the mother. No one in the family has a steady job at the moment; they live off benefits from social assistance. The two oldest generations earn some money by collecting things, penetrating the neighborhood bazaars and markets in search of objects needed to live. They also draw income from the sale of things they find, scrap metal and paper for recycling. The main expense in the family is the rent for the apartment and the purchase of cigarettes, but they are often behind in rent payments. They deal with access to electricity "in another way," - they are illegally connected to the grid. Contacts are very frequent within the family and include various forms of help. The proximity of the homes means that the occupied space is practically being used by all members of the family. The grandchildren sometimes sleep at the grandmother's, sometimes at their parents' and at times at their aunt's or uncle's. "I don't know where the great-grandchildren sleep, at home, at their aunt's, or grandmother's, where they happen to fall asleep they stay." The family members have a very active social life, not only within the family, but also with neighbors and friends, since alcohol accompanies the household members every day "you buy a bottle, someone always drops by, stays a while, has a good time..." Family and neighborly relations also include providing help, even when it is not asked for. The family is surrounded by others like them. Its neighbors are people also on welfare. In the tenement house where they live everyone has some sort of benefit, no one has a steady job. The grandmother who works by collecting things says, "Sometimes I bring so much of everything that my granddaughters are unable to eat everything at once, so I give some away to the neighbors, they're in need too." Nearly everyone takes care of someone at times. The same woman says, "Sometimes I watch over the grandchildren, when they oversleep for school (...) if I don't get back on time and their mother leaves, we ask the neighbor." A younger family member lives with her boyfriend, together with his mother. She receives a social assistance benefit, but her partner does not know this. She talks to the social worker at her mother's because her boyfriend would not allow the visit of a social worker in his 282 home, but probably does not want to disclose all his income. The respondent lives in poverty and is totally excluded socially. "Everything I'm wearing, from head to toe, is things that have been found. I've forgotten by now what it's like to buy things in a clothing store." "Food or other things, everything is brought over (...) there's some bread in the bag on the basket, just touch it, it's fresh, why shouldn't I..." Her boyfriend cleans up the area around a meat shop, "he gets some bones for soup, sometimes even some leftover ham, sausage unfit for sale, people won't buy old food..." so that they don't spend much on food. Only the children visit a doctor, the grownups treat their illnesses by themselves because they are not entitled to use the public health services. "Either a person saves himself or he dies." Case 2: FAMILY S (Golczyska-Grondas 2003: 175-195) Family S lives in one of the poverty enclaves of Lód, and is made up of the father (a nearly 50-year old widower named Z) and two sons: 17 and 25 years old. The family has been receiving social assistance for three generations. Successive generations have a similar low educational attainment, criminal behavior, diseases and addictions. The life of family S runs in a closed poverty enclave, they live here, meet their life partners here, keep social contacts and die here. Z was raised by his mother, an alcoholic, and her successive partners; he spent several years in orphanages. The family is an illustration of the stereotype of entrenched pathological poverty waiting for benefits. There were strong emotional ties between the father, his wife (now deceased) and the youngest son, whom he tries to care for as best he can. The financial standing of Z's family of origin was bad. His mother (who had been a concierge and did small handicrafts) "... she borrowed money, can't say she didn't, she was on borrowed money every month, the pension would come (...) she'd pay back and borrow again, we were on borrowed money all the time. This friend of hers, who lived with us, I called him uncle, he would help, but those were just peanuts, since he used to drink too, he would drink and didn't help much, but Mother, she counted her pennies all the time. Can't say there wasn't enough food..." Today, the legal income of the three-member household consists of PLN 100 from social assistance and a food card for PLN 100, and the rent is PLN 150 a month, the home has a debt totaling PLN 30,000. The social contacts of family S are limited to family contacts. From the beginning, the in- laws used to help the family. The wedding took place at their cost. "Mother didn't help me because she was on a pension then, she didn't help much for the wedding because she had eight hundred zlotys for a pension, or six hundred during communist times, or something, not much, didn't have enough to help from that. The in-laws and me, together we fixed this all up, I had a friend who played the accordion, you know, a tape recorder, the accordion, we partied for a week, poorly, but we had a good time." Two weeks after the wedding a son was born. The young couple together with the child lived with the in-laws; after several months because of frequent conflicts they moved to the apartment of Z's mother. 283 "Later, after the wedding, there were quarrels with the in-laws, running the home together, you know how it is, I didn't allow for much, my father-in-law also liked the bottle, he would get carried away, there were fights usually, I tell my woman, no way, let's go to Chojny, to our mother." In practically each matter the parents helped. When Z was in prison and there was not enough money for food, "Mother would sometimes bring something, there was bread and something to go with it." When Z's wife went to work in a camp and she had to be brought home, "I say to my father-in-law, Dad, let's go to Bydgoszcz and get her, no point in waiting." The wife's death caused a crisis in the family. Z went on an alcoholic binge: "...when my wife died I broke down (...) But soon enough I came to, people helped me and said, friends, neighbors, get a hold of yourself, you still have the youngest son... the younger one was 11 and a half. I didn't go sober for 8 months, after my wife's death I didn't get sober (with sadness) (...). The children were neglected, you know... But the in-laws frowned on all this, they didn't like it, I said leave me alone when I'm drinking, but I woke up, sure enough, for the child's sake, it's not his fault, otherwise they would do something with him, take him to an orphanage the same as me, I wouldn't want that." The grandparents help in feeding the older non-working alcoholic son, which Z holds against them: "It's the in-laws' fault, because when my son isn't drinking, he can sit all day ­ breakfast, dinner, supper, with the in-laws. The in-laws will buy him shoes if he doesn't have any, they feed him all the time, when he's sober. Because when he's drinking he won't go there (...) he can stay there three nights even (...), the in-laws will give him breakfast, dinner, supper, all the time. But I told them, there will come a time when you're no longer there, what'll he do then, he'll grow a beard, drag a cart and collect scrap paper, like they do nowadays, you won't be giving him things all your life. He's taught to be like that and that's that. Now he says, why should I work, I'll go to my grandparents, they'll give me something (...) The grandparents have a pension of over PLN 1,000, so they always eat something better, why should he eat cheap sausage here when he can go there and he's got a character like that, he won't bring anything for the younger one, from supper a slice of something, some cheese or better sausage, he'll stuff himself there and that's that. I myself told my mother-in-law, mother, don't do that, when you give supper to one of them, you should give it to the other, too (...) she'll just stuff it into him instead of sending him off to work (...) I won't give him things (...) buddy, you go your own way because I've still got K to raise and I won't work for you because I'm too sick for that..." The grandparents support the younger grandson, too, when Z is drinking, and sometimes give small loans: "When they know I'm all right, that I run this house, that I don't drink, they'll lend me money, 10 or 20 zlotys, like right now I still owe them 10 zlotys. Mother, wait, as soon as I get the social benefit I'll give it back to you, and 284 so she waits, poor thing, waits and waits, it's not a lot of money, but I wouldn't go further and ask for 100 or 50, it's not quite right, you know, but they'll lend to others. Like their daughter, because they have two more daughters, my two sisters-in-law, and my woman is their third daughter (...) They can go to their parents (my in-laws) sometimes and borrow 100 zlotys and they'll lend them because they'll give back the money on time. But to me, you know, they'll lend 10 or 20 because they know I don't have anything to give it back from. And God forbid that I should take up the bottle like after the wife's death, well, well, good-bye with any money, they'll call the younger one, they have to make dinner, make supper, well, I used to wander around, you know, I didn't think about the kid, so they would call him for dinner." The family is always in need of money. In the 1980s the couple lived off irregular jobs. They also received social assistance. "...Mother ­ says Z's older son ­ wasn't ashamed of anything, she wasn't afraid, if she had anything to say, she said it honestly and openly what she had to say, she would go directly to someone and say it, when she was still alive she used to go to social assistance, she collected money and cards, like Dad does now..." The family is currently also being supported by charitable institutions. The younger son has "...free dinners, I arrange it through social assistance, he has them at school, and when he's home I cook dinner every day, the in-laws can confirm this (...). I buy only for him and no more, for today and tomorrow (...) But, you know, a son is a son and I won't refuse to the other one when he sees it, there's the refrigerator so that I cook dinner for two days. When Sunday comes, I try to cook two dishes, I'll make ground meat, or (...) I'll buy a chicken breast and especially for the other one I have to make a pork cutlet with an egg, with breadcrumbs, only one, the other one only watches, only stares at what the other one's eating (...). And I fix something for myself with him, a fried egg or ground patties, like for Sunday." The older son says: "I go to the aid center [Caritas ­ church charity organization] and go to the church. Well, now I have a break because the younger one isn't here, he's at camp, but sometimes I get things, at Promiskiego there's the church, they give things there, too, a loaf of bread and that's the help you get in that church (...), because that's not my parish, so I looked in there several times. But for good I have a card registered in Caritas on Gdaska, that's Upper Caritas, where he's gone for camp, there are priests there. So I get things from there, I have a fixed hour and month there and my name and it's posted on the door and there you can get everything, clothing, shoes for the kid and for me (...) It's only food and clothing there, no money and it's once a month, every month I can go only for food, because with clothing they have it all written down in a notebook and they know if I've already been given something, so they won't give as much (...). Once a month they'll give me, you know, several cans of these pâtés, they gave some canned juice for the kid, sugar, flour, bread." Z's and his son's contacts are limited to the closest family: "I go to the in-laws to play cards (...) My free time, like I told you, I won't go to the park, I won't look for a woman in the park, well, because of the older one I can't settle down in any way, I'll just visit the in-laws or go to my 285 sister's and sit there, like I was there this morning, I looked in because sometimes she needs to have something done, because my brother-in-law, her husband has died, so I'll look in on her, and that's my visiting people, my sister, father-in-law and sister-in-law." His son says : "Usually my mother's sister drops by, only she was in a coma once, she woke up from it and has a hard time walking, but she comes. Or Grandma looks in, she lives on the other side, so she drops by, Mother's other sister comes here sometimes too. As for friends, they're from work... As for interests, I don't have any, not at this moment, so I either sit at home or go off somewhere with my friends to talk and so on but I won't sit all day, especially now that the television is broken, so there's nothing to sit by and watch." Z has friends only when he drinks: "...I don't have friends, when I drink I do, but it's, you know, I'd rather not look for them, I'm afraid to drink, now that I know my character, that I'm too soft, I can't say no when I drink then I'll look for drinks for a week afterwards, I'm afraid, I don't want to drink, then friends appear, you meet so and so, they live close by, but it's only 'hi there,' I don't bring anyone here, I'm a loner, I don't go out to any friends." 12.50 Their contacts with neighbors have never been good. The nature of neighborly relations has been the result of the alcoholic and criminal lifestyle of the family. Z recalls a situation from childhood, when as a result of the neighbors' reaction his mother ended up in jail and he in an orphanage, because his older brother did not take care of him. "It all happened because of the neighbors, sort of, we had some mean ones, Mother used to have this friend, an uncle, when father wasn't there he used to help out and I remember him as an uncle only, because he used to live with me almost until his death, and because of him Mother started to look into the bottle often. There was this quarrel, Mother quarreled with the neighbor over me, this neighbor also had a son and us boys, we quarreled over something or other, we had a fight and my mother was after a drink or two, she didn't need much, so she said a few words, the other one made a phone call, they came right over, it was communist times (...), she went out on the street, then had a disciplinary case and got three months in jail. When they closed up Mom, there was no one to look after me and I was still a minor, my brother didn't remember, so they took me to the Guardian Home, I came out after three years, I stayed a whole three years there." Contacts with neighbors are an occasion for them to make some money. Since Z is a trained house painter and the older son a construction helper, sometimes they both work at their neighbor's, a shop owner, in exchange for food. Z says that he is a loner and that is why he does not keep contacts with the neighbors: "...me, I'm a loner, you can ask the neighbors, the ones that don't drink, me I don't waste a drink. Every one who lives here, one, two, three, four tenants, I'm the fourth, they all enjoy a drink, but not too much because they work upstairs, these neighbors all live well, they have elegant homes (...), but I won't bring them here, no. When they were moving in I helped them, you know, carry the furniture upstairs, so for a welcome, when my wife was still 286 alive, we went upstairs, it ended with one liter and now it's just good-day, good-day. No loans, you know, not one zloty, I don't like going to ask, I don't want any." 287 13. CIVIL INVOLVEMENT AND CIVIL SOCIETY ORGANIZATIONS AS INFORMAL SAFETY NETS Piotr Gliski, Katarzyna Górniak and Hanna Palska A. CIVIL SOCIETY ORGANIZATIONS (CSOS) IN POLAND 13.1 Civil society organizations in Poland operate mainly on the basis of provisions stipulated in the Constitution of the Republic of Poland, the Law on Foundations dated 1984 with further amendments, and the Law on Associations, dated 1989. The operating conditions of non- governmental organizations (NGOs) are also regulated by a number of other legal acts (for example, the Act on the National Court Register dated 1997 and the Act on Public Finances dated 1998). On May 20, 2003, after seven years of persistent efforts on the part of the non- governmental circles and the central administration, the President of the Republic of Poland signed the Law on Public Benefit Organizations and Volunteerism. This law is a significant step on the path to the institutionalization of CSOs in Poland. The Act regulates a new model for the relationship between the state and the CSOs, based on cooperation instead of dualism (in which the state and non-governmental organizations were separately predominantly responsible for the funding and provision of social services). The development of cooperation and partnership involves, among other things, the possibility for NGOs to take up initiatives, such as submitting an offer for the delivery of public services (Article 12 of the new Law). 13.2 By promoting cooperation between the state entities and the CSOs, the law is a milestone in the process of making CSOs the third sector of the social order. This is supported by another law, according to which CSOs will acquire the status of public benefit organizations, on a real estate tax-exempt basis (Article 24, Section 1) and will be allowed to acquire, under special conditions (e.g., without a tender), the right to use real estate owned by the State Treasury or a local government unit (Article 24, Section 2). It is also important to underscore Article 26, which allows the public benefit organizations to access the public media in order to disseminate information about their activities free of charge. This law should be perceived as a significant element of state reform in the direction of public governance and of the movement of a part of social services to the non-profit sector, with the retention of state funding. Moreover the act will significantly facilitate the operations of Polish CSOs through tax exemptions, the contracting of services, the possibility for the citizens to appropriate 1 percent of their tax for supporting activities of public benefit organizations, etc. It will also regulate the legal status of volunteers (for example, it will provide them with the right to insurance coverage and to compensation for costs incurred during voluntary work). However, the new Law has been criticized for being insufficiently socially-oriented (many decisions concerning CSOs remain within the competences of state administration instead of the civil bodies), for benefiting only strong and professional organizations and for failing to regulate numerous pathological areas and social phenomena that accompany non-governmental organizations (e.g., the issue of patronage between NGOs and the local authorities sector or the central administration sector). 288 13.3 In Poland there are 41,859 registered NGOs (36,791 associations and 5,068 foundations). Taking the broader definition of a non-governmental organization8 (i.e., including trade unions, political parties, voluntary fire brigades, parents' committees, churches and religious congregations) this number would amount to 96,000. According to the estimates of the KLON/JAWOR Association (using the narrower definition of an NGO, namely, associations, foundations and social organizations), only around two-thirds of the registered entities are active, 91 percent were established after 1989, 30 percent are less than three years old and, among those that ceased operations, 70 percent had ceased within the first three years of operations. 13.4 The civil sector employs over 100,000 people, which constitutes, according to various estimates, from 0.58 percent (KLON, 2002) to 1.2 percent (Institute of Political Studies of the Polish Academy of Sciences ­ ISP PAN) of the employed outside of agriculture. Fifty-five percent of organizations do not employ any regular staff. At the same time, in the widely understood sector, approximately 3.3 million volunteers are working (11.1 percent of the Polish population). 13.5 The latest research on the non-governmental sector in Poland (KLON, 2002) shows that the largest number of organizations (36.5 percent of the total or over 13,000) operate in the area of sports, recreation, tourism and leisure; while 12.4 percent operate in the area of education and upbringing; 11.6 percent - in health protection, rehabilitation and assistance to the disabled; 10.1 percent in art, culture, monuments and tradition protection; and only 4.6 percent (fifth position in the hierarchy) in the area of social aid, self-aid and charity work. These are significant data, as they correspond to the results of the flow of public funds from the central administration to the non-governmental sector, which underscores the particularly privileged position of sports organizations and clubs. To a large extent this situation is caused by a lack of vision and a lack of a coherent policy towards NGOs, which is a consequence of maintaining the old post-communist structures of patronage (as in the sports area) and of politicized relationships between the non- governmental sector and the local and central authorities. 13.6 With the broader definition of the CSOs, we are given a different picture of their activity areas (KLON, 2002). The largest group consists of organizations representing the interests of specific professional groups (mostly trade unions), at 18.2 percent, then come religious organizations (mainly churches and religious congregations), at 16.5 percent. The third group is composed of sports and recreation organizations, at 14.2 percent, and the fourth group is comprised of organizations of public security (over 12,000 of voluntary fire brigades). Social aid, self-aid and charity organizations are classified at the eleventh position and cover only 1.8 percent of all entities in the civil sector. 13.7 The economic potential, measured by the employment levels, of NGOs that operate in social aid and social services is much higher: those type of services constitute the third largest area (after culture/sports/recreation and education) in the operations of the non-profit sector in Poland, with 17 percent of the total employed in non-profit organizations. However, the overall financial situation of NGOs is poor, with 77 percent of the organizations complaining about 8In the Polish environment the broader definition of the non-governmental sector hardy reflects the scope of real civil activities. The civil character of a significant number of organizations is very limited or is not civil at all. Many of them do not execute such functions of NGOs as clarification of the public needs and interests, control over state authorities, aid to the people in need, integration or activation of social communities, etc. These organizations are, rather, manifestations of automatic membership (parents' committees,), employees' interests (trade unions), religious needs (churches) or fire protection (voluntary fire brigades). For these reasons, the analysis herein will always refer to non- governmental organizations in the narrow meaning of the term, unless otherwise indicated. 289 difficulties in obtaining financial resources, and the organizations are considerably diversified in terms of their material situation. Apart from a few well-off organizations, the majority of the organizations are small and have unstable economic bases. Over 43 percent of Polish NGOs have an annual budget below 10,000 PLN (2.5k US$); 80 percent have an annual budget up to 100,000 PLN, while 0.5 percent of organizations hold a yearly budget over 10 million PLN. At the same time, 83 percent of organizations lack any significant assets and only 16 percent have financial reserves. 13.8 Financial difficulties for CSOs are the most frequent reason for ceasing operations (in the opinion of 27 percent of organizations that ceased operations). In the 1990s the Polish non- governmental sector was supplied with funds by domestic private sources (individual charity and corporate donations) and by the private incomes of the organizations. According to data on the income structure of CSOs from the Chief Statistical Office (GUS) for 2001, public funds constitute approximately 19 percent of NGOs' incomes, which is a 5 percent decrease in relation to public outlays in 1997 when these funds made up 24 percent of the income of Polish non-profit organizations. Contributions from private individuals constitute approximately 10.3 percent of the sector's income ­ and private generosity has been growing in recent years ­ while donations from corporate bodies constitute 6.2 percent. Foreign funds, which have decreased substantially since 1997, constituted 8.3 percent of income in the non-governmental sector in 2002. B. CIVIL ACTIVITY DEVELOPMENT 13.9 In 1989 and afterward, Poland created the legal grounds for establishing CSOs, which were guaranteed both in the Constitution and in the laws on associations and foundations and other legal regulatory acts (Le, 2001). In comparison with the communist period when there were only several thousand social organizations, the civil society in Poland today is sizable in terms of the number of entities, and diversified in terms of their legal status and economic potential. During the transformation period, CSOs operating in the area of non-public education and the rehabilitation of the disabled were equalized with local authorities with regard to access to public funds and operational possibilities. This was a success and was the end result of their persistent efforts in this area. Owing to the law on Public Benefit Organizations and Volunteerism, the beginning of this decade should be marked by a full establishment of civil initiatives. CSOs are not particularly perceived as a driver of political, economic and social reforms by Polish politicians, and so their role is limited solely to the provision of social services. 13.10 Despite their dynamic development at the beginning of the 1990s, Polish civil organizations are generally rather weak and closed off, or of an enclave-like nature (Gliski, Palska, 1996). There is a sharp imbalance between the civil sector and the massive business and government sector. Social scientists speak openly about the weakness of Polish civil society, in particular when confronted with the strong, informal and privileged groups dominating the political scene (Mokrzycki, 2001). The result has been the ensuing crisis of Polish democracy, which still follows the model of democracy based on majority and on procedures, rather than on participation in public affairs and on respect for the interests of minority groups. 13.11 In this situation civil organizations' ability to exercise control over state authorities, and their participation in the process of political decision-making at any level, are weak. Financial weakness holds these organizations back from the independent execution of many important social projects, while state assistance is very limited, despite the constitutional rule of subsidiarity. The state's involvement in the development of the civil sector weakened after 1989, and the sector progressed mainly because of self-development, self-defense, and self-education (mainly cultural) mechanisms, as well as foreign aid (financial, expert, cultural and political). 290 Domestic political, business, cultural and media elites showed little interest and offered limited support to this development (Gliski, Lewenstein, Siciski, 2002; Fraczak, 2002; Gliski, 2000). 13.12 Despite the challenging environment and the small size of the non-governmental sector, particularly as compared to the enormous scale of social needs, it should be stressed that NGOs undertake various social tasks, including many poverty prevention measures. They are often the sole institutions able to resolve these types of problems, and they do it in a more professional and less costly manner than the first and second sectors' institutions. They constitute real enclaves of civil society in Poland. 13.13 Currently, the non-governmental sector, after its dynamic development in the mid-1990s, is experiencing a participation decline: in 1990, 5 percent of Poles belonged to NGOs (Gliski, 2000), in 1995, 13 percent, in 1999, 9 percent and in 2002, 10 percent (and only 6 percent declared their active participation in organizations and associations) (IFiS PAN, 2002).9 Membership in civil organizations is considered important by 48.3 percent of Poles; over 50 percent say that there is no organization that could support their interests, and over 40 percent think that if such an organization was established they would be willing to join it (IFiS PAN, 2002). These results point to the significant but neglected civil potential of Polish society. 13.14 The main obstacle to this potential and the reason for the crisis in the civil sector in Poland is the combination of numerous unfavorable phenomena that hamper the development of civil society. Some of these impediments have already been mentioned. One is the insufficient number of institutional and legal conditions facilitating the development of democracy based on participation, such as: · Lack of the practice of partnership and social dialogue, common in the EU (including not only associations of employers and trade unions, as well as the government, but also CSOs' representatives on an equal basis in the mechanisms of social dialogue) · The small number of provisions regulating participation in the Polish law · Lack of partner-type rules of intersectoral cooperation · Poor execution or lack of just and clear procedures on the access of civil organizations to public funds · Lack of a tradition of using social participation techniques in governance procedures on all authority levels · Excessive impact of politics on public life in Poland. 13.15 A worrying example of institutional obstacles to civil participation is the operation of local governments. Social research on the local government institutions shows that Polish local authorities, especially in poviats and voivodships, favor the model of "self-governance without participation." Therefore, programs for cooperation between CSOs and local governments are scarce. The first such program, still operating with success, was created in Gdynia in 1995. However, after a short and dynamic period of development of these types of programs in many cities,10 local governments began to withdraw from clear cooperation procedures with NGOs. Currently, following the estimates of the Polish Cities Association, programs of intersectoral 9 Membership in the NGOs in the broader meaning of the definition showed the following results: 1990, 27 percent ("Poles 1990" survey); 1995, 28.6 percent; 1999, 21 percent; 2002, 24.8 percent (IFiS PAN, 2002). It should be noted that membership in trade unions over the same years decreased dramatically: 22 percent; 12.5 percent; 7.5 percent and 6.2 percent respectively (and 1.6 percent of Poles declared their active participation). 10It should be noted that almost all positive examples of participation models were the result of initiatives taken by civil organizations and their pressures. 291 cooperation exist in only 50 cities (from about 700). Legal regulations that enforce some kind of transparency and public participation in the local governments' operations (law on public finances, law on the access to public information, laws on local governments) are only formally complied with. For example, on various levels of local administration (as well as central administration11) there are no bodies that would involve representatives of NGOs or independent observers working on the appropriation of funds or as advisory units. 13.16 Another, previously mentioned, impediment to the development of civil activity is the not always friendly attitude of the majority of political, cultural, media or business elites to the non- governmental sector. The sector is so weak because for years it has been unable to win modern, democratic legal regulations. An obstacle to the development of civil activity is also posed by the lack of pro-civil awareness of Polish society. This is rather the legacy of the communist era, which discredited the ethos of social work and civic activity, and other values of the intelligentsia, although the transformation period is also conducive to anti-civil behaviors. And, as has been mentioned, the bad financial situation is another serious obstacle for NGOs. 13.17 Finally, the non-governmental sector is challenged by its own weaknesses and drawbacks. The overly privileged position of post-communist and patronage-based organizations, as well as the underestimated significance of social control organizations, to a large extent result from external factors. However, the sector also faces many internal issues, such as: poor joint representation; insufficient compliance with internal regulations and ethics; low organization culture (for example, 80 percent of organizations do not acquire new volunteers on a regular basis or do not launch programs of cooperation with volunteers, only 4 percent provide insurance coverage for volunteers, only 3 percent provide them with the necessary medical check-ups, and only 15 percent recompense volunteers with the costs they incur as a result of their work). Another internal problem is the oligarchic character of the sector, especially of the highly professional infrastructure organizations or the politically dependent organizations. Civil Organizations for the Prevention of Poverty and Social Exclusion:12 Number of Organizations, Scope of Activities and Structure 13.18 In the survey conducted by KLON/JAWOR, 17.2 percent of NGOs (about 7,000) said that social aid, self-aid and charity work belonged to the three areas of their main scope of activities. Only 4.3 percent (about 1,800) of CSOs stated that these areas were the most important field of their operations. Therefore, these organizations deal with the widely understood issue of poverty and social exclusion. Although organizations in this area are infrequent, they are not on the outskirts of the third sector. The number of organizations operating in the field is much higher when we also take into account activities aimed at the stimulation of social and professional groups endangered by unemployment, life-long training and labor programs. 13.19 The fact that relatively few organizations are focused on poverty and social exclusion may result from: 11An exception is the Contact Group working at the Ministry of Economy, or the Environment Council at the National Council for European Integration in the European Integration Committee. 12Statistical data on the NGOs operating in the area of poverty and social exclusion prevention (presented in the tables in this chapter) were gathered during the KLON/JAWOR research of 2002 and worked on for the purposes of this document. Aggregated results of KLON, 2002 research concerning the entire third sector were published in Dabrowska, Gumkowska, 2002; and Dabrowska, Gumkowska and Wygnaski, 2002. 292 · The social belief that poverty-related issues should be resolved by the state and that the central administration should handle these problems · The negative perception of the poor in Poland · The (already discussed), lack of central and local support for these issues and the overall poor situation of the CSOs in Poland (lack of regulations, lack of uniform policy, opposition of the elites, etc.). 13.20 Despite the small number of NGOs that deal with poverty and exclusion, it should be stressed that these organizations ­ thanks to their professionalism and the dedication of thousands of activists and volunteers ­ carry out important work that is often neither visible nor appreciated. Their activities are frequently very efficient and low-cost, as they understand better than other institutions both the social needs and the local environment, and they enjoy the higher confidence of aid recipients. Furthermore, NGOs tend to focus on a problem, rather than on material compensation. 13.21 Among the organizations operating in poverty prevention, 9.4 percent have a diversified structure and are "parent bodies" for other units with local outlets, 19.4 percent are parts of more complex structures, 3.2 percent operate on the basis of the legal personality of their parent units, while 23.9 percent have a separate legal personality with the ensuing higher independence. 13.22 Organizations dealing with poverty are focused on a diversified range of activities and the majority of them operate on a wider scale than the closest vicinity (district, housing estate). Only 8.1 percent of all organizations concentrate their operations on their most direct local problems. Others operate in the municipality or the poviat (38.4 percent), in voivodships (28.1 percent), or countrywide (25.3 percent). 13.23 Staff is the key component of the organizations for poverty prevention (see Table 13.1). Less than half of them (46.5 percent) do not provide their people with employment contracts and are based only on voluntary work. Over one-third (34.6 percent) of organizations employ from one to five people, and 18.9 percent employ over five people. The number of staff is estimated at 32,100 (about 26,500 full time employment contracts [FTEs]). Compared to the entire third sector, organizations operating in the area of social aid more frequently conclude formal agreements with employees and hire a larger number of full time staff. 293 Table 13.1: NGOs Operating in the Area of Social Aid with FTEs Population in the Third Social Aid No. Employment Sector Organizations by Organizations Percentage 1 None 54.9 46.5 2 1-5 people 32.3 34.6 3 Over 5 people 12.8 18.9 4 Total 100.0 100.0 Source: Foundation KLON/JAWOR. 13.24 In comparison with other organizations, social aid units have a larger number of working volunteers who are not members of the organization. Volunteers work in 61.5 percent of social aid organizations, with 47 percent working in all CSOs. It should be noted that volunteers working for the poor are often at the same time beneficiaries and members of the organization (for example, in Markot, "Chleb i ycie" [Bread and Life]). Most community activities are performed by people who live in the organizational centers. As these people were not included in the KLON/JAWOR research, we may assume that the participation of volunteers in social aid organizations is much greater than is indicated by the research results. With regard to social and professional groups, these volunteers are usually unemployed and disabled or retired pensioners. Table 13.2 shows the social and professional status of volunteers in organizations, and Table 13.3 gives the numbers of volunteers. Table 13.2: Social and Professional Status of the Volunteers No. Social and Professional Status of Percentage of Organizations with Frequently the Volunteers Active Volunteers 3rd sector population Social Aid Organizations (percent) (percent) 1 Employed (at a different workplace) 60.6 48.3 2 Secondary school students 42.3 32.0 3 University students 39.7 38.3 4 Retired/ disabled pensioners 27.5 38.1 5 Unemployed ­ secondary school 17.2 15.7 graduates 6 Unemployed ­ university graduates 11.0 13.4 7 Unemployed ­ others 9.1 13.3 8 Professionally inactive (e.g., bringing 9.0 17.8 up children) Source: Foundation KLON/JAWOR. 294 Table 13.3: Volunteers in NGOs No. Number of Volunteers 3rd Sector Social Aid in NGOs Population Organizations (percent) (percent) 1 1-5 22 16.8 2 6-10 20 14.6 3 11-15 10 14.5 4 16-30 21 25.8 5 31-60 11 9.0 6 61-100 8 10.1 7 Over 100 7 9.2 Source: Foundation KLON/JAWOR. 13.25 The data show that CSOs for poverty prevention constitute a distinct and developing subsystem of institutionalized services which, by operating for the sake of the excluded social groups, becomes a kind of employer. Although the focus on the issues of poverty and social exclusion in the third sector is limited, the organizations operating in this area are more developed, with a broader structure and a larger number of active volunteers than those in the rest of the sector. Funding 13.26 Where funding is concerned, 76 percent of organizations for the prevention of poverty and social exclusion signal problems with obtaining financial resources and goods. This points to their financial instability and to the element of uncertainty embedded in their daily operations. Financial instability leads to a series of negative consequences that affect the units' efficiency. Organizations without a stable financial background are unable to develop and execute long-term strategies and to provide regular assistance. In 2002 the income of half of the social aid organizations did not exceed 25,897.5 PLN (the median), which is a significant amount if compared to the income of the remaining organizations in the third sector (the median for all NGOs did not exceed 19,000 PLN). To sum up, the income of social aid organizations is higher than the average in the sector. 13.27 Among the organizations operating in the poverty prevention area we encounter units that report an income of 1,000 PLN and units that declare an income of over 10 million PLN (Table 13.4). The budget of a large number of organizations (37.6 percent) ranges between 10,000 and 1 million PLN. Organizations against poverty have a slightly higher income than the total of third sector organizations, which (under Polish conditions) does not result from their special strength but rather from their greater opportunity to obtain funds from local governments (which perceive social purposes as more important than ecological or cultural ones) and from the interest of citizens who show greater generosity for charitable purposes.13 This thesis is supported by the research results presented in Table 13.5, where we can see that, in comparison with other CSOs, social aid organizations are more frequently supported by local authorities (as much as 31 percent of resources compared to 19.6 percent in the entire sector) and common citizens (as much as 20.4 percent of their budgets is a result of contributions from physical persons while in the whole 13In 2002 as much as 84 percent Poles claimed their participation in at least one charity event and 32 percent declared that they have offered financial support to a specific family or individual in need (CBOS, 2000). 295 sector this share amounts to 10.3 percent). It should be noted that the share of central government funds in social charity organizations is much smaller, constituting only 6.4 percent of social aid organizations' resources (with the average for the whole sector at 13.5 percent). Table 13.4: Income of NGOs Operating in the Area of Social Aid No. Income Spread 3rd Sector Social Aid (PLN) Population Organizations (percent) (percent) 1. 0-1 thousand. 15.3 11.7 2. 1.-10 thousand 26.6 25.1 3. 10-100 thousand 35.6 37.6 4. 100 thousand.-1m 17.5 19.4 5. 1m-10m 4.5 5.3 6. over 10m 0.4 0.9 7. Total 100.0 100.0 Source: Foundation KLON/JAWOR. 13.28 The sources of income of social aid organizations are diversified. It is noteworthy that almost one-third of the income of CSOs comes from public funds (central and local government funds), while the rest comes from contributions and donations. In practice, these organizations use any available methods to raise funds, as presented in Column A of Table 13.5. In relation to the total third sector population, social aid organizations are more often based on traditional financial sources such as contributions from physical persons and corporate bodies, organized campaigns, public fund-raising and other charity actions. Owing to the material situation of the beneficiaries, the CSOs services rarely generate income from fees. 13.29 The table shows how important particular sources are in the budget of an average organization and in the total budget of social aid organizations. Column B depicts the budget structure of an average social aid organization (that is, it presents the average share of individual sources in the income of the organization) and verifies these data against the budget structure of other associations and foundations. The structure of the budget of an average organization calculated here is closer to the budgets of numerous rather "poor" organizations than to the budgets of the smaller group of better-off organizations. 13.30 Column C shows the share of individual sources in the total budget of social aid organizations and the total resources of the whole sector. The difference from Column B consists in the fact that, although a given source may substantially feed the budgets of numerous organizations, the money from this source may not have a significant share in the whole sector (for example, membership fees are a common source of income for many organizations but their share in the total amount is small). This is true not only for membership fees, but also for income from campaigns, public fund-raising and charity events. On the other hand, there are rarely used financial sources which make a large contribution to the entire budget, such as income from business operations, which is allowed, although it is not a direct statutory activity. 296 Table 13.5: Sources of Income for NGOs Symbol A B C Social aid All organizations Income from Income from Share of Share of organizations using the source the source in the source in individual individual sources Financial sources using the (%) social aid the budgets of sources in the in the budget of source organizations' all budget of the entire sector (%) budgets organizations social aid (%) (%) (%) organizations (%) 1 Public sources ­ central government 22.1 17 6.3 6 6.4 13.5 2 Public sources ­ local government 44.9 48.8 21.5 20.2 31.0 19.6 3 Public sources ­ foreign aid programs (including EU funds, e.g. Phare, Sapard, Access) 4.7 3.4 2.5 1.1 2.5 5.7 4 Support from other domestic NGOs 11.0 10.9 3.9 2.5 1.7 2.6 5 Support from other foreign NGOs 6.4 3.2 1.9 1.1 6.1 2.8 6 Contributions from physical persons (which are not fees for services) 47.2 38.8 9.4 10.5 20.4 10.3 7 Contributions from corporate bodies (which are not fees for services, e.g., for sponsor agreements) 41.8 40 13.7 10.3 6.3 6.2 8 Donations from other units of the organization 2.4 5.7 0.7 1.7 0.1 1.8 9 Membership fees 66.8 69.7 19.7 26.8 3.3 8.3 10 Income from campaigns, fund-raising, charity events 18.9 10.7 2.8 1.7 0.6 0.9 11 Bank interests, capital profits, shares, income from assets, e.g., rental of premises, etc. 24.5 21,2 2.4 1.6 1.1 6.2 12 Fees for statutory activities execution, (including donations in the form of services payments) 11.9 24.1 4.4 6.5 2.3 7.2 13 Income form business operations which is not a direct statutory activity 6.4 6.4 2.7 4.2 10.2 10.4 14 Other 16.6 10.2 7.7 4.2 8.0 4.7 Source: Foundation KLON/JAWOR. 297 13.31 According to qualitative research on the issues of poverty and social exclusion, organizations do not use some sources of income because of certain rules that are generally followed. For example, they do not accept donations from political parties or politically active individuals (a finding supported by data on contacts and cooperation: 72.5 percent of organizations declare having no relations with political parties and organizations). Neither do they accept contributions from individuals and institutions whose actions do not comply with certain defined standards (e.g., from tobacco or spirits producers, especially if the organization deals with poverty among children and youth). Organizations acting for the benefit of the poor, which base their actions on social trust and solidarity, follow the ethical rules of the non- governmental sector14 and are guided by the principles of economy, resourcefulness and transparency, in particular in the management of their financial resources. Cooperation with Other Entities 13.32 The relations that organizations against poverty have established with other entities and institutions are limited mainly to those in their nearest vicinity and to those that can grant them the necessary support. In other words, they have regular contacts with the business environment (16.4 percent) more often than with other NGOs (14.9 percent) and have contacts more rarely with public institutions such as schools and museums (38.6 percent versus 49 percent for the other NGOs) (see Table 13.6). These contacts to some extent result from the specific nature of the undertaken activities and the existing issues and are, in a sense, obligatory (mainly applying to local authorities). Regular and frequent contacts with municipal authorities are recorded by 51.2 percent of organizations, and 48.5 percent of organizations report frequent contacts with poviat units. Interactions with other local units such as social aid centers (OPS) or poviat centers for family assistance (PCPRs) (natural allies in poverty prevention) are also frequent. The frequency of contacts with the municipal office is similar to that of other third sector organizations (although the financial outcome of these interactions is more significant). Over one-fifth of social organizations (22.3 percent) are in regular contact with voivodship authorities: in comparison with all NGOs this is a less frequent contact. The general trend is: the higher the administration level is, the less frequent the contacts are. This means that CSOs in the area of social aid cooperate mainly with the local structures of the widely understood political system, with state administration units (on the voivodship level rather than on the central level), and with the business community. 13.33 In local communities the interests of the poor and the socially marginalized are mainly represented via central and local administration bodies and businesses, excluding political parties. Compared to other CSOs, social aid organizations have much more regular and sustainable relations with the Catholic Church and church-related institutions (22.5 percent versus 12.7 percent for total NGOs). This is to some extent due to tradition and to some extent to the fact that the Church undertakes similar actions to fight poverty. 13.34 Although, social aid organizations cooperate with other NGOs more willingly than with other social units, these contacts are still rather sporadic (21.6 percent declare having no contacts with other NGOs, while the figure for the total population is 30.4 percent). Among social aid organizations, 27.6 percent belong to domestic, regional or professional NGOs' agreements (formally or informally) and 6.8 percent to foreign/international agreements. This is slightly less than the total of NGOs. 14The charter of ethical values in the non-governmental sector was adopted in the second half of the 1990s. 298 Table 13.6. Frequency of Contacts with Other Entities 3rd Sector Social Aid Population Organizations (%) (%) No. Institution Type Frequent Frequent Regular Sporadic No Regular Sporadic No Contacts Contacts Contacts Contacts Contacts Contacts 1. Municipal authorities 52.2 29.0 18.8 51.2 30.9 17.9 2. Public institutions (schools, 49.0 26.1 24.9 38.6 35.5 25.9 museums) 3. Local authorities and public 45.9 35.7 18.5 48.5 37.7 13.7 institutions at poviat level 4. Local media 38.0 41.7 20.3 31.5 50.2 18.4 5. Other Polish NGOs 30.4 39.2 30.4 33.0 45.4 21.6 6. Academic and scientific circles 20.4 28.1 51.5 19.1 20.5 60.3 7. Expert magazines 18.1 32.0 49.9 15.9 31.1 52.9 8 Administration at the voivodship 16.1 39.0 44.8 22.3 39.5 38.1 level 9. Business 14.9 41.5 43.6 16.4 41.7 42.0 10. Church, religious congregations 12.7 21.7 65.6 22.5 23.9 53.5 11. Governments, central 10.7 21.9 67.4 10.5 22.9 66.5 administration institutions 12. Nationwide media 10.7 29.4 59.9 9.8 32.1 58.0 13. Foreign NGOs 8.9 21.5 69.6 8.6 19.9 71.5 14. Political parties and organizations 6.1 17.9 76.0 5.7 21.8 72.5 15. Institutions responsible for EU 4.8 15.0 80.1 2.9 19.4 77.6 integration Source: Foundation KLON/JAWOR. 13.35 In assessing their cooperation with various institutions, we find that organizations against poverty seem rather reserved, though we can generally say that they are satisfied: the number of positive evaluations exceeds the number of negative and indifferent ones (see Table 13.7). Where the organizations appear to find these interactions most difficult, we would point to the business environment (12.6 percent) and central administration institutions (10.9 percent). 299 Table 13.7: Assessment of Cooperation with Various Institutions 3rd Sector Social Aid Population Organizations (%) (%) No. Cooperation Assessment Neither Neither Generally good or Generally Generally good or Generally Good bad bad good bad bad 1. Public institutions (schools, 84.7 12 3.2 73.1 23.2 3.8 museums) 2. Foreign NGOs 81.9 15.3 2.8 78.6 16.0 5.3 3. Other Polish NGOs 79.7 16.4 3.8 74.0 21.3 4.7 4. Academic and scientific 79.4 16.1 4.5 69.0 27.3 3.6 circles 5. Local media 79 17.7 3.2 72.7 24.6 2.7 6. Church, religious 76 20.4 3.6 74.4 19.6 6.0 congregations 7. Expert magazines 75.5 20.7 3.8 63.9 33.1 3.0 8. Nation-wide media 72.4 21.8 5.8 66.8 25.8 7.4 9. Municipal authorities 67.4 24.5 8.1 60.9 31.2 7.9 10. Business 64.4 28.6 7 54.8 32.6 12.6 11. Local authorities and public 64.1 26.9 8.9 63.6 26.6 9.8 institutions on poviat level 12. Administration at the 57.9 31.6 10.4 48.1 42.9 9.0 voivodship level 13. Governments, central 57.3 33.9 8.8 51.1 38.1 10.9 administration institutions 14. Institutions responsible for 56.8 34.1 9.1 EU integration 38.5 53.6 7.8 15. Political parties and 54.8 38.3 6.8 54.3 34.2 11.5 organizations Source: Foundation KLON/JAWOR. 13.36 On the government level, NGOs dealing with poverty and social exclusion cooperate with the Ministry of Economy, Labor and Social Policy, which attempts to include NGOs in the resolution of specific poverty-related problems. In 2002, the Ministry launched two social projects directed to NGOs: "Homelessness" and "The Governmental Program for Supporting Municipalities in School Children's Nourishment."15. The aim of these competition-based programs was to activate social organizations in these matters. The Ministry offered mainly financial support, with the emphasis that formal bodies established to deal with such issues can be helped by the operations of NGOs. However, this declaration seems to be only a postulate. The same applies to the provision of the "Homelessness" program which says that the purpose of direct support of the Ministry provided to specific NGOs is to create sustainable grounds for social policy system in the area of social exclusion prevention. One-off donations to some organizations, with a lack of meaningful coordination of activities, cannot create any stable 15This program was addressed not only to NGOs but also to municipalities and individuals or groups involved in children's nourishment. 300 grounds for the development of social policy. In the program focused on school children's nourishment, the emphasis was placed on long-term cooperation manifested in signing the "Declaration on Cooperation in the Execution of the Governmental Program for Supporting Municipalities in School Children's Nourishment" in 2003. The signatories of this document from the government side, together with five representatives of NGOs, had announced their direct collaboration to enhance the efficiency of these programs. They plan to call up a working group where people and organizations involved in this issue can join their efforts. 13.37 In analyzing the cooperation between the central administration and NGOs in dealing with poverty prevention, attention should be paid to the establishment of the Task Team for Social Reintegration by the Ministry of Economy, Labor and Social Policy at the end of 2002. This was done during the work on the Social Integration Strategy aimed at preventing social exclusion. In this project, NGOs are one of the main (and equal) partners and consultants that are to create solutions to key social problems. Review of the Activities Undertaken by Organizations for Poverty and Social Exclusion Prevention 13.38 The KLON/JAWOR research shows that organizations for poverty prevention concentrate their efforts on direct support for the people in need (see Table 13.8). The forms of support include: direct provision of services to the people in need (68.8 percent of organizations); acting as representatives of the interests of beneficiaries (50 percent); granting financial support (35 percent). Organizations also undertake broader actions beyond their primary scope of focus, such as: mobilizing and educating the public to win support for their operations, leading social campaigns, etc. (42.8 percent); organizing debates, seminars and conferences on related topics (30.8 percent); participating in debates with public administration at various levels (e.g., organizing or participating in social consultations, campaigns, protests, petitions, etc. ­ 25.7 percent). The purpose of all these actions is to make public opinion more sensitive to the issue of poverty and social exclusion. 13.39 Comparing the activities undertaken by other NGOs with those of organizations focused on poverty and social exclusion prevention, one notices that the poverty organizations more often act as representatives of the people in need (see items 4,6,11 in Table 13.8). The beneficiaries rarely take part in the public discourse, and are frequently perceived as unaware of and deprived of their full social rights. 301 Table13.8: Forms of Actions Undertaken by Organizations 3rd Sector Social Aid Population Organizations (percent) (percent) No. Actions Important, Additional, Important, Additional, basic, assisting, basic, assisting, regular sporadic regular sporadic 1. Direct services to members, beneficiaries or 65 12.5 68.8 17.6 clients 2. Cooperation with other organizations / 41 36 34.5 42.5 institutions In Poland (joint actions, meetings, exchange of experience) 3. Mobilizing and educating the public to win 39 35 42.8 29.7 their support for organizations, social campaigns, etc. 4. Representing the interests of members, 38 28 50.0 29.4 beneficiaries / clients of the organization 5. Organization of debates, seminars, 28 30 30.8 33 conferences on topics important to the organization 6. Participating in debates with public 19 27 25.7 27.1 administration at various levels (e.g., organizing or participating in social consultations, campaigns, protests, petitions 7. International cooperation with institutions 18 24 14.1 28.7 and organizations of a similar profile 8. Publishing magazines, bulletins, reports on 16 28 20.5 25.2 subjects related to the mission of the organization 9. Financial or tangible support of individuals 12 26.5 35 36.3 10. Supporting other NGOs by providing 12 35 14.4 37.0 information, advice, consultations, training and other 11 Lobbing, influencing system changes, e.g., 11 18 14.5 20.1 amendments to legal regulations 12. Scientific research, collection and 10 20 11.2 16.8 processing of data 13. Financial support of projects launched by 5 14 2.8 14.5 other organizations or institutions In Poland Source: Foundation KLON/JAWOR. 13.40 The qualitative analysis of actions undertaken by organizations for poverty prevention shows that they operate on three levels: · Structural ­ by aiming at standards that could be widely applied; radical changes to political, economic and social conditions (e.g., by participation in the coordination of the support of legal solutions); consultation; the development of a problem resolution matrix. 302 · Indirect ­ by supporting actions whose recipients are not fully defined (e.g., purchase of equipment, actions for the development of the local environment, provision of school facilities). · Direct ­ by executing projects/programs focused on defined groups; and by taking specific actions that have immediate results. 13.41 The focus and operations of social aid organizations are also limited to some extent by the two areas of poverty (poverty among the old and poverty in rural areas being peripheral to the scope of their interests). Actions undertaken to fight poverty in rural and urban areas are completely different. In the case of urban areas, the organizations usually undertake direct and irregular actions (individual assistance, mainly in the form of medicines, accommodation, support in learning, care, compensation of health care costs). Fighting poverty in rural areas is a part of the comprehensive rural focus16 in which the poor are not distinctly identified. The overall purpose of these activating and modernizing actions is the improvement of the general situation in rural areas, including that of the poor. These types of actions (structural) are performed by organizations based in large cities -- the so-called infrastructural or "umbrella" organizations. 13.42 Actions addressed to the inhabitants of rural areas do not provide direct assistance but are aimed at stimulating their initiatives, which would consequently reduce poverty17 (e.g., by establishing a strawberry plantation, by organizing an art competition at the schools and buying necessary facilities from the money collected during the auction of the works, etc.). These types of activities are mainly initiated by small local organizations 13.43 Social aid organizations find it difficult to perceive their activities within a broad perspective and to work on complex strategies that would reduce specific social problems. They act on a "here and now" basis, rather than undertaking long-term and far-flung projects, and they are primarily focused on eliminating the symptoms, not the roots, of the problems. Beneficiaries 13.44 The way in which organizations for poverty and social exclusion prevention treat their beneficiaries is ambivalent in two respects. On the one hand, the assistance is not based on the dichotomy of those who deserve and those who do not deserve help. The assistance is not driven by pity or sympathy, but rather by the aspects of human dignity. On the other hand, organizations often perceive their beneficiaries as people who are helpless in life, who are unable to make rational decisions even though they have worked out the entire array of methods for survival and adaptation. If possible, assistance is provided to all people in need, regardless of the causes of their difficult situation. However, in spite of the equal opportunities declaration, organizations set specific selection criteria to ensure that the aid is directed to those who most require it. The assistance is both available and based on entitlement. It should be noted that none of the 16 The number of NGOs operating in rural areas is difficult to estimate, because the available statistics do not differentiate between rural and urban organizations. Data are unrepresentative because they are based on voluntary reporting and come from various databases. The Foundation for the Development of Polish Agriculture reported having 455 registered organizations on April 17, 2003. It should be noted that the browser in this database does not specify the category of organizations operating in the interest of the poor. The only available category is very broad: "social and cultural activities" (127 organizations) and probably all poverty prevention efforts are included therein. 17This conclusion is supported by the results of the third edition of the competition for NGOs, "Our way to fight poverty" which has been organized in the past four years by the Foundation for Rural Area Support. This year (the fourth edition) 598 applications were submitted. 303 organizations analyzed by K. Górniak (2002) is focused on the identification of people in need. Organizations develop new methods and strategies of operations, but they do not actually search for their beneficiaries. Street workers are the only exception to the rule, but they are not really common in Poland. We can therefore say that organizations wait for their beneficiaries but do nothing to identify them. Thus, the assistance is given to those who seek help (those who are sufficiently informed and who know how to fill in an application form, etc.), but not necessarily to those who need it most. 13.45 Qualitative research on the poverty areas shows that NGOs are absent from the lives of the respondents. Caritas is the only organization that systematically "knocks on their door" and that provides the majority of financial and material help. Respondents incidentally also mention the Polish Red Cross (PCK) (as a source of clothing), MONAR, and (only once) the Association for Assistance to Children with Down's Syndrome. The assistance from various sources (Caritas, PCK, school collections) is incomplete: some things are offered in excess while others are in scarcity. Beneficiaries receive mainly clothes. Paradoxically, the recipients rarely have a proper place or washing powder to be able to make proper use of all these goods. However, out of the fear of even greater impoverishment, they accept everything that they are given. 13.46 The Catholic Church is an important source of assistance for the poor. Unlike the municipality, the respondents treat the Church as a gift-giver, and not as a source of income to which they are entitled. Respondents use church assistance on an irregular basis, receiving mainly clothing, food and medicines. In smaller communities "the church comes to the people" and looks for families in need. In large cities the situation is reversed, "the people go to the church" when they learn about a gift-giving event. In several cases respondents also accepted help from religious orders, on the basis of a personal relationship rather than on the institutionalized contact. 13.47 Interestingly, the Church plays the role of gift-giver while on the other hand it involves a serious material burden (collecting the offertory). Assistance from the parish is based on specific rules of exchange. Poor people are usually isolated from the broader parish community; only a few families investigated under the research actively participate in the life of the parish by taking part in pilgrimages, processions and preparations for religious festivals. The bonds with the parish community are weak, just as all other social bonds are. Often, social isolation is strengthened by space isolation. The Church is financially supported by the believers and the respondents take this role very seriously. Sacraments (baptisms, weddings, funerals) are the most heavy financial burden for the respondents. Although, in principle, the respondents make an offering relevant to what they can afford, the parish imposes informal "rates" for its services, which are followed by parishioners (as this is an element of belonging to the parish community). If the priest postpones the fee payment or waives it entirely, people perceive it as a serious allowance. Other forms of offerings include the yearly distribution of Christmas wafers and priests' visits to parishioners after Christmas. In these cases the poor are also discharged of offerings, which is regarded as a specific gift from the church. 13.48 Despite their ambivalent attitudes (sometimes regret and disappointment, sometimes relief and security) respondents do not mention the moral or psychological discomfort they experience as a result of accepting Church aid (while they frequently complain about various unpleasant situations related to contacts with social aid organizations). It is easier for them to accept assistance from the Church than to apply for help to formal organizations. This may result from the fact that they benefit from the social aid on a regular basis and treat it as something to which they are fully entitled, while accepting the Church assistance involves belonging to a community and is subject to some specific rules of exchange. The conditions of this exchange are clearly defined: faith is required in return for help. 304 13.49 It is also worth analyzing what the perception is of the public life and political reality by poor families and how they participate in these matters. As a result of unemployment and poverty, even formerly active people tend to withdraw from social life. We observe the progressive self-isolation of these people and their growing resentment towards both local and central authorities which they blame for their difficult conditions. C. CONCLUSIONS AND RECOMMENDATIONS 13.50 The scale of the problem of social marginalization and its effective prevention is frequently beyond the capabilities of CSOs. The pace of development of these organizations is insufficient -- to a large extent because of the little support that is offered by state institutions. The crisis of the civil sector in Poland prevents organizations from having a deeper involvement in the resolution of social marginalization. In some regions, especially in rural areas, there is a complete lack of civil involvement. 13.51 Despite these unfavorable conditions for social activity, about 1,800 organizations and thousands of devoted volunteers work for the sake of the poor and socially marginalized. Temporary activities prevail, which fight the symptoms but do not treat the root causes of poverty and social marginalization. 13.52 If this situation is to be mitigated, the state should encourage the mobilization of the civil sector, for example, through launching a nationwide program against social exclusion. Such a program would underscore the role of NGOs, which are the only units that can execute some specific social and charity tasks. A recently passed law on public benefit organizations and volunteerism, as well as the draft of the law on social employment are signs of the government's support for the idea of introducing active employment opportunities for people threatened by poverty and unemployment. 13.53 Practical actions should make use of the existing experience of NGOs and of the programs and initiatives that have proved efficient to date (such as the local funds program launched by the Academy for Philanthropy Development in Poland, the Food Banks program, the "Older Brother ­ Older Sister" program, street workers' programs, etc.). Although numerous initiatives of that type are undertaken in Poland, they still cannot satisfy the enormous needs. Regarding the above-mentioned programs, the majority are well known and documented by NGOs: for example, in the Atlas of Model Local Initiatives, prepared by KLON/JAWOR (2002) in cooperation with the Academy for Philanthropy Development in Poland and in the analysis of model local social initiatives against poverty and social marginalization published by the Institute of Public Affairs (Hrynkiewicz, 2002). However, lack of both funds and political initiative prevents the massive implementation of these projects. 13.54 Some of the NGOs have formulated new approaches to poverty prevention and social exclusion. Special attention shall be drawn to those that activate the beneficiaries and make them independent. The Mutual Aid Foundation "Barka" ("Boat") should be recognized for its systemic support of the marginalized groups' integration process, which is based on the idea of mutual assistance and coordination with the local community. The Barka Foundation stresses the importance of education, the creation of new jobs, and the development of inexpensive housing. NGOs also become involved in the management of professional organizations, such as the agricultural organization of professional activities program run by the Work for the Blind Foundation. 305 13.55 Examples of good practice are also disseminated by: · The Local Activities Centers (Centra Aktywnoci Lokalnej - CAL) which work towards enhancement and development of the local communities' potential. The CAL program's goal is to integrate local communities and civil groups, to strengthen local solidarity and the ties of mutual help, and to form groups of professional social animators. The CAL program proved that existing local institutions and organizations, both public and social, are able to act more broadly, inspiring and supporting citizens' organizations. The CAL method effectively supplemented activities of the local organizations (the public organizations and NGOs) activating the hidden potential of institutions that often lack resources and feel lost in the new reality. · The Small School Program, coordinated by the Federation of Educational Initiatives, had similar objectives. The program was initiated as a result of protests against the closing of schools in rural areas during the introduction of the education system reform in 1999. The program's mission was to assist and rescue schools in rural areas by providing them with a new institutional status: that of small schools, like the public schools, managed, however, by local associations. The introduction of such a program proves that the vigor of protests may be transformed into constructive activities, thanks to the cooperation of all participants involved in the problem resolution. Schools organized by local associations form new ties among schools, teachers, local communities and public authorities. 13.56 The activities of local associations uniting the inhabitants of a village who participate in the Small School Program are not limited to the management of an educational unit. They usually undertake various initiatives towards the general development of the village and become its spiritus movens. The scale of the associations' activities keeps growing. Associations participating in the Small School Program can provide examples of good practice for social organizations that aim at assistance and social dialogue, as the basis for an effective functioning of NGOs. 13.57 NGOs, and particularly those that operate in the field of social assistance, undoubtedly use available financial, material and organizational resources in a very effective way. This is the enormous appeal of the third sector. The effectiveness of social organizations is paradoxically an outcome of their weak financial basis, as well as their ingeniousness and skill. The authors of a report (Fatyga, Zieliski, 2002) in which projects sponsored by the Program Helpful Local Community, funded by the Batory Foundation, are evaluated, clearly state that most social organizations, despite encountering numerous difficulties, are able to evaluate realistically their financial needs, and to forge improvement strategies, as well as to carefully manage their available financial resources. This conclusion finds confirmation in various analyses and quality evaluations. 13.58 The following are some examples: · Tomasz Polkowski, leader of "Our Home" Society, says (in an interview by K. Górniak) that supporting one child in a state childcare institution reaches a monthly cost of 1,600 PLN or even as much as 3,000 PLN, while in a substitute family or another informal institution the costs do not exceed 800 PLN. In this case, the financial effectiveness is accompanied by extraordinary social effectiveness. The Society attempts to reform the children's support system by limiting typical charity activities and concentrating on programs that reform existing childcare institutions and develop in-home childcare, as 306 well as transform traditional childcare institutions into networks of small apartments, which make it possible to conduct programs that better prepare young people for an independent and valuable life. The Society's important task is to prepare young people for adult life through equipping them with the social skills and abilities that are required in a fully active social life. These are activities that will counter a deepening of the inequalities that lead to the exclusion of those who are not well adapted. · In 2002 the Warsaw "Food Bank" Foundation, having at its disposal a budget of 380,000 PLN, managed to provide food to the total value of 4.2 million PLN to organizations dealing with poverty and social exclusion. It can be said that each zloty spent by the organization accumulated the value of 11 zlotys. 13.59 The above-mentioned examples of valuable initiatives in the third sector are disseminated through networking. The networking method is based on the assumption that the dissemination of an initiative should rely on the existing infrastructure -- for example, on schools in rural areas, not necessarily on NGOs. The priority of the programs should be to identify and train young leaders from local communities who are willing to work in the local environment, to shape it, and to engage in actions against poverty and social marginalization -- for example, by means of bonds between neighbors, which are fairly strong in Poland. It must be borne in mind that no program for poverty and marginalization prevention will be successful as long as we fail to deal with two main root causes of these problems: namely, alcoholism and unemployment. 13.60 The NGO sector should cooperate closely with central administration institutions, local authorities, local institutions, and businesses, as well as with non-formal social institutions, such as the neighbors' assistance network. It should be borne in mind that the essential elements of the third sector are non-formal or spontaneous initiatives that significantly help to resolve problems. This type of initiative addresses local needs and thus is more effective. 307 REFERENCES CBOS (2000), Dobroczynno ­ dzielo pomocy potrzebujcym, (Charity ­The Art Of Giving to the Needy) Research report, no 2298, Warsaw. Dbrowska, J., and M. Gumkowska (2002), Wolontariat i filantropia w Polsce ­ raport z bada 2002, (Voluntary Work and Philanthropy in Poland ­ Research Report 2002) in P Warsaw: KLON/JAWOR Association. Dbrowska, J., M. Gumkowska, and J. Wygnaski (2002), Podstawowe fakty o organizacjach pozarzdowych ­ raport z badania 2002, (Basic Facts about NGOs ­ Research Report 2002) Warsaw: KLON/JAWOR Association. Fatyga, B., and P. Zieliñski, Ewaluacja projektów dotowanych przez program Pomocna Spoleczno Lokalna Fundacji im. S. Batorego. Raport z badañ, [Evaluation of the Projects Subsidized by the Local Community Program of the Batory Foundation, Research Report, Warsaw, 2001]. Frczak, P. (2002), Trzeci sektor w III Rzeczypospolitej (Third Sector in the III Poland's Republic), Warsaw. Frczak, P., and R. Skrzypiec (eds.) (2002), Przejrzysta gmina, organizacje pozarzdowe, korupcja (Transparent Local Authorities, NGOs and corruption), Biblioteka Aktywnoci Obywatelskiej, Vol. 1, Warsaw: Centre for Research of the Local Activities at ASOCJACJE Association. Gciarz, B., and W. Paków (2003), Procesy instytucjonalizacji samorzdu lokalnego i regionalnego. Aktorzy, efekty i beneficjanci (The Insitutionalization Process of Local and Regional Governments. Actors, results and beneficiaries), a report presented at the session "Self-organization of the Polish Society: The Third Sector in the United Europe," Warsaw: IFiS PAN, April 15. Gliski, P., 2000, "O pewnych aspektach obywatelskoci. Aktywno spoleczna a integracja wspólnot obywatelskich," (About Aspects of Civil Society. Social Activity and Integration of Civil Organizations), (with) H. Domaski and others (eds.), Jak yj Polacy (How the Poles Live), Warsaw: Published by IFiS PAN. Gliski, P., and H. Palska (1996), "Cztery wymiary aktywnoci obywatelskiej," ("Four Dimensions of Civil Engagement" (with) H. Domaski and A. Rychard (eds.), Elementy nowego ladu, (Elements of New Deal) Warsaw: Published by IFiS PAN. Gliski, P., B. Lewenstein, and A. Siciski (eds.) (2002), Samoorganizacja spoleczestwa polskiego: Trzeci sector (Self-organization of the Polish Society: Third Sector), Warsaw: Published by IFiS PAN. Górniak, K., (2002), "Ekskluzja spoleczna a nowe formy dobroczynnoci. Analiza dziala wybranych organizacji spolecznych," ("Social Exclusion and New Forms of Charity. Analysis of Several NGOs), wyniki bada przeprowadzonych w ramach przygotowywanej rozprawy doktorskiej (unpublished results of doctoral research). 308 Hrynkiewicz, J. (2002), Przeciw ubóstwu i bezrobociu: lokalne inicjatywy obywatelskie (Against Poverty and Unemployment: Local Social Initiatives), Warsaw: Institute of Public Affairs. IFiS PAN (2002), Europejski Sonda Spoleczny (European Social Survey), research results. ISP PAN (2001), Sektor non-profit. Nowe dane i nowe spojrzenie na spoleczestwo obywatelskie w Polsce (Non-Profit Sector. New data and new outlook on the Polish Civil Society), Warsaw: Institute for Political Studies PAN. KLON (2002), Podstawowe fakty o organizacjach pozarzdowych ­ raport z badania 2002 (Basic Facts about NGOs ­ Research Report 2002), Warsaw: KLON/JAWOR Association. Le, E (2001), Od filantropii do pomocniczoci (From Philanthropy to Assistance), Warsaw. Le, E., and S. Nalcz (2002), Midzynarodowe badania sektora non-profit w Polsce (International Research Results on the Non-Profit Sector in Poland), Non-Profit Organizations Research Unit PAN, Warsaw. Mokrzycki, E. (2001), Bilans niesentymentalny (Non-Sentimental Balance), Warsaw: IFiS PAN. Skrzypiec, R. (2002), Lokalne uczestnictwo obywatelskie (Local Civil Engagement), Warsaw: Centre for Research on Local Activities at the ASOCJACJE Association. Starosta, P., M. Szczepaski, and J. Kurczewski (2003), Badania spolecznoci lokalnych, wypowiedzi na Konwersatorium "Samoorganizacja spoleczestwa polskiego: Trzeci sektor w jednoczcej si Europie," (Local Communities' Survey, presentations from the Conference: ,,Self-Organization of the Polish Society: Third Sector in Uniting Europe), Warsaw: IFiS PAN. Stepie, B. (2003), Polityka rzdu wobec organizacji pozarzdowych w Polsce, referat na konferencji "Rola organizacji pozarzdowych w budowaniu spoleczestwa obywatelskiego," (Government Policies towards NGOs in Poland,, presentation at tej conference: The Role of NGOs in Civil Society Building". Non-governmental Organizations Center in Gliwice, April 5. Wybrane krajowe fundusze publiczne dostpne dla organizacji pozarzdowych w Polsce w latach 1998-1999. (Chosen National Public Funds Available to NGOs in Poland in 1998-1999). Report prepared by the Network for NGOs Support SPLOT, NGOs Data Bank KLON/JAWOR and Association for FIP, Warsaw, 1999. 309 14. SOCIAL BENEFITS AS ALTERNATIVE TO WAGE INCOME IN POLAND Agnieszka Chlon-Dominczak, Edwarda Dabrowska and Piotr Stronkowski 14.1 The aim of this chapter is to analyze the sources of income of the unemployed in Poland. Unemployed individuals usually remain without jobs for more than one year. Unemployment benefits are traditionally perceived as the main source of income for the jobless. However, as unemployment benefits in Poland are granted only for a short period of time, only a small and decreasing number of the unemployed are entitled to such benefits. In 2002, on average only 18.3 percent of the unemployed received this form of support. Therefore, this chapter also discusses other systems of social protection in Poland, such as social assistance and social security. The potential benefits that the unemployed can receive are of three main kinds: (i) benefits for the unemployed; (ii) social assistance for the poorest; and (iii) disability pensions. 14.2 The first type of benefits is benefits available to the unemployed based on the provisions of the law on employment and unemployment prevention. These benefits are described in the first section, and include: (i) unemployment benefits during the first 6 to 18 months of unemployment; (ii) pre-retirement allowances; and (iii) pre-retirement benefits. 14.3 The latter two benefits apply to unemployed individuals who meet specified age and work experience criteria until they are eligible for old-age pensions. 14.4 The second section of this chapter discuses social assistance benefits for those unemployed who meet the income criteria necessary to receive these benefits. The benefits covered include: temporary social assistance benefits; family benefits; and housing allowances. 14.5 The third section discusses benefits available from the social security system, in particular, disability pensions. Though pensions are usually not treated as support for the unemployed, in the early 1990s the pension system played a significant role in absorbing people who had dropped out from the labor force. At the end of the 1990s the criteria for receiving disability pensions were tightened. Nevertheless, the question of whether the social security system is more attractive for the unemployed remains. All listed transfers are available to people of working age. 14.6 The fourth section of the chapter presents a set of micro-simulations showing how social transfers can substitute for wage income. The analysis aims to show whether, and in what ways, the system of social transfers creates incentives for people to remain unemployed. 14.7 The fifth section analyzes the actual income of the unemployed on the basis of Labor Force Survey results, and the final section provides conclusions to the discussion. 310 A. BENEFITS FOR THE UNEMPLOYED 14.8 Poland offers two types of benefits for the unemployed paid from the Labor Fund (see section under Financing of Unemployment Policies, below). The first type consists of standard unemployment benefits that can be paid to all registered unemployed. The second type consists of pre-retirement allowances and pre-retirement benefits. These benefits are granted to unemployed individuals who are relatively close to retirement. They have replaced the early retirement pensions granted until 1997. Pre-retirement allowances were granted until the end of 2001, while pre-retirement benefits are still being granted to the unemployed. The level of pre-retirement benefits is linked to the level of the accrued old-age pension, while pre-retirement allowances were linked to the level of unemployment benefits. The former are granted to persons closer to fulfilling the retirement criteria. The institutional setting for unemployed programs is given in Box 14.1. Box 14.1: Institutional Setting for Unemployment Programs The system of institutions responsible for implementing labor market policies in Poland is decentralized. The legal regulations for labor market operations are the responsibility of the Minister of Labor. The Minister manages the Labor Fund, which is a source of funding for passive and active labor market policies. The administrative structure in Poland consists of three levels: regional, poviat and authorities. Regional authorities are mainly responsible for the preparation of regional strategies for labor market development and for the initiation and organization of some activities in the area of active labor programs. They also prepare statistical analyses of regional labor markets. Poviat authorities manage poviat labor offices. The offices register the unemployed, grant them the right to unemployment or pre-retirement benefits, and pay these benefits. The labor offices also conduct active labor market programs. District authorities are not directly involved in the labor market policy and institutions. Financing of Unemployment Policies 14.9 Labor market policies are financed from the Labor Fund, which is a government special purpose fund. The Labor Fund obtains money from two main sources: mandatory contributions paid by employers, and state budget subsidies. In recent years the income of the Labor Fund did not cover its expenditure. As a result, the Fund needed to take out loans from commercial banks. 14.10 Labor fund expenditure consists of two main items. The first item is spending on financial benefits, which are distributed across the country according to actual needs (i.e., the number of eligible persons). The second item is spending on active labor market policies (ALMPs), such as training, job placement, and subsidized employment. Funds for ALMPs are distributed across regions based on the formula defined by the Minister of Labor. The formula takes into account the regional unemployment level compared to the national level, as well as specific types of unemployed persons (long-term unemployed and school leavers). As a result, those regions that have more unemployed (and more long-term unemployed and school leavers) 311 receive more funds. The Minister of Labor distributes the funds. The Minister also holds a special reserve that can be used to finance specific programs if they are beneficial for the labor market. 14.11 The regulations adopted in recent years have increased the funds appropriated for financial benefits. This has squeezed the active labor market program (Figure 14.1). The increase results entirely from higher expenditure on pre-retirement benefits and pre-retirement allowances. Rising unemployment, which in the period 2001-03 has been above 15 percent, combined with a low share of expenditure on ALMPs, shows that the existing labor market policy is overly focused on financial benefits. At the same time, the benefits provided to the unemployed are in most cases insufficient and unemployment has become one of the most important causes of poverty. Figure 14.1: Structure of Labor Fund Expenses, 1999-2003 100% 11.5 6.2 8.1 5.6 6.1 90% 5.4 11.1 7 8.2 80% 19.2 70% 60% 50% 89 40% 82.8 84.8 85.6 69.2 30% 20% 10% 0% 1999 2000 2001 2002 2003* Financial benefits ALMPs Other *Forecast. Source: MGPiPS. Unemployment Benefits 14.12 Unemployment benefits are provided to people who have lost their jobs, and are aimed at providing them with an income that allows them to search for work. 14.13 Benefits are paid to an unemployed person if he/she cannot find an appropriate job offer, or placement in a subsidized job. The beneficiary should have worked for at least 365 days during the 18 months prior to registration. 14.14 The level of unemployment benefit varies according to an individual's years of service before becoming unemployed. The basic amount of the benefit is set in nominal terms and is raised quarterly based on the inflation index. In December 2003, this amount was 503.20 PLN (about US$128) per month.18 If the unemployed person has less than 5 years of service, he or she 18In comparison, the minimum wage in 2003 was 800 PLN per month. 312 receives 80 percent of the base amount. If he or she has more than 20 years of service, he or she receives 120 percent of the base amount. 14.15 The duration of the benefit depends on the unemployment rate in the region. The unemployment benefit is paid for 6 months in regions where the unemployment rate is below the national average, for 12 months in regions where the unemployment rate is higher than the national average, but less than twice that level, and for 18 months, in regions where the unemployment rate is at least double the national average. 14.16 It is important to note that unemployment benefits provide income for a steadily falling percentage of the unemployed. As shown in Figure 14.2, at the end of 2002 only 16.7 percent of the unemployed were entitled to the benefit. This is due to the fact that most of the unemployed have been without jobs for a long period of time and have therefore lost the right to unemployment benefits. At the end of 2002 the proportion of long-term unemployed (defined as having spent more than 12 months without a job) exceeded 51 percent. 14.17 Additionally, a smaller share of those who register themselves as unemployed is eligible to receive unemployment benefits. This reflects two factors: the high percentage (over 70 percent) of people who re-enter the unemployment register; and the large number of school leavers. 14.18 However, a complete picture needs to take into account the data on the inflows into unemployment and the outflows from unemployment. There is a higher percentage of people eligible for the unemployment benefit among those who move out of unemployment than among those who move into unemployment. This indicates that a person entitled to the unemployment benefit is more likely to find work. This broad finding suggests that unemployment benefits can play a supportive role in finding a job. Figure 14.2: Percentage of Unemployed Eligible for the Unemployment Benefit in Various Categories, 1996-2002 80 70 60 50 40 30 20 10 0 1996 1997 1998 1999 2000 2001 2002 unemployed receiving benefit as % of all unemployed unemployed receiving benefit are % of unemployed flowing into unemployment unemployed receiving benefit as % of unemployed flowing out of unemployment Source: MGPiPS data. 313 14.19 The Ministry of Economy, Labor and Social Policy (MGPiPS) has prepared a new law on employment promotion that should come into force in mid-2004. Proposed changes include revised rules for unemployment benefit payments. In the original design, the changes were intended to reduce the unemployment trap. However, the proposal currently under discussion only slightly modifies current regulations. In addition to slight changes to the criteria that determines the duration of the entitlements, the proposal is to implement a re-employment bonus19. The bonus would amount to 30 percent of the unemployment benefit and the duration would be equal to half of the duration of unemployment benefit payment. For the unemployed who were directed by labor office to a part-time job, in case of the wage below minimum wage, the re-employment benefits will be equal to the difference between minimum and actual wage, but not more than 50 percent of unemployed benefit. Pre-retirement Transfers 14.20 The system of pre-retirement benefits and allowances was introduced in 1997. The main aim of this system was to replace early retirement schemes with benefits targeted to elderly workers who cannot find jobs. The system included two types of benefits: · Pre-retirement allowances for unemployed individuals who faced a relatively long wait to be eligible for the old-age pension (usually more than 5 years). For these individuals, the pre-retirement allowance was 120 percent of the base amount of the unemployment benefit. · Pre-retirement benefits granted to unemployed individuals who had roughly less than 5 years to wait before becoming eligible for the old-age pension. For these individuals, the pre-retirement benefit was equal to 80 percent of their accrued old-age pension, but not less than 120 percent of the base unemployment benefit and no more than 200 percent of the base unemployment benefit.20 14.21 Since 2002, only pre-retirement benefits have been provided; pre-retirement allowances have been discontinued. Labor offices pay all pre-retirement transfers on a monthly basis; payment depends on the date on which the applicant was declared eligible for the benefit. The change was caused by the need to reduce a rapidly rising expenditure on pre-retirement transfers. 14.22 If persons receiving pre-retirement transfers also receive income from work, become self- employed, or receive other payments linked to their work, benefits are suspended. This does not apply if the sum of the pre-retirement transfer and work income does not exceed 200 percent of the base unemployment benefit amount. 19According to the new proposal the duration would be equal to 6 months if the local unemployment rate is below 150% of national average, 12 months if the unemployment rate is above 150% but below 200% of national average and 18 months if the local unemployment rate is above 200% of national average). 20This limitation was introduced in 2002. It does not cover the following: · Individuals who accrued the right to pre-retirement benefits before January 1, 2002 · Persons who registered in the labor offices and filed an application for pre-retirement benefits by December 31, 2001, and who met all eligibility conditions on that date · Individuals who registered after January 1, 2002, and who accrued the right to pre-retirement benefits as a result of being dismissed from state-owned companies in the steel industry under restructuring · Employees whose employment contract was terminated due to difficulties faced by their employers in the period between July 1, 1998 and December 31, 2001 who are entitled to pre-retirement benefits equal to the full amount of the accrued old-age pension. 314 14.23 Pre-retirement transfers are financed from the state budget subsidy to the Labor Fund. Since the beginning of 2002, pre-retirement benefits and allowances have been subject to annual indexation based on the inflation index. Eligibility criteria for pre-retirement benefits are given in Box 14.2. Box 14.2: Eligibility Criteria for Pre-retirement Benefits The criteria for granting pre-retirement benefits include: age; years of service; and the way in which the employment contract was terminated. Pre-retirement benefits are granted to persons who: · Are older than 58 years (women) or 63 years (men) and have worked for more than 20 years (women) or 25 years (men), or · Are older than 50 years (women) or 55 years (men) and have worked for more than 30 years (women) or 35 years (men), who were employed for more than 6 months by their last employer and whose employment contract was terminated owing to difficulties faced by their employer, or · Have worked for more than 35 years (women) or 40 years (men) how were employed for more than 6 months by the last employer and whose employment contract was terminated owing to difficulties faced by their employer, or · Have worked for more than 34 years (women) or 39 years (men) by December 31 of the year preceding contract termination and when the termination was due to the employer's insolvency. 14.24 Between 1997 and 2001 there was a dynamic increase in the number of recipients of pre- retirement benefits and allowances. The number of pre-retirement benefit recipients rose from 102,000 at the end of 1997 to 365,000 at the end of 2001. The number of pre-retirement allowances recipients at the end of 2001 was 113,500 compared to 1,300 at the end of 1997 (see Table 14.1). Table 14.1: Individuals Entitled to Pre-retirement Benefits and Allowances in 1997 ­ 2002 (end of year) Pre-retirement benefits Pre-retirement allowances Year total Women men total women men 1997 102,409 52,025 50,384 1,357 487 870 1998 131,628 68,612 63,016 4,754 2,050 2,704 1999 196,328 103,847 92,481 32,443 14,345 18,098 2000 270,975 142,301 128,674 72,593 31,745 40,848 2001 365,621 183,670 181,951 113,546 47,977 65,569 2002 334,348 163,265 171,083 162,792 75,400 87,392 Source: Labor Market Department, MGPiPS. 14.25 The increase in the number of beneficiaries was accompanied by a rapid rise in Labor Fund expenditures for this purpose. In 2001, expenditures on pre-retirement benefits amounted to 2,196.1 million PLN, accounting for 25.8 percent of total Labor Fund expenditures; pre- retirement allowances expenditures amounted to 1,187.3 million PLN (13.9 percent of total expenditures). 315 B. SOCIAL ASSISTANCE AND INCOME SUPPORT FOR THE POOREST 14.26 The social protection system aimed at the poorest individuals and families consists mainly of social assistance activities and non-insurance benefits designed to support families with children. This section describes the three main types of income support ­ social assistance, family benefits and housing allowances. 14.27 The current social assistance system was initiated with the transfer of social assistance- related tasks from the Ministry of Health to the Ministry of Labor and Social Policy and by the passing of a law on social assistance in 1991 (for the first time since World War II).21 The law re- shaped the institutional structure of social assistance and introduced criteria for its availability. The criteria include an income test as well as a list of social dysfunctions that can determine eligibility for social assistance. 14.28 Dysfunction in an individual or a family is defined to include: orphanhood, homelessness, the need to protect motherhood or many children, unemployment, disability, long-term sickness, lack of capacity for housekeeping and child rearing (in particular in multi-children families or broken homes), alcoholism, drug addition, problems with adaptation after leaving prison, natural disasters, and ecological catastrophes. 14.29 The income test for social assistance depends on the composition of the household and is defined in net terms.22 For example, in 2002 a single person household could receive support if the monthly income (net of taxes and social security contributions) was less than 454.30 PLN. In a two-person (with two adults) household, the limit was 702.10 PLN. 14.30 Social assistance financing is decentralized. There are two sources of financing ­ state and municipality budgets. Most of the benefits provided by social welfare are financed from the state budget. The financing is provided in two ways ­ either through special purpose subsidies to finance mandatory tasks defined by the central administration, or through subsidizing local government activities. The local authorities' resources finance only lump-sum benefits ­ benefits in kind and services. The system is organized at all three levels of regional and local administration (see Box 14.3). 21By 1989 the law on social assistance dating from 1923 was effectively supported only by unpublished instructions from the Ministry of Health and Social Protection. 22Income without taxes and social contributions is taken into account. 316 Box 14.3: Institutional Framework of Social Assistance The social assistance system's organizational model is decentralized, with each level of local government accountable for a different area of activities. Social assistance centers at the district level are responsible for all benefits provided in the domicile of a recipient. Centers for family assistance at the poviat level are responsible for moving the recipient to a new place of residence -- i.e., to a social assistance facility (operating beyond the borders of the municipality), a child care center, a foster family, etc. The poviat administration is also responsible for granting assistance to refugees. Regional centers for social assistance are responsible for organizing and administering centers at the regional level, preparing financial information on regional social assistance expenditure and revenue, and training social assistance employees. 14.31 As a result of decontrolling rent prices in 1990, the government launched a program aimed at providing the poorest families with compensation for the increased costs of housing. Housing allowances, which were primarily a social assistance activity, were transferred to the ministry responsible for housing policy. Housing assistance is provided at the district level. All regulations are created centrally (currently in the State Office for Housing and Urban Development). Housing allowances are financed by local authorities and the state budget, in a similar way to the financing of social assistance. The income test for housing allowances is linked to the minimum pension. For a one-person household, income cannot exceed 150 percent of the minimum pension. In households with more than one member, income per household member should not be higher than 100 percent of the minimum pension. 14.32 Family allowances apply to families with children (aged up to 16 years, or up to 20 years if still being educated). Until 1995 these benefits were paid to all families with children. Since 1995, family benefits have been means-tested. As a result, where total expenditure on family allowances has not changed, the level of individual benefits has increased, as the number of families entitled to this benefit is smaller. Family benefits are paid by the Social Security Administration (for employees and farmers separately), companies, labor offices and social assistance centers. Family benefits are financed from the state budget. Since June 2002, the income test for family benefits has been defined in net terms (548 PLN per person in the household). It had previously been linked to the percentage of the average gross wage in the economy. Social Assistance for the Unemployed 14.33 The unemployed face a high risk of poverty. In 2001, 37 percent of the unemployed had incomes below 60 percent of the median income (which is the relative poverty line). According to the law, social assistance is not provided to the unemployed in every case. For example, a social assistance worker can decline to grant financial assistance if an unemployed person refuses employment for unjustified reasons. 14.34 Social assistance offers financial and non-financial benefits to the unemployed. The financial benefits include temporary benefits, guaranteed temporary benefits, and "purpose" 317 benefits. An indirect form of financial benefit is the payment of health premiums and social security premiums by the state. Non-financial benefits include various forms of assistance, such as accommodation, clothing, etc. 14.35 The guaranteed temporary benefit is obligatory when an individual meets certain specific conditions (an income criterion, unemployment benefit expiry, single parenthood) for a defined period of time (36 months). This benefit is thus addressed to a narrow group of the unemployed, such as a single person bringing up a child. Entitlement to the guaranteed temporary benefit does not require readiness to take up work; however, it can be suspended if an individual takes up temporary employment and can be re-granted at the end of the temporary assignment if the 36- month period has not expired. 14.36 The temporary benefit is addressed to a wide group of the unemployed. However, this is a short-term form of assistance (offered for three months on average) in a defined amount (up to a statutory limit of the maximum amount of social assistance). Social assistance offices have discretion in granting these benefits. In reality, given the scarcity of financial resources for social assistance, the amount of benefit depends on the financial situation of the local social assistance budget after all mandatory benefits have been paid. 14.37 In special cases temporary benefits are granted to people with an income exceeding the statutory defined poverty limit. Unemployed people may also belong to this group. Such assistance is called a temporary special benefit. 14.38 "Purpose" benefits and special purpose benefits are granted mainly to buy medicines or fuel, to cover the costs of kindergarten or nursery care, to cover apartment renovation, or to cover costs resulting from difficult life events. Selected Information on the Provision of Social Assistance 14.39 In recent years almost half of all families applying for social assistance give unemployment as one of the reasons for applying for help. In 2000, the number of families citing unemployment was 709,000; in 2002 it reached 861,000 families (Table 14.2). Those families accounted for 43.3 percent of all beneficiaries in 2000 and 47.5 percent in 2002. 318 Table 14.2: Social Assistance Recipients in Crisis Situations, 1999-2002 Reasons for crisis situation 1999 2000 2001 2002 Number of Families Poverty 712,193 928,878 922,726 972,466 Orphanhood 14,212 20,676 18,925 20,490 Homelessness 20,746 19,621 22,812 24,916 Necessity of motherhood protection 147,357 143,195 139,434 127,553 Unemployment 678,505 708,649 737,207 861,139 Disability 354,081 392,261 427,510 491,938 Long-term sickness 355,702 365,157 364,618 403,103 Lack of capacity for housekeeping and child rearing - total 419,550 415,979 392,888 435,452 - one parent families 188,589 185,646 274,272 193,206 - multi-children families 159,777 152,093 142,865 157,446 Alcoholism 119,143 114,831 106,443 115,392 Drug addiction 3,151 3,501 3,164 3,809 Adaptation difficulties after leaving prison 16,207 13,709 13,487 15,119 Natural disasters 2,499 5,424 18,607 2,752 Note: Data cannot be summed up as the reasons for applying for assistance may be multiple and the family can give more than one reason (maximum three). Source: MGPiPS, Social Assistance Department data. 14.40 In recent years the number of persons receiving temporary benefits has fluctuated. It decreased from 387,000 in 1999 to 195,000 in 2001, but increased again to 354,000 in 2002. Expenses for temporary benefits granted to the unemployed decreased from 275 million PLN in 1999 to 71 million PLN in 2001 (i.e., by almost 75 percent). In 2002, following the increase in the number of beneficiaries, expenditure on these benefits increased to 173 million PLN. The average duration of the temporary benefit payment to the unemployed in 1999 was 4.2 months; in 2001 it was 2.6 months, and in 2002 it was 2.9 months. The average amount of temporary benefit granted for unemployment was 168 PLN in 1999; by 2001 it was 140 PLN, (17 percent lower). In 2002 the average amount rose again to the 1999 level (167 PLN). 14.41 The number of guaranteed temporary benefits recipients decreased from 56,000 in 1999 to 31,000 in 2001 (i.e., by almost 45 percent), and in 2002 by another 11.3 percent to 27,500. The average duration of the guaranteed temporary benefit payment fell from 10.5 months in 1999 to 8.3 months in 2001 (i.e., by about 20 percent). In 2002 the period lengthened by one month. The average amount of the guaranteed temporary benefit was 295 PLN in 1999 and 369 PLN in 2001; in 2002 it increased to 386 PLN. 14.42 Special temporary benefits were paid to 57,000 people in 1999 (among whom there could be unemployed persons), in the amount of 180 PLN on average for a period of 2.7 months. In 2001, these benefits were paid to about 11,000 people, in the amount of 163 PLN on average for a period of 1.8 months. In 2002, 19,600 recipients received the benefit for approximately 2.1 months in the amount of 174 PLN. 319 14.43 Current social assistance reporting only provides overall expenditure data on benefits, with no information on individuals. This breakdown will be possible under the POMOST system, which will feed the central database, facilitating analysis and providing detailed information on recipients. 14.44 The POMOST system is still being built and is operational in half of the social assistance centers. The list in Table 14.3 does not provide a complete picture of the assistance granted to the unemployed, and therefore the analysis does not give the number of recipients and focuses only on the percentages applied to different benefits. Table 14.3: Most Frequently Granted Benefits to Families with Unemployed Applying for Social Assistance, 2000-2002 Benefit 2000 2001 2002 Fixed benefit for a person who gave up work in order to bring up a child requiring permanent care 0.6 0.9 0.8 Temporary benefit because of involuntary unemployment 30.6 12.9 17.11 Guaranteed temporary benefit for a person who lost the right to unemployment benefits because of the 3.9 1.3 1.1 expiry of the payment period Other temporary benefits because of the impossibility of employment 1.6 0.4 0.6 Pregnancy benefit / child care benefit 2.4 3.7 0.6 Meals for schoolchildren 16.1 26.4 32.0 Meals for other people 0.5 1.4 1.3 Purpose benefit to cover medical costs 2.9 5.0 - Purpose benefit to cover fuel costs 7.7 9.2 3.7 Purpose benefit for purchase of clothes 4.9 5.0 3.7 Purpose benefit to satisfy other needs 28.8 33.9 38.3 Source: MGPiPS based on POMOST. Projected Changes in Social Assistance and Family Benefits 14.45 Currently, the social assistance system is undergoing significant legal changes. The Ministry of Economy, Labor and Social Policy has undertaken actions aimed at standardizing family benefits and allowances by defining clear rules and eligibility criteria and establishing a unified organizational system with one payer for all beneficiaries. Family benefits will be supported by an education allowance addressed to families with schoolchildren who incur the costs of accommodation outside of the permanent place of residence, the costs of travel to school, and the purchase of school materials and books. 14.46 The family benefits system will take over part of the mandatory allowances for social assistance. As a result, social workers will have more time to work with social assistance beneficiaries. 14.47 Another proposal includes introduction of a contract between the recipient of support and the social worker. The contract would oblige the recipient to actively search for work. 14.48 Housing allowances will be included in the social assistance packages and the income test will be equalized for both types of support. 320 14.49 Currently, the government is progressing with a draft law on social employment, which is aimed at creating special forms of employment for people threatened by social exclusion. Social employment is envisaged as partially therapeutic (for example, for former drug or alcohol addicts), and connected with professional re-adaptation and training. C. SOCIAL SECURITY SYSTEM DISABILITY PENSIONS 14.50 People who have not yet reached the retirement age and who, for health reasons, are unable to work may apply for the disability pension. In the past, disability pensions, along with early retirement pensions, were frequently perceived as alternative source of income for workers who had lost their jobs and were unable to find new ones. This was particularly common at the beginning of the 1990s. Consequently, in the current social security system, some 2 million beneficiaries receive early retirement or disability pensions granted under very relaxed conditions. Not surprisingly, one-third of all pensioners are disability pensioners. Poland ranks among the countries with the highest percentage of citizens claiming disability compared to the total population. Institutional Framework 14.51 The pension system for employees is a form of social insurance that covers pension insurance (old-age, disability and survivor), sickness insurance, and work injury. The system is administered by the Social Security Office (ZUS). The ZUS structure consists of a head office, 51 branches, and 216 inspectorates, with some 50,000 employees. The institution is responsible for the collection of contributions and the approval and payment of benefits. 14.52 Eligibility for a disability pension requires a work disability confirmation issued by a ZUS doctor. The disability pension system was significantly modified in 1998 when the eligibility criteria were tightened considerably. Before 1998, any health deterioration was a basis for granting a pension, but since 1998 the only eligibility criterion is the confirmed inability to work. The system of work disability certification was also changed, with medical commissions being replaced by special doctors employed by ZUS. ZUS also provides health rehabilitation programs for people threatened by permanent work disability. 14.53 Disability pensions are paid from the Social Security Fund (FUS). FUS is fully financed by social security contributions. The disability and survivor insurance contribution amounts to 13 percentage points of the contribution (the total contribution is about 37 percent23 of gross wage). Pension contributions are paid in equal share by the employee and the employer. Eligibility Criteria24 14.54 In order to be eligible for the disability pension, the following criteria must be met: work disability; a minimum of five years of social security contribution payments25; and a work disability that occurred no more than 18 months since the individual ceased to contribute. 23The total amount of the contribution depends on the amount of the casualty insurance contribution, which in turn depends on the risk of accidents in a given industry. 24Eligibility criteria for disability pensions are specified by the law on retirement and disability pensions, from the Social Security Fund, dated December 30, 1998 (Journal of Laws, 1998, No. 162, item 1118 with further amendments). 25In the case of people below the age of 30, this period is shortened depending on age (one year for people below 20, to four years for people between 25 and 30 years of age) 321 14.55 A ZUS-employed doctor must certify work disability. A person is declared disabled for work if he/she has partially or completely lost the ability to work for reasons of health and is unlikely to regain such ability after retraining. There are two levels of work disability: · Complete disability, when a person has lost the ability to do any sort of work · Partial disability, when a person has substantially lost the ability to do work matching his/her qualifications. 14.56 The level of disability and the potential for recovery are assessed on the basis of: · The level of health deterioration and the likelihood of recovery in the course of medical treatment and rehabilitation · The possibility of taking up existing or other work and the chances for successful retraining, taking into account the type and character of the existing occupation, the individual's level of education, age, and the individual's psychological as well as physical condition. 14.57 There can be permanent and term-limited disability, depending on the projections for recovery. 14.58 The disability pension amount is subject to the total period over which social insurance contributions have been paid and the salary amount in the period before the receipt of the disability benefit. The disability pension is granted according to a defined-benefit formula. The benefits consist of two elements: · A flat rate component equal to 24 percent of the average wage in the economy · An individual component representing 1.3 percent of the individual's average wage26 for each year of employment and 0.7 percent of the individual's wage for other periods (such as maternity and child leave or periods of university education). 14.59 The disability pension for persons with complete disability is 100 percent of the calculated benefit. The disability pension for an individual with partial disability equals 75 percent of the calculated amount. All benefits are indexed annually to reflect changes in inflation and one-fifth of the real growth of average wages. Basic Information about Disability Pensions 14.60 Since 1990 both the number of disability pensioners and their percentage share in the total number of beneficiaries has changed significantly (Figure 14.3). The number of disability pensioners rose from 2.16 million in 1990 to 2.70 million in 1998. The costs of disability pensions increased from 2.5 percent of GDP in 1990 to 4.1 percent of GDP in 1994. These costs fell to 3.2 percent of GDP in 2000. Between 1995 and 2000 the average disability pension amount fell from 50.9 percent of average pay to 44.9 percent (of a salary reduced by social security premiums). In the following years this ratio rose slightly to the level of 46.8 percent. 14.61 After 1999, mainly as a result of regulatory changes in the disability declaration process, the number of pensioners decreased, reaching 2.4 million in 2002. But, despite the rise in the 26An individual's average wage is calculated based on the average wage for 10 consecutive years of the last 20 years of recorded wage. 322 number of pensioners in the 1990s, their percentage share in the total number of beneficiaries decreased from 40 percent in 1990 to 34 percent in 2002. Figure 14.3: Number of Disability Pensioners, 1990-2002 3 000 42% ) 2 500 40% hst( sr 2 000 38% onssr neoi pe 1 500 36% ns ofr 1 000 34% be pellafo % num 500 32% - 30% 1990 1992 1994 1996 1998 2000 2002 Source: Authors' analysis of ZUS data. 14.62 The growth in the number of disability pensioners was particularly rapid at the beginning of the 1990s, when the old age and disability pensions systems were treated as a buffer absorbing the excess workforce on the labor market. The number of newly granted benefits of both types started to decrease after 1991, but the number was still high (more than 150,000 new pensioners annually). From 2000, following the changes in the assessment regulations introduced in 1998, the number of newly granted pensions has been below 100,000 annually (Figure 14.4). Figure 14.4: Number of Newly Granted Benefits, 1990-2002 600 500 snosreP.s 400 300 th 200 100 0 1990 1992 1994 1996 1998 2000 2002 old-age disability Source: ZUS. 14.63 The likelihood of a person's becoming a disability pensioner increases with the person's age (up to the 45-49 year old age group). In 2002, the average age of this type of pensioner was 46.8 years (47.4 years for men and 45.9 years for women) (Figure 14.5). Most frequently, 323 disability pensions were granted to people between 45 and 54 years of age (over half of the benefits granted by value). Indeed, at this age the opportunity to find another job for the unemployed decreases significantly. Figure 14.5: Structure of Newly Granted Disability Pensions by Age in 2002 35 Total 30 wolf Men Women inlatotfotnecrep25 20 15 10 5 0 and ssle 24- 29- 34- 39- 44- 49- 54- 59- 64- and er 19 20 25 30 35 40 45 50 55 60 65 mo Source: ZUS. 14.64 The structure of the newly granted disability pensions determines the age structure of pensioners: almost three-fourths of the pensioners are over 45 years of age (Figure 14.6). Figure 14.6: Structure of Disability Pensioners by Age in 2002 100 sr 90 Total ne 80 Men io ns 70 Women pe 60 ytilibas 50 40 30 di 20 of 10 % 0 ssel 34- 39- 44- 49- 54- 59- 64- 69- 74- 79- erom 30 35 40 45 50 55 60 65 70 75 and and 29 80 Source: ZUS. 324 14.65 Disability pensioners have a low ratio of professional activity, which means that the disability pension can be treated as an alternative source of income if no other earning possibilities exist. In 1995, the labor force participation rate of disability pensioners aged 15-64 was 28.4 percent,27 while in 2002 the rate fell to 24.5 percent.28 Box 14.4 discusses the need for additional changes in the disability pension system. Box 14.4: The Need for Further Changes in the Disability Pension System Despite the decreasing number of new pensioners, their total is still high and the expenses for disability pensions constitute a serious burden for the state budget. This leads to high costs of labor, while at the same time the professional activity of disability pensioners is very low. In the future, further actions aimed at reducing expenditures for the disability pension system will need to be undertaken. At the same time, the government will have to develop incentives that increase the professional activity of disability pensioners. In 2003, as an element of a social expenditure review, the government proposed a review of some disability pensioners as well as further changes to the pension and rehabilitation system for the disabled that would reduce expenditures on these policies and would stimulate higher employment of this group. D. THE ROLE OF SOCIAL BENEFITS IN REPLACING WAGE INCOME 14.66 This section presents the results of a micro simulation of how unemployment and social assistance benefits affect the financial situation of households with an unemployed family member. The simulation covers social assistance benefits, family benefits and housing allowances. 14.67 The simulation does not take into account pre-retirement benefits or disability pensions, because: (i) the size these of benefits is linked to the previous wage; and (ii) entitlement for these types of transfers depends on meeting additional criteria (of age, years of service, disability certification). Assumptions for the Simulation 14.68 In order to test the impact of unemployment on the financial situation of selected types of families, we performed simulations with regard to the following household categories: · Single person · Parents (one of whom is working) with two children to be supported (the most typical family constellation in Poland) · Parents (one of whom is working) with four children to be supported (to serve as an example of a multi-child family) · Single person with two children to be supported (to serve as an example of a family to which special benefits are addressed). 27Data for November 1995, based on LFS. 28Mid-year data based on LFS. In the fourth quarter the activity rate of people on disability pensions was 23.3 percent. 325 14.69 The simulation took into account two levels of past earnings: 1) The family member earned an average salary before losing his or her job29 2) The family member earned a minimum salary before losing his or her job. 14.70 The simulation was performed for each of the listed family types. Such a set of microsimulations allows us to compare the level of income from unemployment benefits and social assistance benefits to the income from the salary earned before the loss of a job. For each of the family types and past salary levels we calculate an "income substitution ratio" which reflects the extent to which social benefits replace the previous level of earnings. 14.71 However, such an analysis shows only how benefits, based on existing regulations, can provide income to a standardized family. In reality, social transfers to beneficiaries may differ significantly (and will usually be lower). 14.72 In Table 14.4 we list the income tests used to assess the right to a given type of benefit. The table shows that the income criteria usually increase with the number of persons in the family. As a result, a single parent with two children has a lower income test than a family consisting of parents with two children. Table 14.4: Income Test (per family) for Benefits by Household Type in 2002 Income criteria Single person Parents with two Parents with four Single person children children with two children Average per month in PLN Social assistance (net) 454.30 1,115.10 1,528.10 836.00 Family benefits (gross) 548,00 2,192.00 3,288.00 1,644.00 Housing allowances (gross) 758,04 2,127.24 3,190.86 1,595.43 Source: Authors' calculation. Results of the Simulation 14.73 We present below the analysis of the income substitution ratio for selected categories of families and wage levels. Summary results of the estimates are presented in Table 14.5.30 We take into account the hypothetical situation in which the unemployed person in the family receives a benefit equal to 80 percent, 100 percent or 120 percent of the base benefit amount as well as a situation in which the unemployed person does not receive a benefit. 29In 2002 the average monthly wage was 2133.31 PLN and the minimum monthly wage was 760 PLN. 30Detailed tables are included in the annexes. 326 Table 14.5: Income Substitution Ratio by Family Type and Income Level Unemployment benefits after work loss Assistance after Family type/past wage level 80% 100% 120% unemployment benefit expiry Single person, average wage 35.2 40.5 45.8 7.0 Single person, minimum wage 91.8 105.7 119.6 47.0 Parents with two children, average wage 46.2 50.1 54.0 28.5 Parents with two children, minimum wage 81.6 88.5 95.4 50.3 Parents with four children, average wage 49.2 52.5 55.8 34.2 Parents with four children, minimum wage 84.8 90.5 96.2 58.9 Single parent with two children, 44.4 48.5 52.8 (43.1) 47.5 average wage Single parent with two children, 78.7 86.1 93.7 (76.5) 84.3 minimum wage Note: Values including the change of guaranteed temporary benefit in the second and third year of payment are in parentheses. Source: Authors' estimates. 14.74 In the case of a single person household, if this person earned an average wage, having lost his/her job, the person became entitled to unemployment benefit and could also apply for a housing allowance because the benefit amount (regardless of the level) met the income criteria for additional benefits. The person's income has now become about 35 to 45 percent of his or her former wage income. After the unemployment benefit expiry date, this person retains the right to the housing allowance and acquires the right to temporary social assistance. As the social assistance benefit is a short-term benefit, the person's income only reaches 7 percent of his/her previous wage income. But if such a person earned the minimum wage, his/her income, while the unemployment benefit is being paid, may exceed the salary income by about 5 to 20 percent. 14.75 In the case of a household consisting of parents with two children, before one of the parents lost his/her average wage job, the family had already met the income criterion and was eligible for family benefits paid by the employer. The family slightly exceeded the eligibility limit for the housing allowance. After losing a job, one of the parents is entitled to unemployment benefits, family benefits, and the housing allowance. The low level of income per family member (including all of the assistance benefits) makes the family eligible for temporary social assistance. The family income while the unemployment benefit is paid equals 46 to 54 percent of wage income. After the unemployment benefit expires, the family retains the right to receive family benefits and the housing allowance and can receive a temporary benefit in a slightly higher amount. The family income at this point is about 28 percent of the former wage income. 14.76 The income of a similar family when a parent earned the minimum wage is significantly different. Already, when a parent is working, the family can receive family benefits and a housing allowance. The total income is still low enough to make the family eligible for the temporary 327 social assistance benefit. With one family member unemployed, the family receives, apart from the unemployment benefit, a family benefit, a housing allowance, and a temporary social assistance benefit. Their income at that time equals 80 to 95 percent of the former wage income. After the unemployment benefit expires, the remaining sources of income do not change and the total income is still about 50 percent of the former wage income. 14.77 The income substitution ratio increases for larger families. A family with four children with an average wage earner is entitled to family benefits and the housing allowance. When receiving the unemployment benefit, the family is eligible for temporary social assistance and its income amounts to about 50 to 56 percent of the former wage income. After the unemployment benefit expires, this family's income is reduced to about 34 percent. A similar family with a minimum wage earner is eligible for all of the analyzed benefits during the employment of the parent. When one of the parents becomes unemployed, this family's income is about 85 to 96 percent of the former wage income. After the unemployment benefit expires, this family's income is reduced to about 60 percent. 14.78 In the case of a household consisting of a single parent with an average wage and two children, in which the parent becomes unemployed, the family is eligible for all family benefits for children, the housing allowance and the temporary social assistance benefit. The family's income at this point is about 44 to 53 percent of the previously earned average wage. After the unemployment benefit expires, the single parent with at least one child up to 7 years of age is eligible for a three-year guaranteed temporary social assistance benefit. In the first year this benefit amount is 100 percent, and is then reduced to 80 percent. The family income at this point is about 48 percent of the former wage income; in the two following years it falls to about 43 percent. 14.79 The single parent household with a minimum wage and two children is eligible for family benefits for children, as well as the housing allowance and a minimum temporary social assistance benefit. The right to these benefits is also retained during the unemployment period. The family's income during this period is about 80 to 94 percent of the previously earned wage. After the unemployment benefit expires, this income is slightly lower ­ about 84 percent in the first year and about 77 percent in the second and third years. 14.80 Before 2002, a fairly long (and a priori defined) period of receiving the temporary social assistance benefit payment (without any check being made of the person's readiness for work), and a fairly high age limit for the child (16 years) created a disincentive among these beneficiaries for seeking work. Since 2002, the child's maximum age has been reduced from 16 to 7 years for families applying for a guaranteed temporary benefit. E. INCOME OF INDIVIDUALS VERSUS THEIR LABOR MARKET SITUATION 14.81 This section discusses the income analysis of individuals in relation to their status in the labor market. The focus of the analysis is on the sources of income of the unemployed. In contrast to the previous section, which presented hypothetical benefits, this section shows the actual sources of income of the unemployed, including, in particular, social transfers. 14.82 Labor market status is one of the most important factors determining an individual's financial situation. According to the GUS31 survey on the living standards of families in Poland, 31Central Statistical Office. 328 the unemployed are most frequently affected by poverty. A family with an unemployed member has a four times higher risk of falling into poverty below the subsistence level and a three times higher risk of relative poverty (expressed as 50 percent of the average monthly consumption of the households in total) than the average. 14.83 In 2002, about 36 percent of households with an unemployed family member had an income below the relative poverty line, and this figure has not changed over recent years ­ in 1999 the proportion was about 35 percent of such households. By comparison, in 2002 only 13 percent of families that did not have an unemployed member were below the poverty line. 14.84 According to the research on "Social Diagnosis 2000," between 1999 and 2000, 12 percent of respondents who assessed their financial situation as poor or very poor declared that the main reason for their situation was unemployment. The same research showed that 6 percent of all families had at least one unemployed member and neither of the adults was working. A third of such families lived only on social transfers. These households also declared receiving support from their families or friends in the form of goods and sometimes money. Assistance from charity organizations (civil and church) and from trade unions and business companies was marginal. Sources of Income of the Unemployed 14.85 According to Labor Force Survey research in the fourth quarter of 2002, around one-sixth of the unemployed declared the unemployment benefit as their main source of income (which corresponds to the share of the registered unemployed who were entitled to such a benefit). A similar share of the unemployed declared other, non-earning, sources as their main source of support. Other people, either employed or obtaining income from other sources, supported two- thirds of the unemployed (Figure 14.7). 14.86 In order to analyze possible changes in sources of income, we also looked at the sources of income of those who were unemployed 12 months before the survey was conducted, which is also illustrated in Figure 14.7. As shown later in this section, some two thirds of this group was still unemployed. 14.87 In the case of the second group, the unemployment benefit was the main source of income for a relatively small group of the unemployed, as the right to the benefit expired. The larger percentage of this group, compared to all of the currently unemployed, received income from other non-earning sources (such as pre-retirement benefits or social assistance benefits). Only around one-fifth of persons unemployed one year before the survey declared work income as their main source of income. This means that the percentage of people supported by others was lower, compared to the total unemployed population. 14.88 A relatively small percentage of respondents who were unemployed when the survey was conducted (or declared unemployment a year before the survey) indicate the disability pension as their main source of income. 329 Figure 14.7: Main Sources of Income of People Unemployed at the Time of the Survey and Unemployed 12 Months Before the Survey Supported by a person with non-earning i 100% income 90% Supported by a person working in agriculture 80% Supported by a person working outside of agriculture 70% Other non-earning sources 60% 50% Unemployment benefit 40% Disability pension 30% retirement 20% 10% Work in agriculture 0% Work outside of agriculture Sources of income of Sources of income of people unemployed the unemployed In the previous year No answer Source: Authors' estimates based on the LFS. Changes in the Labor Market 14.89 It is worth tracing the way in which the situation of the unemployed changed in the course of a year. The LFS data show that a significant proportion of the unemployed in the previous year still do not have work (over 60 percent). The percentage is similar for both men and women. However, gender differences are more significant in the case of the unemployed who changed their status in the labor market (Table 14.6). 14.90 The analysis of persons who drop out of the labor market shows that the probability of becoming inactive is much higher in the case of the unemployed compared to the employed. This is particularly clear in the case of unemployed women who dropped out of the labor force twice as frequently as unemployed men. As the survey results show, unemployed men tend to find work more often than unemployed women. 330 Table 14.6: Change in Labor Market Status After One Year, Based on LFS, 2002 (in percent) Current status Employed Unemployed Inactive Total Work 91.5 5.7 2.9 48.4 Unemployment 22.1 67.0 10.9 10.7 Education, training 6.5 7.7 85.8 14.7 Retirement, early retirement 2.6 0.8 96.6 14.4 before Disability 4.2 3.7 92.1 9.1 Compulsory military 30.8 62.4 6.8 0.7 year service Family duties 15.7 21.7 62.7 0.1 Other forms of inactivity 6.0 15.6 78.4 1.9 Situation Men Total 48.7 12.2 39.0 100.0 Current status Employed Unemployed Inactive Total Work 90.9 4.1 4.9 41.1 Unemployment 16.2 64.4 19.4 10.9 Education, training 6.0 7.7 86.2 15.1 Retirement, early retirement 1.8 0.6 97.7 21.4 before Disability 2.0 2.4 95.6 11.2 Compulsory military 0.0 0.0 0.0 0.0 year service Family duties 5.7 9.8 84.5 6.8 Other forms of inactivity 2.3 6.8 91.0 4.0 Situation womenTotal 37.2 10.2 52.6 100.0 Source: Authors' estimates, based on the LFS. 14.91 Table 14.7 presents a more detailed analysis of persons who were outside of the labor force in 2002. The table shows the sources of income of these persons and gives their labor market status in the previous year, separately for men and women. The sample is also divided into two age categories: 15 to 44 years (in the so-called mobile age) and older than 45 years (in the so- called immobile age). 14.92 Significant differences are apparent between the situations of younger and older people. Younger people more frequently tend to be supported by others. Among people over 45 years of age, apart from the obvious percentage of old-age pensioners, disability pensions and other non- earning sources of income form a significant income alternative. Men are more likely to support themselves from one of the two latter sources. The percentage of people unemployed in the previous year, supporting themselves from non-earning sources of income, is particularly high for both men and women across both age groups. 331 332 Table 14.7: Structure of Persons Outside of the Labor Market by Source of Income, Age, Gender, and Labor Market Status in the Previous Year (in percent) MEN Last year's situation Current source of income: 15 ­ 44 years 45 years and more employed unemployed total employed unemployed total Work outside of agriculture 1.5 0.0 0.1 0.5 0.0 0.0 Work in agriculture 0.7 0.2 0.1 0.3 0.0 0.1 Retirement, early retirement 4.8 0.0 0.3 37.6 3.0 60.1 Disability pension 14.1 3.8 7.7 19.1 8.1 30.2 Unemployment benefit 23.0 2.2 0.8 5.7 4.3 0.5 Other non-earning sources 24.8 21.6 8.6 33.5 47.9 7.4 Supported by a person employed outside of agriculture 19.3 28.4 57.7 1.3 11.5 0.6 Supported by a person employed in agriculture 1.5 3.3 8.5 0.1 0.3 0.1 Supported by a person with non- earning sources of income 10.4 40.4 16.3 1.9 25.0 1.1 WOMEN Last year's situation Current source of income: 15 ­ 44 years 45 years and more employed unemployed total employed unemployed total Work outside of agriculture 0.9 0.6 0.1 0.4 0.0 0.0 Work in agriculture 0.0 0.0 0.0 0.1 0.4 0.1 Retirement, early retirement 0.1 0.0 0.1 51.5 3.5 56.0 Disability pension 4.2 1.8 4.9 11.5 4.2 23.9 Unemployment benefit 16.0 1.8 0.9 3.5 4.8 0.3 Other non-earning sources 35.4 22.0 13.7 29.0 31.6 12.3 Supported by a person employed outside of agriculture 34.4 53.4 58.0 2.1 26.5 2.3 Supported by a person employed in agriculture 3.9 5.0 8.2 0.3 1.4 0.2 Supported by a person with non- earning sources of income 5.0 15.4 13.9 1.6 27.7 4.9 Source: Authors' estimates based on the LFS. 14.93 The analysis presented in Table 14.7 is also confirmed by the results of the logit model that analyzed the impact of the individual's past and current position in the labor market with respect to the probability of receiving income from non-earning sources or from a disability pension.32 On the basis of this regression the following conclusions can be drawn: 32Detailed results of the model are presented in Annex 2 333 · Women support themselves from non-earning sources more frequently than men · The higher the education level is, the lower is the chance that people will support themselves from non-earning sources of income · If people were unemployed 12 months before the survey, the chance that they would support themselves from non-earning sources is over three times higher than for people who were not unemployed at that time · The likelihood of seeking support from non-earning sources of income is also higher (but to a lesser extent) among respondents who were unemployed during the time the survey was conducted · The likelihood of seeking support from non-earning sources of income increases with age. 14.94 As far as disability pension recipients are concerned, the individual characteristics that increase the likelihood of supporting oneself from the disability pension are sex (women live solely on the disability pension more frequently than men) and age (the likelihood of supporting oneself from the disability pension increases with age up to age 49). People who were unemployed a year before the research took place and during the research are less likely to support themselves with the disability pension than those who were not unemployed at that time. This shows that disability pensions are no longer treated as an alternative source of income for the unemployed. F. CONCLUSIONS 14.95 The social transfers system is a source of support for a significant group of the unemployed. But some transfers in the system under certain conditions create incentives for the unemployed not to look for a job (particularly a low-paid one). 14.96 The level of unemployment benefits and social assistance benefits does not depend primarily on previously earned income. The eligibility for potential transfers depends instead on the family type and income received per family member. In particular, in the case of minimum wage earners the potential social benefits received during the period of unemployment (especially when a person is eligible for the unemployment benefit) are similar to the previous wage income. The level of living in these families, though it is very low, does not fall significantly as a result of loss of work. This kind of relationship between the minimum wage and social transfers may create disincentives to re-employment. However, this is not the main drawback of the social protection system in Poland. 14.97 The social protection system in Poland has too few incentives to encourage re- employment. Despite various requirements obliging the unemployed to actively look for ways of improving their financial situation before they are granted social assistance benefits, there is no institutional support in the form of social work with the family, counseling or simple encouragement. Both the labor offices and the social assistance system focus on their mandatory tasks. In the case of the labor offices, these tasks mean mainly payments of benefits. Social assistance support is often directed to a narrow group of individuals. Meanwhile, families affected by long-term unemployment often remain unsupported. 14.98 There is also a fiscal side to the problem. It is possible to discuss the disincentive character of social assistance benefits provided to families stricken by unemployment if this assistance were more widely available. Social security system employees cannot refuse to grant 334 obligatory benefits to any person who is eligible for them. However, owing to the lack of financial resources, they have to define the amount and duration of optional benefits in relation to the budget available after all mandatory benefits have been paid. Optional temporary benefits, which are the basic interim state assistance to poverty-stricken families, have become an insignificant element of social assistance. This is also reflected in the proportion of funds devoted to mandatory tasks (90 percent) and to optional tasks (10 percent of funds and about 35 percent of beneficiaries). By comparison in 1995, the breakdown of funds for these benefits was 45 percent applied to mandatory tasks and 55 percent to optional tasks. 14.99 Older unemployed people at the "immobile" age can apply for pre-retirement benefits (if they meet the required age and work experience criteria). It should be noted that the disability pension system, which provided incomes to a large group of people who had lost their jobs as a result of the transition, played a significant role as an alternative source of income, but now the system is abandoning its role of "income provider." The number of pensioners is decreasing slightly, but the system itself is still an onerous obligation for employees and employers, increasing the tax burden. 14.100 Pre-retirement transfers are becoming the source of income for a growing number of people. In 2002, there were almost half a million beneficiaries receiving pre-retirement benefits and allowances. Expenditures on financial benefits are a growing item on the Labor Fund's list of total expenditures, which, in turn, reduces the availability of funding for active labor market programs for unemployment prevention, in turn reducing the potential availability of help to the unemployed who want to find work. 14.101 The results of the logit analysis show that people who have been unemployed for over a year are more likely to use social transfers and pre-retirement benefits. Hence, these benefits may act as disincentives to job searching. On the other hand, the low availability of unemployment benefits, in particular for people threatened by unemployment traps (older people and mainly women), forces them to apply for pre-retirement benefits, which provide few opportunities for the transition back to employment. Those who can, most frequently use pre-retirement benefits among all available sources of social transfers. As this system is becoming increasingly expensive, it contributes in part to the squeezing of resources allocated to active labor market programs for unemployment prevention. In 2003, the Polish government proposed discontinuing the pre-retirement benefits program after 2006. However, this proposal gained little support in the society. 14.102 Although the disability pension system no longer provides an alternative source of income for the unemployed, it is nevertheless important to consider options for reducing its continuing high costs. Poland spends some 4 percent of GDP on disability pensions, which raises labor costs and makes the country one of the countries with the highest share of indirect costs in total labor costs, which in turn leads to higher unemployment. 335 REFERENCES GUS, Labor Force Survey. Czapinski, Janusz (ed.) (2001), Social Diagnosis 2000, The Council of Social Monitoring, Pedagogical Academy. ZUS (2003), Background Information about Social Insurance System in 2002. The Act of 29 November 1990 on social assistance (full text: Official Journal, No. 64 of 1998, item 414 with later amendments). The Act of 2 July 1994 on renting houses and housing benefits (Official Journal, No. 120 of 1998, item 787 with later amendments). The Act of 1 December 1994 on family, nursing and child-care allowances (Official Journal, No. 102 of 1998, item 651 with later amendments). The Act of 14 December 1994 on employment and counteracting unemployment (full text: Official Journal, No. 6 of 2001, item 56). The Act of 17 December 1998 on old-age, disability and survivors pensions from the Social Insurance Fund (Official Journal, No. 162, item 1118 with later amendments). The Act of 25 June 1999 on cash benefits from social insurance in case of sickness and maternity (Official Journal, No. 60, item 636 with later amendments). 336 Annex 1: Results of the Micro Simulation Because of a complicated formula for housing allowance calculation and the lack of data on individual levels, the authors took an average amount of this benefit. Because temporary benefits are optional and their scope is limited for financial reasons, the authors applied the maximum amount of temporary benefit paid for 2.7 months, which was the average period of such benefit payment in 2002. Table A1.14.1: Replacing the Average Income of a Single Person with Social Transfers, 2002 (in PLN) Employment Unemployment benefits after work loss Assistance after income 80% 100% 120% unemployment benefit expiry Gross income 2,133,.1 386.05 482.56 579.07 - Net income From work 1,466.08 - - - - Unemployment benefits - 353.81 431.98 510.16 - Housing allowances * - 162.00 162.00 162.00 162.00 Temporary social assistance - - 102.22 benefit ** Family income 1,466.08 515.81 593.38 672.16 264.22 Income substitution 35.2 40.5 45.8 7.0 ratio * Country average ** Maximum value Source: Authors' estimates. 337 Table A1.14.2: Replacing the Minimum Income of a Single Person with Social Transfers, 2002 (in PLN) Employment Unemployment benefits after work loss Assistance after income 80% 100% 120% unemployment benefit expiry Gross income 760.00 386.05 482.56 579.07 - Net income From work 561.89 - - - - Unemployment benefits - 353.81 431.98 510.16 - Housing allowances * - 162.00 162.00 162.00 162.00 Temporary social assistance - - - - 102,22 benefit ** Family income 561.89 515.81 593.98 672.16 264.22 Income substitution ratio 91.8 105.7 119.6 47.0 * Country average ** Maximum value Source: Authors' estimates. Table A1.14.3: Replacing the Average Income of a Four-person Family with Social Transfers (in PLN) Employment Unemployment benefits after work loss Assistance after income 80% 100% 120% unemployment benefit expiry Gross income 2,133.21 386.05 482.56 579.07 - Net income From work 1,466.08 - - - - Family benefits 83.92 83.92 83.92 83.92 83.92 Unemployment benefits - 353.81 431.98 510.16 - Housing allowances * 162.00 162.00 162.00 162.00 Temporary social assistance - 115.96 98.37 80.78 195.57 benefit ** Family income 1,550.00 715.69 776.27 836.86 441.49 Income substitution ratio 46.2 50.1 54.0 28.5 * Country average ** Maximum value Source: Authors' estimates. 338 Table A1.14.4: Replacing the Minimum Income of a Four-person Family with Social Transfers (in PLN) Employment Unemployment benefits after work loss Assistance after income 80% 100% 120% unemployment benefit expiry Gross income 760.00 386.05 482.56 579.07 - Net income From work 561.89 - - - - Family benefits 83.92 83.92 83.92 83.92 83.92 Unemployment benefits - 353.81 431.98 510.16 - Housing allowances * 162.00 162.00 162.00 162.00 162.00 Temporary social assistance 69.14 115.96 98.37 80.78 195.57 benefit ** Family income 876.95 715.69 776.27 836.86 441.49 Income substitution ratio 81.6 88.5 95.4 50.3 * Country average ** Maximum value Source: Authors' estimates. Table A1.14.5: Replacing the Average Income of a Family with Four Children with Social Transfers (in PLN) Assistance after Employment Unemployment benefits after work loss unemployment income benefit expiry 80% 100% 120% Gross income 2,133.21 386.05 482.56 579.07 - Net income From work 1,466.08 - - - - Family benefits 200.67 200.67 200.67 200.67 200.67 Unemployment benefits - 353.81 431.98 510.16 - Housing allowances * 162.00 162.00 162.00 162.00 162.00 Temporary social assistance - 182.61 165.03 147.44 262.22 benefit ** Family income 1,828.75 899.09 959.68 1,020.27 624.89 Income substitution ratio 49.2 52.5 55.8 34.2 * Country average ** Maximum value Source: Authors' estimates. 339 Table A1.14.6: Replacing the Minimum Income of a Family with Four Children with Social Transfers (in PLN) Employment Unemployment benefits after work loss Assistance after income 80% 100% 120% unemployment benefit expiry Gross income 760.00 386.05 482.56 579.07 - Net income From work 561.89 - - - - Family benefits 200.97 200.97 200.97 200.97 200,97 Unemployment benefits - 353.81 431.98 510.16 .- Housing allowances* 162.00 162.00 162.00 162.00 162.00 Temporary social assistance 135.80 182.61 165.03 147.44 262.22 benefit ** Family income 1,060.36 899.09 959.68 1,020.27 624.89 Income substitution ratio 84.8 90.5 96.2 58.9 * Country average ** Maximum value Source: Authors' estimates. Table A1.14.7: Replacing the Average Income of a Single Parent with Two Children with Social Transfers (in PLN) Employment Unemployment benefits after work loss Assistance after income 80% 100% 120% unemployment benefit expiry (***) Gross income 2,133.21 386.05 482.56 579.07 - Net income From work 1,466.08 - - - - Family benefits 83.92 83.92 83.92 83.92 Unemployment benefits - 353.81 431.98 510.16 - Housing allowances * 162.00 162.00 162.00 162.00 Temporary guaranteed (330.40) 413.00 benefit Temporary social assistance - 51.91 33.32 18.58 (56.18) 37.59 benefit ** Family income 1,466.08 650.64 711.22 774.66 (632.50) 696.51 Income substitution ratio 44.4 48.5 52.8 (43.1) 47.5 * Country average ** Maximum value (***) Values for the second and third years of the guaranteed temporary benefit are in parentheses. Source: Authors' estimates. 340 Table A1.14.8: Replacing the Minimum Income of a Single Parent with Two Children with Social Transfers (in PLN) Employment Unemployment benefits after work loss Assistance after income 80% 100% 120% unemployment benefit expiry (***) Gross income 760.00 386.05 482.56 579.07 - Net income From work 561.89 - - - - Family benefits 83.92 83.92 83.92 83.92 83.92 Unemployment benefits - 353.81 431.98 510.16 - Housing allowances * 162.00 162.00 162.00 162.00 162.00 Guaranteed temporary (330.40) 413.00 benefit Temporary social assistance 18.58 51.91 33.32 18.58 (56.18) 37.59 benefit ** Family income 826.39 650.64 711.22 774.66 (632.50) 696.51 Income substitution ratio 78.7 86.1 93.7 (76.5) 84.3 * Country average ** Maximum value (***)Values for the second and third years of the guaranteed temporary benefit are in parentheses. Source: Authors' estimates. 341 Annex 2: Results of the Logit Regression 14.103 This annex presents detailed results of a logit regression aimed at checking the impact of the current and past labor market position of an individual, as well as of his/her other individual characteristics with respect to the likelihood of supporting himself/herself from a disability pension or other non-earning source (excluding unemployment benefits, retirement benefits, early retirement benefits, or disability pension). The research is based on the results of the LFS conducted in the fourth quarter of 2002. Two defined variables were adopted. The first variable: D1: is a binary variable taking the value of 1 if the main source of income is the "income from other non-earning sources." Otherwise its value is zero. In the research sample, 1,955 people (out of 47,095 respondents), namely 4.15 percent, declared other non-earning sources of income as their main means of support. The second variable: D2: is also a binary variable, taking the value of 1, if the main source of income is a disability pension. Otherwise its value is zero. In the research sample, 5,175 people, that is, 10.99 percent of respondents, declared a disability benefit as their main source of income. Explanatory variables were based on individual characteristics, such as sex, education, age, and whether the respondent was unemployed 12 months earlier or during the time the survey was conducted. Sex_w woman Edu_high secondary school education Edu_univ higher education Un12 binary variable, taking the value of "1" if the respondent declared unemployment 12 months before the survey. Otherwise the variable takes the value of "0". Un binary variable, taking the value of "1", if the respondent was unemployed at the time when the survey was conducted Age1 age (from 15 to 24 years of age) Age3 age (from 45 to 65 years of age). 342 Table A2.14.1 presents the estimate for a non-earning sources of income model, and Table A2.14.2 presents the estimate for a disability benefit model. Table A2.14.1: Supporting Oneself from Non-earning Sources of Income (logit estimates) Logit Estimates Number of obs = 47094 chi2(7) =1629.30 Prob > chi2 = 0.0000 Log Likelihood = -7319.5316 Pseudo R2 = 0.1002 D1 Odds Ratio Std. Err. Z P>|z| [95% Conf. Interval] Sex w 1.44 0.07 7.54 0.000 1.31 1.59 Edu high 0.88 0.05 -2.43 0.015 0.79 0.97 Edu univ 0.40 0.04 -8.21 0.000 0.32 0.49 Un12 3.17 0.22 16.93 0.000 2.77 3.62 Un 2.74 0.19 14.66 0.000 2.39 3.13 Age1 0.89 0.06 -1.61 0.107 0.78 1.03 Age3 2.23 0.12 15.48 0.000 2.01 2.47 Source: Authors' estimates. Table A2.14.2: Supporting Oneself from a Disability Pension (logit estimates) 14.104 Logit Estimates Number of obs = 47094 chi2(7) =3269.37 Prob > chi2 = 0.0000 Log Likelihood = -14672.939 Pseudo R2 = 0.1002 D2 Odds Ratio Std. Err. Z P>|z| [95% Conf. Interval] Sex w 1.49 0.05 12.88 0.000 1.40 1.59 Edu high 0.40 0.02 -21.34 0.000 0.40 0.47 Edu univ 0.13 0.01 -23.55 0.000 0.11 0.15 Un12 0.17 0.02 -15.64 0.000 0.14 0.21 Un 0.65 0.05 -5.13 0.000 0.56 0.77 Age1 0.27 0.02 -22.36 0.000 0.20 0.26 Age3 1.88 0.06 19.81 0.000 1.77 2.00 Source: Authors' estimates. 343 STATISTICALAPPENDIX 344 Table A1. Regional deflators Voivodship/ residence /hhld category Mean Nof hhlds St Dev By voivodship DOLNOLSKIE 1.0080 2,541 0.0352 KUJAWSKO-POMORSKIE 0.9698 1,775 0.0363 LUBELSKIE 0.9611 1,866 0.0449 LUBUSKIE 1.0069 850 0.0292 LÓDZKIE 0.9769 2,478 0.0369 MALOPOLSKIE 1.0241 2,495 0.0427 MAZOWIECKIE 1.0433 4,282 0.0617 OPOLSKIE 1.0002 902 0.0281 PODKARPACKIE 0.9542 1,568 0.0275 PODLASKIE 0.9560 1,042 0.0389 POMORSKIE 1.0268 1,725 0.0450 LSKIE 1.0135 4,098 0.0258 WITOKRZYSKIE 0.9714 1,080 0.0434 WARMISKO-MAZURSKIE 0.9743 1,129 0.0302 WIELKOPOLSKIE 1.0002 2,597 0.0468 ZACHODNIOPOMORSKIE 1.0037 1,420 0.0412 By place of residence Urban area 1.0211 20,615 0.0473 Town 500T + 1.0824 4,196 0.0420 Town 200T - 500T 1.0450 3,637 0.0282 Town 100T - 200T 1.0204 2,530 0.0239 Town 20T - 100T 0.9921 6,296 0.0207 Town - 20T 0.9805 3,957 0.0224 Rural area 0.9629 11,232 0.0279 By socio-economic category* of household Worker/employee 1.0123 12,494 0.0494 Farmer/worker 0.9637 2,525 0.0310 Farmer 0.9545 1,335 0.0318 Self employed 1.0130 2,056 0.0503 Pensioner (retiree) 1.0025 7,485 0.0502 Pensioner (disable/ survivor) 0.9933 4,331 0.0445 Living on soc income/ part time jobs 1.0000 1,620 0.0493 TOTAL 1.0006 31,847 0.0499 * According to the main source of income Source: HBS 2001, WB staff computation [MS]. 345 Table A2. Consumption composition by quartile and place of residence ALL HOUSEHOLDS Ex post consumption* quartile group Consumption components* All households 1 2 3 4 PLN/month, per household Food 529.64 582.09 613.20 680.46 601.35 of which: in kind food 85.74 77.51 60.62 49.80 68.42 Clothing and footwear 37.56 68.46 100.81 184.74 97.89 Energy and housing utilities 137.42 194.24 239.08 335.71 226.61 Health 38.45 65.21 92.86 147.18 85.92 Transportation and communication 96.74 161.50 210.80 350.09 204.78 Education, culture and recreation 44.36 73.73 100.83 192.96 102.97 Imputed rent 107.12 117.02 123.05 134.68 120.47 Imputed cons of durables 60.91 83.92 109.28 159.85 103.49 Total consumption (WB2) 1,163.98 1,493.55 1,782.72 2,527.27 1,741.87 Percent [WB2 = 100] Food 45.5 39.0 34.4 26.9 34.5 of which: in kind food 7.4 5.2 3.4 2.0 3.9 Clothing and footwear 3.2 4.6 5.7 7.3 5.6 Energy and housing utilities 11.8 13.0 13.4 13.3 13.0 Health 3.3 4.4 5.2 5.8 4.9 Transportation and communication 8.3 10.8 11.8 13.9 11.8 Education, culture and recreation 3.8 4.9 5.7 7.6 5.9 Imputed rent 9.2 7.8 6.9 5.3 6.9 Imputed cons of durables 5.2 5.6 6.1 6.3 5.9 Total consumption (WB2) 100.0 100.0 100.0 100.0 100.0 Memo: WB2 per hhld, PLN/month 1,163.98 1,493.55 1,782.72 2,527.27 1,741.87 * Quartile groups are for households, set for equivalent WB2. URBAN HOUSEHOLDS Ex post consumption* quartile group Consumption components* All households 1 2 3 4 PLN/month, per household Food 467.13 525.40 567.58 650.34 560.71 of which: in kind food 26.36 24.68 23.12 22.66 24.02 Clothing and footwear 34.55 62.81 95.21 185.39 100.96 Energy and housing utilities 166.21 216.64 245.89 307.90 240.40 Health 34.22 60.89 90.18 149.85 88.86 Transportation and communication 87.79 150.41 203.01 359.35 211.79 Education, culture and recreation 96.74 96.74 96.74 96.74 96.74 Imputed rent 104.10 112.44 118.63 130.29 117.53 Imputed cons of durables 54.03 76.75 102.52 156.32 101.89 Total consumption (WB2) 1,107.74 1,431.97 1,722.53 2,499.10 1,750.40 Percent [WB2 = 100] Food 42.2 36.7 33.0 26.0 32.0 of which: in kind food 2.4 1.7 1.3 0.9 1.4 Clothing and footwear 3.1 4.4 5.5 7.4 5.8 Energy and housing utilities 15.0 15.1 14.3 12.3 13.7 Health 3.1 4.3 5.2 6.0 5.1 Transportation and communication 7.9 10.5 11.8 14.4 12.1 Education, culture and recreation 8.7 6.8 5.6 3.9 5.5 Imputed rent 9.4 7.9 6.9 5.2 6.7 Imputed cons of durables 4.9 5.4 6.0 6.3 5.8 Total consumption (WB2) 100.0 100.0 100.0 100.0 100.0 346 Memo: WB2 per hhld, PLN/month 1,107.74 1,431.97 1,722.53 2,499.10 1,750.40 * Quartile groups are for households, set for equivalent WB2. RURAL HOUSEHOLDS Ex post consumption* quartile group Consumption components* All households 1 2 3 4 PLN/month, per household Food 600.01 675.91 715.22 768.66 675.95 of which: in kind food 152.59 164.92 144.48 129.28 149.90 Clothing and footwear 40.96 77.79 113.32 182.83 92.25 Energy and housing utilities 105.02 157.17 223.86 417.12 201.30 Health 43.21 72.35 98.86 139.36 80.53 Transportation and communication 106.82 179.86 228.23 322.96 191.92 Education, culture and recreation 96.74 96.74 96.74 96.74 96.74 Imputed rent 110.53 124.61 132.92 147.53 125.87 Imputed cons of durables 68.66 95.79 124.40 170.18 106.43 Total consumption (WB2) 1,227.29 1,595.46 1,917.31 2,609.74 1,726.22 Percent [WB2 = 100] Food 48.9 42.4 37.3 29.5 39.2 of which: in kind food 12.4 10.3 7.5 5.0 8.7 Clothing and footwear 3.3 4.9 5.9 7.0 5.3 Energy and housing utilities 8.6 9.9 11.7 16.0 11.7 Health 3.5 4.5 5.2 5.3 4.7 Transportation and communication 8.7 11.3 11.9 12.4 11.1 Education, culture and recreation 7.9 6.1 5.0 3.7 5.6 Imputed rent 9.0 7.8 6.9 5.7 7.3 Imputed cons of durables 5.6 6.0 6.5 6.5 6.2 Total consumption (WB2) 100.0 100.0 100.0 100.0 100.0 Memo: WB2 per hhld, PLN/month 1,227.29 1,595.46 1,917.31 2,609.74 1,726.22 * Specified components sum up to WB0 aggregate, but some items (such as alcohol, tobacco, restaurants, not specified) are not displayed. ** Quartile groups are for households, set for equivalent WB2. Source: HBS 2001, own computation. 347 Table A3. Poverty according to various LS concepts: overview "Medium Poverty" LS* concept Poverty measures GUS GUS reg WB0 WB2 Poverty line [PLN/month] ---> 389 389 349 404 Headcounts [%] persons H 15.17 14.24 14.84 14.24 households Hh 11.27 10.61 11.23 10.22 Poverty depth [%] P1 3.35 3.09 3.02 2.72 Poverty gap [%] PG 22.09 21.68 20.35 19.11 Poverty deficit [%] PD 2.38 2.19 2.15 1.92 Macro estimates** Population in poverty [mln] 5.86 5.50 5.73 5.50 Total gap as % GDP 0.57 0.53 0.46 0.48 "Hard Poverty" LS* concept Poverty measures GUS GUS reg WB0 WB2 Poverty line [PLN/month] ---> 336 336 302 355 Headcounts [%] persons H 9.60 9.01 8.86 8.60 households Hh 6.92 6.48 6.54 5.94 Poverty depth [%] P1 1.94 1.76 1.66 1.53 Poverty gap [%] PG 20.24 19.57 18.70 17.86 Poverty deficit [%] PD 1.37 1.25 1.17 1.07 Macro estimates** Population in poverty [mln] 3.71 3.48 3.43 3.33 Total gap as % GDP 0.29 0.26 0.22 0.24 "Extreme Poverty" LS* concept Poverty measures GUS GUS reg WB0 WB2 Poverty line [PLN/month] ---> 263 263 240 283 Headcounts [%] persons H 3.81 3.45 3.56 3.22 households Hh 2.65 2.37 2.47 2.05 Poverty depth [%] P1 0.67 0.60 0.56 0.48 Poverty gap [%] PG 17.59 17.30 15.61 14.95 Poverty deficit [%] PD 0.47 0.42 0.39 0.33 Macro estimates** Population in poverty [mln] 1.47 1.33 1.38 1.25 Total gap as % GDP 0.08 0.07 0.06 0.06 Notes: * LS = Living Standard. For methodology - see Section1 (main text) and Methodological Appendix. ** Estimates, based on H and PD, and : population = million 38.641; GDP = PLN million 749311. PovLines in constant 1998 PPP$/month Medium poverty 190 190 171 198 Hard poverty 165 165 148 174 Extreme poverty 129 129 118 139 Exchange rate: 2.042 WB Staff [MS} Source: HBS 2001, own computation. 348 Table A4. Poverty by voivodship Hard poverty Medium poverty Voivodship GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Poverty line [PLN/month] 336 336 302 355 389 389 349 404 Headcount [%] DOLNOLSKIE 11.4 11.5 12.1 11.3 17.4 17.4 18.6 17.6 KUJAWSKO-POMORSKIE 10.1 8.5 9.2 8.9 15.7 14.4 15.4 15.9 LUBELSKIE 14.1 11.5 10.7 10.6 21.0 17.6 18.0 16.9 LUBUSKIE 8.1 8.3 8.3 8.1 13.8 13.8 15.2 15.4 LÓDZKIE 7.5 6.6 6.5 6.9 12.9 11.1 10.7 11.2 MALOPOLSKIE 7.9 8.3 6.9 7.0 13.4 13.9 13.8 14.1 MAZOWIECKIE 6.7 6.6 6.1 5.9 11.5 11.5 10.7 10.6 OPOLSKIE 8.5 7.9 7.6 6.7 12.2 11.0 11.7 11.0 PODKARPACKIE 11.9 10.1 9.1 8.5 20.2 17.0 16.2 15.7 PODLASKIE 9.3 7.2 7.6 7.5 14.7 11.9 14.1 12.2 POMORSKIE 12.4 12.8 13.7 13.7 19.1 18.5 19.9 18.6 LSKIE 6.8 6.8 7.7 6.7 10.6 10.6 12.7 11.0 WITOKRZYSKIE 14.3 12.0 9.7 10.6 19.6 18.0 16.6 16.7 WARMISKO-MAZURSKIE 14.0 12.9 13.0 13.5 21.0 18.9 20.3 19.4 WIELKOPOLSKIE 9.1 8.9 8.4 8.2 14.6 13.6 14.5 13.5 ZACHODNIOPOMORSKIE 11.4 11.1 11.7 11.1 17.6 17.3 19.0 18.1 POLAND -TOTAL 9.6 9.0 8.9 8.6 15.2 14.2 14.8 14.2 Relative headcount DOLNOLSKIE 1.19 1.28 1.37 1.31 1.15 1.22 1.25 1.23 KUJAWSKO-POMORSKIE 1.06 0.94 1.04 1.04 1.04 1.01 1.03 1.12 LUBELSKIE 1.47 1.27 1.21 1.23 1.39 1.24 1.21 1.19 LUBUSKIE 0.85 0.93 0.94 0.94 0.91 0.97 1.03 1.08 LÓDZKIE 0.78 0.73 0.73 0.80 0.85 0.78 0.72 0.79 MALOPOLSKIE 0.82 0.92 0.77 0.81 0.88 0.98 0.93 0.99 MAZOWIECKIE 0.70 0.74 0.69 0.68 0.76 0.81 0.72 0.74 OPOLSKIE 0.89 0.88 0.85 0.77 0.80 0.77 0.79 0.78 PODKARPACKIE 1.24 1.12 1.02 0.99 1.33 1.19 1.09 1.10 PODLASKIE 0.97 0.80 0.86 0.87 0.97 0.84 0.95 0.86 POMORSKIE 1.29 1.42 1.55 1.59 1.26 1.30 1.34 1.30 LSKIE 0.71 0.76 0.86 0.78 0.70 0.75 0.86 0.78 WITOKRZYSKIE 1.49 1.33 1.09 1.23 1.29 1.26 1.12 1.18 WARMISKO-MAZURSKIE 1.45 1.43 1.47 1.56 1.38 1.33 1.37 1.36 WIELKOPOLSKIE 0.95 0.98 0.95 0.95 0.96 0.96 0.98 0.95 ZACHODNIOPOMORSKIE 1.19 1.23 1.32 1.29 1.16 1.21 1.28 1.27 POLAND -TOTAL 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Source: HBS 2001, own computation. 349 Table A5. Poverty by place of residence Hard poverty Medium poverty Place of residence GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Poverty Line (PLN/month) 336 336 302 355 389 389 349 404 Headcount [%] Urban 5.93 6.06 6.70 6.51 9.85 9.99 11.60 11.22 500 000+ 2.13 2.77 3.28 3.12 3.59 4.74 6.42 5.97 200 - 500 000 4.27 4.91 5.69 5.48 7.36 8.18 9.76 9.37 100 - 200 000 6.35 6.34 7.41 7.04 10.34 10.66 12.79 12.31 20 - 100 000 5.90 5.91 6.37 6.31 9.90 9.74 11.62 11.30 20 000 - 10.31 9.86 10.51 10.21 16.82 15.85 16.75 16.40 Rural 15.16 13.48 12.14 11.76 23.24 20.70 19.76 18.83 TOTAL 9.60 9.01 8.86 8.60 15.17 14.24 14.84 14.24 Relative to TOTAL Urban 0.62 0.67 0.76 0.76 0.65 0.70 0.78 0.79 500 000+ 0.22 0.31 0.37 0.36 0.24 0.33 0.43 0.42 200 - 500 000 0.44 0.55 0.64 0.64 0.49 0.57 0.66 0.66 100 - 200 000 0.66 0.70 0.84 0.82 0.68 0.75 0.86 0.86 20 - 100 000 0.61 0.66 0.72 0.73 0.65 0.68 0.78 0.79 20 000 - 1.07 1.09 1.19 1.19 1.11 1.11 1.13 1.15 Rural 1.58 1.50 1.37 1.37 1.53 1.45 1.33 1.32 TOTAL 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Relative to urban Urban 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 500 000+ 0.36 0.46 0.49 0.48 0.36 0.47 0.55 0.53 200 - 500 000 0.72 0.81 0.85 0.84 0.75 0.82 0.84 0.84 100 - 200 000 1.07 1.05 1.11 1.08 1.05 1.07 1.10 1.10 20 - 100 000 0.99 0.97 0.95 0.97 1.00 0.98 1.00 1.01 20 000 - 1.74 1.63 1.57 1.57 1.71 1.59 1.44 1.46 Rural 2.55 2.22 1.81 1.81 2.36 2.07 1.70 1.68 Source: HBS 2001, own computation. 350 Table A6. Rural vs. urban poverty by voivodship WB0 poverty headcounts [%] Sorted by 'Rural/Urban' for medium poverty Hard poverty Medium poverty Rural Voivodship population Rural Urban Rural/ Rural Urban Rural/ share [%] areas areas Urban areas areas Urban POMORSKIE 31.8 22.7 9.5 2.37 31.4 14.5 2.16 MAZOWIECKIE 36.8 10.0 3.8 2.62 16.1 7.6 2.12 OPOLSKIE 48.5 9.9 5.4 1.84 15.6 8.0 1.96 LUBUSKIE 37.0 13.0 5.6 2.32 21.8 11.4 1.92 LUBELSKIE 55.7 14.5 6.0 2.42 22.8 11.9 1.91 ZACHODNIOPOMORSKIE 32.6 20.1 7.6 2.64 27.7 14.8 1.88 MALOPOLSKIE 49.6 9.1 4.7 1.95 17.8 9.9 1.80 DOLNOLSKIE 30.1 19.4 9.0 2.15 26.9 15.0 1.79 KUJAWSKO-POMORSKIE 38.0 14.2 6.2 2.28 20.9 12.0 1.74 WARMISKO-MAZURSKIE 42.1 17.2 10.0 1.71 26.8 15.6 1.72 PODKARPACKIE 62.6 10.1 7.2 1.40 19.1 11.2 1.70 WIELKOPOLSKIE 44.0 10.7 6.7 1.59 18.4 11.4 1.61 PODLASKIE 44.2 8.6 6.9 1.24 17.1 11.7 1.46 LÓDZKIE 36.2 8.1 5.6 1.44 13.3 9.2 1.46 LSKIE 20.8 8.5 7.4 1.14 15.4 12.0 1.28 WITOKRZYSKIE 57.7 10.2 8.9 1.14 18.2 14.6 1.25 TOTAL 39.7 12.1 6.7 1.81 19.8 11.6 1.70 WB2 poverty headcounts [%] Sorted by 'Rural/Urban' for medium poverty Rural Hard poverty Medium poverty Voivodship population Rural Urban Rural/ Rural Urban Rural/ share [%] areas areas Urban areas areas Urban MAZOWIECKIE 36.8 9.3 3.9 2.42 16.3 7.2 2.26 POMORSKIE 31.8 22.7 9.5 2.39 28.7 13.9 2.07 LUBELSKIE 55.7 13.6 6.8 1.99 21.8 10.7 2.04 LUBUSKIE 37.0 13.7 4.8 2.84 22.2 11.4 1.95 ZACHODNIOPOMORSKIE 32.6 18.5 7.5 2.47 26.9 13.8 1.95 MALOPOLSKIE 49.6 9.3 4.7 1.98 18.5 9.8 1.89 OPOLSKIE 48.5 8.1 5.3 1.54 14.3 8.0 1.78 DOLNOLSKIE 30.1 17.8 8.5 2.10 25.3 14.2 1.77 PODKARPACKIE 62.6 9.6 6.6 1.47 18.1 11.5 1.58 KUJAWSKO-POMORSKIE 38.0 13.4 6.2 2.17 20.5 13.1 1.56 WARMISKO-MAZURSKIE 42.1 17.2 10.7 1.60 24.5 15.7 1.55 WIELKOPOLSKIE 44.0 10.9 6.1 1.80 16.6 11.0 1.50 LÓDZKIE 36.2 8.3 6.1 1.35 14.2 9.5 1.49 PODLASKIE 44.2 8.8 6.5 1.37 14.2 10.6 1.34 WITOKRZYSKIE 57.7 12.0 8.7 1.38 18.0 15.0 1.20 LSKIE 20.8 6.8 6.7 1.02 11.4 11.0 1.04 TOTAL 39.7 11.8 6.5 1.81 18.8 11.2 1.68 Source: HBS 2001, own computation. 351 Table A7. Poverty by socio-economic category of a household Socio-economic category of Hard poverty Medium poverty household GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Poverty Line (PLN/month) ------> 336 336 302 355 389 389 349 404 Headcount [%] Worker/Employee 7.31 7.17 7.63 7.51 11.75 11.40 12.87 12.55 Worker-farmer 12.11 10.81 9.36 9.06 20.43 17.79 16.74 15.95 Farmer 13.34 10.58 8.76 7.57 23.48 19.03 16.33 14.70 Self-employed 4.91 4.87 5.19 4.34 8.62 8.72 8.52 8.00 Retiree 5.46 4.98 5.13 4.79 8.68 8.10 9.10 8.26 Disable / Survivor 14.38 13.40 12.79 12.24 21.31 20.55 20.97 20.03 Living on soc transf and part-time 29.69 28.96 27.71 28.96 41.21 40.04 40.79 40.71 TOTAL 9.60 9.01 8.86 8.60 15.17 14.24 14.84 14.24 Relative to TOTAL Worker/Employee 0.76 0.80 0.86 0.87 0.77 0.80 0.87 0.88 Worker-farmer 1.26 1.20 1.06 1.05 1.35 1.25 1.13 1.12 Farmer 1.39 1.17 0.99 0.88 1.55 1.34 1.10 1.03 Self-employed 0.51 0.54 0.59 0.50 0.57 0.61 0.57 0.56 Retiree 0.57 0.55 0.58 0.56 0.57 0.57 0.61 0.58 Disable / Survivor 1.50 1.49 1.44 1.42 1.40 1.44 1.41 1.41 Living on soc transf and part-time 3.09 3.21 3.13 3.37 2.72 2.81 2.75 2.86 TOTAL 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Relative to Worker Worker/Employee 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Worker-farmer 1.66 1.51 1.23 1.21 1.74 1.56 1.30 1.27 Farmer 1.83 1.48 1.15 1.01 2.00 1.67 1.27 1.17 Self-employed 0.67 0.68 0.68 0.58 0.73 0.76 0.66 0.64 Disable / Survivor 0.75 0.69 0.67 0.64 0.74 0.71 0.71 0.66 Disable 1.97 1.87 1.68 1.63 1.81 1.80 1.63 1.60 Living on soc transf and part-time 4.06 4.04 3.63 3.86 3.51 3.51 3.17 3.24 Source: HBS 2001, own computation. 352 Table A8. Poverty by land area Hard poverty Medium poverty Land area GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Poverty Line (PLN/month) 336 336 302 355 389 389 349 404 Headcount [%] 0.00 - 0.09 8.14 8.00 8.39 8.25 12.75 12.47 13.72 13.32 0.10 - 1.00 12.98 11.56 10.28 8.94 20.17 18.67 17.43 16.14 All non-farm households 8.66 8.38 8.60 8.32 13.55 13.14 14.12 13.63 1.01 - 1.99 15.08 13.13 10.96 10.95 22.74 20.49 19.61 20.17 2.00 - 4.99 16.38 14.43 11.11 11.57 25.67 22.34 20.59 19.70 5.00 - 6.99 16.42 13.08 11.58 11.28 26.39 22.24 18.79 17.49 7.00 - 9.99 16.23 13.73 11.61 10.69 25.63 21.43 19.19 17.84 10.00 - 14.99 9.94 7.80 8.82 8.48 20.21 14.05 15.76 13.33 15.00 + 7.18 6.05 5.12 3.70 13.95 12.51 11.93 9.64 All farm households 14.06 11.95 10.10 9.84 22.97 19.51 18.21 17.10 TOTAL 9.60 9.01 8.86 8.60 15.17 14.24 14.84 14.24 Relative to TOTAL 0.00 - 0.09 0.85 0.89 0.95 0.96 0.84 0.88 0.92 0.94 0.10 - 1.00 1.35 1.28 1.16 1.04 1.33 1.31 1.17 1.13 All non-farm households 0.90 0.93 0.97 0.97 0.89 0.92 0.95 0.96 1.01 - 1.99 1.57 1.46 1.24 1.27 1.50 1.44 1.32 1.42 2.00 - 4.99 1.71 1.60 1.25 1.35 1.69 1.57 1.39 1.38 5.00 - 6.99 1.71 1.45 1.31 1.31 1.74 1.56 1.27 1.23 7.00 - 9.99 1.69 1.52 1.31 1.24 1.69 1.50 1.29 1.25 10.00 - 14.99 1.04 0.87 1.00 0.99 1.33 0.99 1.06 0.94 15.00 + 0.75 0.67 0.58 0.43 0.92 0.88 0.80 0.68 All farm households 1.46 1.33 1.14 1.15 1.51 1.37 1.23 1.20 TOTAL 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Notes: aTotal agricultural land area over 1 ha is treated as a 'farm' in GUS and IERiG statistics. Grouping - according to GUS and IERiG. Source: HBS 2001, own computation. 353 Table A9. Poverty by the employment status of a household head Employment status of Hard poverty Medium poverty a household head GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Poverty Line (PLN/month) 336 336 302 355 389 389 349 404 Headcount [%] Employed* 7.52 6.99 6.93 6.66 12.87 11.84 12.41 11.83 Unemployed* 37.89 36.52 35.51 36.56 48.52 47.50 49.43 48.28 Other* 11.97 11.32 11.04 10.68 17.64 16.93 17.49 16.90 TOTAL 9.60 9.01 8.86 8.60 15.17 14.24 14.84 14.24 Relative to TOTAL Employed* 0.78 0.78 0.78 0.77 0.85 0.83 0.84 0.83 Unemployed* 3.95 4.05 4.01 4.25 3.20 3.34 3.33 3.39 Other* 1.25 1.26 1.25 1.24 1.16 1.19 1.18 1.19 TOTAL 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Relative to employed Employed* 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Unemployed* 5.04 5.23 5.13 5.49 3.77 4.01 3.98 4.08 Other* 1.59 1.62 1.59 1.60 1.37 1.43 1.41 1.43 Notes: * Employed ­ has a permanent job (including farmers, self-employed and professionals). * Unemployed - does not have a job (over the last week), and is looking for a job or is ready to take a job. * Other - lives on social transfers, on part time etc.(but does not enter into 'unemployed' category). Source: HBS 2001, own computation. 354 Table A10. Poverty by unemployment status of household members Number of unemployed in a Hard poverty Medium poverty household GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Poverty Line (PLN/month) 336 336 302 355 389 389 349 404 Headcount [%] Nof unempl = 0 6.86 6.29 6.17 5.69 11.37 10.48 10.90 10.33 Nof unempl = 1 or more 21.07 20.39 20.14 20.78 30.87 29.79 31.13 30.39 of which: 1 unemployed 16.70 16.29 16.47 16.47 25.81 24.78 26.33 25.13 2 or more unemployed 38.49 36.75 34.78 37.99 50.86 49.55 50.05 51.21 TOTAL 9.60 9.01 8.86 8.60 15.17 14.24 14.84 14.24 Relative to TOTAL Nof unempl = 0 0.72 0.70 0.70 0.66 0.75 0.74 0.73 0.73 Nof unempl = 1 or more 2.19 2.26 2.27 2.42 2.03 2.09 2.10 2.13 of which: 1 unemployed 1.74 1.81 1.86 1.92 1.70 1.74 1.77 1.76 2 or more unemployed 4.01 4.08 3.92 4.42 3.35 3.48 3.37 3.60 TOTAL 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Relative to households without unemployed Nof unempl = 0 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Nof unempl = 1 or more 3.07 3.24 3.26 3.65 2.71 2.84 2.86 2.94 of which: 1 unemployed 2.43 2.59 2.67 2.89 2.27 2.36 2.42 2.43 2 or more unemployed 5.61 5.84 5.63 6.67 4.47 4.73 4.59 4.96 Notes: Unemployed - does not have a job (over the last week) and is ready to take a job. Source: HBS 2001, own computation. 355 Table A11. Poverty by employment status of household members Employment of household Hard poverty Medium poverty members GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Poverty Line (PLN/month) 336 336 302 355 389 389 349 404 Headcount [%] Nof employed = 0 13.26 12.96 12.99 12.50 19.18 18.61 19.37 18.74 Nof employed = 1or more 8.40 8.40 8.40 8.40 13.87 12.82 13.36 12.77 of which: 1 employed 10.11 9.48 9.59 9.46 16.38 15.73 16.62 15.89 2 or more employed 7.08 6.35 5.91 5.67 11.92 10.56 10.84 10.36 2 working 6.22 5.56 5.14 4.78 10.55 9.41 9.77 9.19 3+ working 10.94 9.88 9.35 9.61 18.00 15.66 15.59 15.53 TOTAL 9.60 9.01 8.86 8.60 15.17 14.24 14.84 14.24 Relative to TOTAL Nof employed = 0 1.38 1.44 1.47 1.45 1.26 1.31 1.31 1.32 Nof employed = 1or more 0.88 0.93 0.95 0.98 0.91 0.90 0.90 0.90 of which: 1 employed 1.05 1.05 1.08 1.10 1.08 1.10 1.12 1.12 2 or more employed 0.74 0.70 0.67 0.66 0.79 0.74 0.73 0.73 2 working 0.65 0.62 0.58 0.56 0.70 0.66 0.66 0.65 3+ working 1.14 1.10 1.05 1.12 1.19 1.10 1.05 1.09 TOTAL 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Relative to households without employed Nof employed = 0 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Nof employed = 1or more 0.63 0.65 0.65 0.67 0.72 0.69 0.69 0.68 of which: 1 employed 0.76 0.73 0.74 0.76 0.85 0.85 0.86 0.85 2 or more employed 0.53 0.49 0.46 0.45 0.62 0.57 0.56 0.55 2 working 0.47 0.43 0.40 0.38 3+ working 0.82 0.76 0.72 0.77 Notes: Employed ­ has a permanent job (including farmers, self-employed professionals). Source: HBS 2001, own computation. 356 Table A12. Poverty by disability status of household members Disability of household Hard poverty Medium poverty members GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Poverty Line (PLN/month) 336 336 302 355 389 389 349 404 Headcount [%] Hhld without the disable 9.09 8.59 8.57 8.16 14.15 13.40 13.97 13.22 Hhld with the disable 10.72 9.95 9.50 9.55 17.44 16.12 16.77 16.50 of which: 1 disable person 10.84 10.10 9.67 9.60 17.47 16.18 16.92 16.42 2 or more disable persons 10.37 9.49 8.98 9.42 17.35 15.93 16.30 16.75 TOTAL 9.60 9.01 8.86 8.60 15.17 14.24 14.84 14.24 Relative to TOTAL Hhld without the disable 0.95 0.95 0.97 0.95 0.93 0.94 0.94 0.93 Hhld with the disable 1.12 1.10 1.07 1.11 1.15 1.13 1.13 1.16 of which: 1 disable person 1.13 1.12 1.09 1.12 1.15 1.14 1.14 1.15 2 or more disable persons 1.08 1.05 1.01 1.10 1.14 1.12 1.10 1.18 TOTAL 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Relative to households without the disabled Hhld without the disable 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Hhld with the disable 1.18 1.16 1.11 1.17 1.23 1.20 1.20 1.25 of which: 1 disable person 1.19 1.18 1.13 1.18 1.23 1.21 1.21 1.24 2 or more disable persons 1.14 1.10 1.05 1.15 1.23 1.19 1.17 1.27 Notes: Disable: a person 15+ declaring diability status. Source: HBS 2001, own computation. 357 Table A13. Poverty by family type Hard poverty Medium poverty Family type GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Poverty Line (PLN/month) 336 336 302 355 389 389 349 404 Headcount [%] Single person 3.64 3.35 3.91 2.26 5.76 5.51 6.46 4.51 Couple, 0 childrenb 1.88 1.76 1.92 1.41 3.69 3.31 3.68 2.76 Couple, 1 child 4.16 4.04 4.41 3.70 6.86 6.48 6.96 6.17 Couple, 2 children 6.94 6.54 6.70 5.61 11.85 11.52 12.19 10.99 Couple, 3 children 13.70 12.75 12.06 11.67 21.92 20.26 20.48 19.59 Couple, 4+ children 33.10 30.95 30.74 32.68 44.80 42.63 41.84 44.94 Single motherc with children 11.87 12.46 12.79 13.82 20.49 19.65 20.64 20.74 All other households 11.61 10.81 10.26 10.45 18.39 17.04 17.88 17.46 TOTAL 9.60 9.01 8.86 8.60 15.17 14.24 14.84 14.24 Relative to TOTAL Single person 0.38 0.37 0.44 0.26 0.38 0.39 0.44 0.32 Couple, 0 childrenb 0.20 0.20 0.22 0.16 0.24 0.23 0.25 0.19 Couple, 1 child 0.43 0.45 0.50 0.43 0.45 0.46 0.47 0.43 Couple, 2 children 0.72 0.73 0.76 0.65 0.78 0.81 0.82 0.77 Couple, 3 children 1.43 1.41 1.36 1.36 1.44 1.42 1.38 1.38 Couple, 4+ children 3.45 3.44 3.47 3.80 2.95 2.99 2.82 3.16 Single motherc with children 1.24 1.38 1.44 1.61 1.35 1.38 1.39 1.46 All other households 1.21 1.20 1.16 1.22 1.21 1.20 1.20 1.23 TOTAL 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Relative to single person Single person 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Couple, 0 childrenb 0.52 0.53 0.49 0.62 0.64 0.60 0.57 0.61 Couple, 1 child 1.14 1.21 1.13 1.64 1.19 1.18 1.08 1.37 Couple, 2 children 1.90 1.95 1.71 2.49 2.06 2.09 1.89 2.44 Couple, 3 children 3.76 3.81 3.08 5.17 3.80 3.68 3.17 4.34 Couple, 4+ children 9.09 9.24 7.86 14.49 7.77 7.73 6.47 9.96 Single motherc with children 3.26 3.72 3.27 6.13 3.56 3.57 3.19 4.59 All other households 3.19 3.23 2.62 4.64 3.19 3.09 2.77 3.87 Notes: bChild - a dependent household member, age 24-. cSingle fathers (less then 0.1 percent of the sample) are not displayed. Source: HBS 2001, own computation. 358 Table A14. Poverty by education of a household head Education level* of a Hard poverty Medium poverty household head GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Poverty Line (PLN/month) 336 336 302 355 389 389 349 404 Headcount [%] Tertiary 0.57 0.56 0.82 0.44 1.29 1.31 1.60 1.26 Secondary general 3.75 3.56 3.85 3.33 6.96 6.54 7.56 6.58 Secondary vocational 12.16 11.51 11.04 10.72 19.01 17.95 18.78 18.03 Primary 17.72 16.41 16.01 16.24 26.76 24.88 24.62 24.69 TOTAL 9.60 9.01 8.86 8.60 15.17 14.24 14.84 14.24 Relative to TOTAL Tertiary 0.06 0.06 0.09 0.05 0.09 0.09 0.11 0.09 Secondary general 0.39 0.40 0.43 0.39 0.46 0.46 0.51 0.46 Secondary vocational 1.27 1.28 1.25 1.25 1.25 1.26 1.27 1.27 Primary 1.85 1.82 1.81 1.89 1.76 1.75 1.66 1.73 TOTAL 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Relative to tertiary Tertiary 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Secondary general 6.56 6.38 4.68 7.57 5.40 4.99 4.73 5.20 Secondary vocational 21.27 20.63 13.45 24.36 14.73 13.71 11.76 14.27 Primary 30.99 29.40 19.49 36.91 20.74 19.00 15.41 19.53 Notes: Secondary general includes uncompleted tertiary; primary includes uncompleted primary Source: HBS 2001, own computation. 359 Table A15. Population in poverty Population in poverty by voivodship Hard poverty Medium poverty Voivodship Total GUS GUS Population reg WB0 WB2 GUS GUS reg WB0 WB2 DOLNOLSKIE 9.2% 9.9% 10.5% 10.1% 8.8% 9.4% 9.7% 9.5% 7.7% KUJAWSKO-POMORSKIE 6.1% 5.4% 6.0% 6.0% 6.0% 5.8% 6.0% 6.5% 5.8% LUBELSKIE 8.8% 7.6% 7.2% 7.4% 8.3% 7.4% 7.3% 7.1% 6.0% LUBUSKIE 2.4% 2.6% 2.6% 2.6% 2.6% 2.7% 2.9% 3.0% 2.8% LÓDZKIE 5.7% 5.3% 5.3% 5.8% 6.2% 5.7% 5.2% 5.7% 7.3% MALOPOLSKIE 7.0% 7.9% 6.6% 6.9% 7.5% 8.3% 7.9% 8.4% 8.5% MAZOWIECKIE 8.8% 9.2% 8.6% 8.6% 9.5% 10.1% 9.0% 9.3% 12.5% OPOLSKIE 2.5% 2.5% 2.4% 2.2% 2.2% 2.2% 2.2% 2.2% 2.8% PODKARPACKIE 7.1% 6.3% 5.8% 5.6% 7.6% 6.8% 6.2% 6.2% 5.7% PODLASKIE 3.1% 2.6% 2.8% 2.8% 3.1% 2.7% 3.1% 2.8% 3.2% POMORSKIE 7.1% 7.7% 8.5% 8.7% 6.9% 7.1% 7.3% 7.1% 5.5% LSKIE 8.5% 9.2% 10.4% 9.4% 8.4% 9.0% 10.3% 9.4% 12.1% WITOKRZYSKIE 5.1% 4.6% 3.7% 4.2% 4.4% 4.3% 3.9% 4.0% 3.4% WARMISKO-MAZURSKIE 5.2% 5.1% 5.3% 5.6% 5.0% 4.8% 4.9% 4.9% 3.6% WIELKOPOLSKIE 8.3% 8.6% 8.4% 8.4% 8.5% 8.4% 8.6% 8.3% 8.8% ZACHODNIOPOMORSKIE 5.2% 5.4% 5.8% 5.6% 5.1% 5.3% 5.6% 5.6% 4.4% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Population in poverty by place of residence Hard poverty Medium poverty Place of residence Total GUS GUS Population reg WB0 WB2 GUS GUS reg WB0 WB2 Urban 37.3% 40.6% 45.6% 45.7% 39.1% 42.3% 47.1% 47.5% 60.3% 500 000+ 2.4% 3.4% 4.1% 4.0% 2.6% 3.7% 4.8% 4.6% 11.0% 200 - 500 000 4.5% 5.5% 6.5% 6.5% 4.9% 5.8% 6.7% 6.7% 10.2% 100 - 200 000 4.9% 5.2% 6.1% 6.0% 5.0% 5.5% 6.3% 6.3% 7.3% 20 - 100 000 11.6% 12.3% 13.5% 13.8% 12.3% 12.9% 14.7% 14.9% 18.8% 20 000 - 13.9% 14.1% 15.3% 15.4% 14.3% 14.4% 14.6% 14.9% 12.9% Rural 62.7% 59.4% 54.4% 54.3% 60.9% 57.7% 52.9% 52.5% 39.7% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Population in poverty by land area Hard poverty Medium poverty Land area Total GUS GUS Population reg WB0 WB2 GUS GUS reg WB0 WB2 0.00 - 0.09 62.7% 65.7% 70.1% 71.0% 62.2% 64.8% 68.5% 69.3% 74.0% 0.10 - 1.00 12.1% 11.5% 10.4% 9.3% 11.9% 11.8% 10.5% 10.2% 9.0% All non-farm households 74.9% 77.2% 80.5% 80.4% 74.1% 76.6% 79.0% 79.4% 83.0% 1.01 - 1.99 4.8% 4.5% 3.8% 3.9% 4.6% 4.4% 4.0% 4.3% 3.1% 2.00 - 4.99 9.2% 8.7% 6.8% 7.3% 9.1% 8.5% 7.5% 7.5% 5.4% 5.00 - 6.99 3.4% 2.9% 2.6% 2.6% 3.4% 3.1% 2.5% 2.4% 2.0% 7.00 - 9.99 4.0% 3.6% 3.1% 2.9% 4.0% 3.6% 3.1% 3.0% 2.4% 10.00 - 14.99 2.1% 1.8% 2.0% 2.0% 2.7% 2.0% 2.2% 1.9% 2.0% 15.00 + 1.6% 1.5% 1.3% 0.9% 2.0% 1.9% 1.8% 1.5% 2.2% All farm households 24.8% 22.5% 19.3% 19.4% 25.7% 23.2% 20.8% 20.4% 17.0% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% aTotal agricultural land area over 1 ha is treated as a 'farm' in GUS and IERiG statistics. Grouping - according to GUS and IERiG. 360 Table A15 cont. (1) Population in poverty by socio-economic category of household Socio-economic category of Hard poverty Medium poverty Total household GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Population Poverty Line** (PLN/month) ------> 389 389 349 404 389 389 349 404 Percent Percent Worker/Employee 33.3% 34.7% 37.6% 38.2% 33.8% 35.0% 37.9% 38.5% 43.7% Worker-farmer 14.5% 13.8% 12.2% 12.1% 15.5% 14.4% 13.0% 12.9% 11.5% Farmer 7.8% 6.6% 5.5% 4.9% 8.7% 7.5% 6.2% 5.8% 5.6% Self-employed 3.9% 4.1% 4.5% 3.8% 4.3% 4.7% 4.4% 4.3% 7.6% Retiree 9.1% 8.8% 9.3% 8.9% 9.2% 9.1% 9.8% 9.3% 16.0% Disable / survivor 15.8% 15.7% 15.2% 15.0% 14.8% 15.2% 14.9% 14.9% 10.6% Living on soc transf and part-time 15.6% 16.2% 15.8% 17.0% 13.7% 14.2% 13.9% 14.4% 5.0% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Population in poverty by the employment status of a household head Employment status of a hhld Hard poverty Medium poverty Total head GUS GUS Population reg WB0 WB2 GUS GUS reg WB0 WB2 Employed** 52.3% 51.7% 52.2% 51.7% 56.6% 55.5% 55.8% 55.4% 66.7% Unemployed** 9.1% 9.4% 9.3% 9.8% 7.4% 7.7% 7.7% 7.8% 2.3% Other** 38.6% 38.9% 38.6% 38.5% 36.0% 36.8% 36.5% 36.7% 31.0% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Population in poverty by the unemployment status of household members Unemployment of household Hard poverty Medium poverty Total members GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Population Poverty Line** (PLN/month) ------> 389 389 349 404 389 389 349 404 Hhld without unemployed 57.6% 56.2% 56.1% 53.3% 60.4% 59.2% 59.1% 58.4% 80.5% Hhld with unemployed 42.8% 44.1% 44.3% 47.1% 39.6% 40.8% 40.9% 41.6% 19.5% of which: 1 unemployed 27.1% 28.1% 28.9% 29.8% 26.5% 27.1% 27.6% 27.4% 15.6% 2or more unemployed 15.8% 16.1% 15.4% 17.4% 13.2% 13.7% 13.3% 14.2% 3.9% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Population in poverty by the employment status of household members Employment of household Hard poverty Medium poverty Total members GUS GUS Population reg WB0 WB2 GUS GUS reg WB0 WB2 Nof employed = 0 34.0% 35.4% 36.0% 35.8% 31.1% 32.1% 32.1% 32.4% 24.6% Nof employed = 1or more 66.0% 70.3% 71.5% 73.7% 68.9% 67.9% 67.9% 67.6% 75.4% of which: 1 employed 34.7% 34.7% 35.6% 36.2% 35.6% 36.4% 36.9% 36.7% 32.9% 2 or more employed 31.3% 29.9% 28.3% 28.0% 33.4% 31.5% 31.0% 30.9% 42.5% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 361 Table A15 cont. (2) Population in poverty by the disability status of household members Disability of household Hard poverty Medium poverty Total members GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Population Hhld without the disable 65.2% 65.6% 66.6% 65.4% 64.2% 64.8% 64.8% 63.9% 68.9% Hhld with the disable 34.8% 34.4% 33.4% 34.6% 35.8% 35.2% 35.2% 36.1% 31.1% of which: 1 disable person 26.4% 26.2% 25.6% 26.1% 27.0% 26.6% 26.7% 27.0% 23.4% 2 or disable persons 8.3% 8.1% 7.8% 8.5% 8.8% 8.6% 8.5% 9.1% 7.7% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Disable: a person 15+ declaring diability status. Population in poverty by family type Hard poverty Medium poverty Family type Total GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Population Single person 2.0% 1.9% 2.3% 1.4% 2.0% 2.0% 2.3% 1.7% 5.2% Couple, 0 childrenb 2.2% 2.2% 2.4% 1.8% 2.7% 2.6% 2.8% 2.2% 11.1% Couple, 1 child 5.0% 5.2% 5.7% 5.0% 5.2% 5.2% 5.4% 5.0% 11.5% Couple, 2 children 13.0% 13.1% 13.6% 11.8% 14.1% 14.6% 14.8% 13.9% 18.0% Couple, 3 children 11.5% 11.4% 11.0% 10.9% 11.7% 11.5% 11.1% 11.1% 8.1% Couple, 4+ children 16.2% 16.1% 16.3% 17.8% 13.9% 14.0% 13.2% 14.8% 4.7% Single motherc with children 3.4% 3.8% 3.9% 4.4% 3.7% 3.8% 3.8% 4.0% 2.7% All other households 46.6% 46.2% 44.5% 46.8% 46.6% 46.0% 46.4% 47.2% 38.5% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% bChild - a dependent household member, age 24-. cSingle fathers (less then 0.1 percent of the sample) are not displayed. Population in poverty by education level of the household head Education level*of a Hard poverty Medium poverty Total household head GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Population Tertiary 0.6% 0.6% 1.0% 0.5% 0.9% 1.0% 1.1% 0.9% 10.4% Secondary general 11.9% 12.1% 13.3% 11.8% 14.0% 14.0% 15.5% 14.1% 30.5% Secondary vocational 47.2% 47.6% 46.4% 46.5% 46.7% 47.0% 47.2% 47.2% 37.3% Primary 40.2% 39.7% 39.4% 41.2% 38.4% 38.1% 36.2% 37.8% 21.8% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% *Secondary general includes uncompleted tertiary; primary includes uncompleted promary Source: HBS 2001, own computation. 362 Table A16. Poverty in 1994 - 2001 according to various LS concepts Poverty headcounts and lines 1994 1995 1996 1997 1998 1999 2000 2001 MEDIUM POVERTY Headcounts [%] GUS 19.97 18.27 16.07 15.04 12.87 14.85 15.06 15.17 GUS reg 19.16 17.29 15.30 14.15 12.04 14.00 14.31 14.24 WB0 17.99 16.93 14.65 13.59 12.59 13.48 14.50 14.84 Poverty lines* [PLN/month] GUS and GUS reg 159 203 243 279 312 335 369 389 WB0 142 182 218 250 280 300 331 349 Memo: Published headcounts (for S A threshold)* N/A N/A N/A 13.3 12.1 14.4 13.6 15.0 Publ poverty line (annual avg for a single person), PLN N/A N/A N/A ? 333 367 356 385 HARD POVERTY Headcounts& [%] GUS 12.36 10.84 9.52 9.08 7.67 8.91 9.20 9.60 GUS reg 11.66 10.19 8.83 8.35 6.99 8.08 8.65 9.01 WB0 10.53 9.57 8.15 7.99 6.77 7.74 8.57 8.86 Poverty lines** [PLN/month] GUS and GUS reg 137 175 210 241 270 289 318 336 WB0 123 157 189 217 242 260 286 302 Memo: Published headcounts (for S M threshold)* 6.4 N/A 4.3 5.4 5.6 6.9 8.1 9.5 Publ poverty line (annual avg for a single person), PLN N/A N/A 177 213 247 273 308 334 Note: * S A threshold - social assistance; S M threshold - Subsitence Minimum. As declared in the published reports, following methods have been adopted : OECD 50/70 scales, specific concept of the number of household members, and quarterly adjusted poverty lines. N/A: Figures Not Available. For methodology (LS concepts and details) - see also Section 1 (main text) and Methodological Appendix. Source: HBS 1994-2001, own computation. Published headcounts: GUS (1999, 1998, 2001, 2002) Living Conditions of the Population in [year}. (various editions). 363 Table A17. Consumption aggregates 1994 - 2001: an overview HBS consumption mean, Gini and consumption growth 1994 1995 1996 1997 1998* 1999 2000 2001 Avg. equivalent consumption per month (PLN, current) GUS 280.78 362.89 458.83 556.43 651.95 680.56 753.13 789.56 GUS reg** 280.68 362.93 459.00 556.30 650.46 679.06 749.87 787.53 WB0** 237.34 305.25 380.11 453.99 514.15 552.85 603.90 629.63 WB0/GUS reg 0.85 0.84 0.83 0.82 0.79 0.81 0.81 0.80 Gini coefficient*** GUS 0.318 0.312 0.326 0.336 0.334 0.336 0.340 0.341 GUS reg** 0.313 0.307 0.321 0.330 0.327 0.329 0.333 0.333 WB0** 0.279 0.275 0.278 0.284 0.279 0.287 0.290 0.290 Avg equivalent consumption**** growth rate (%) GUS 101.1 105.5 105.5 104.8 97.3 100.5 99.4 GUS reg** 101.2 105.5 105.5 104.6 97.3 100.3 99.5 WB0** 100.6 103.9 103.9 101.3 100.2 99.2 98.8 Note: * Some corrections of expenditure concepts have been introduced by GUS in 1998 (see: published HBS 1998, Methodological. Notes, p. XXV). f Equivalence scales according to the Polish social assistance system (= OECD 70/50, but a single person = 1.1). All aggregates are expressed in June prices of a given year. Avg - weighted with the number of household members. ** Updated regional deflators for 1994-2000; first version of deflators for 2001. ***Gini coefficients are for per capita aggregates, weighted with persons (no sample weights have been used). ****Consumption in constant prices (deflator: annual CPI). See also Section 1 (main text), and Methodological Appendix. Source: HBS, own computation 364 Table A18. Selected household characteristics, 1994 - 2001 HBS characteristic 1994 1995 1996 1997 1998 1999 2000 2001 Sample size (nof households) 32,085 32,009 31,907 31,776 31,756 31,427 36,163 31,847 Avg. household sizea 3.05 3.06 3.17 3.12 3.04 3.10 3.12 3.11 Vulnerable households [%] Farmer and mixedb 13.2 13.3 13.5 14.6 13.9 13.7 12.5 12.1 Living on disability pensions na na na na 13.2 13.3 13.5 13.6 Living on social transfersc 4.8 4.8 3.7 3.4 2.9 3.4 4.5 5.1 Vulnerable population [%] Share of children 14- 23.2 22.8 23.0 22.9 22.0 21.8 20.8 20.4 Share of the elderly 65+ 10.9 11.1 10.4 10.7 11.0 10.8 11.1 11.0 Dependency ratee [%] 56.6 55.6 56.8 56.7 54.0 Unemployment indices Hhlds with the unemployed [%] 15.0 14.5 13.9 12.2 10.8 10.4 13.9 15.3 Unemployment ratiod 8.7 8.3 7.6 6.9 6.0 5.8 7.8 8.6 Notes: aAverage household size is based on the actual number of household members, neglecting the mumber of days in or out the household (see: GUS approach). This size is usually lower than reported by GUS. bHouseholds declaring 'agriculture' as main or additional source of income. cMostly living on unemployment benefits and on part-time jobs. dRatio of the number of the unemployed (registered or not) to the number of household members age 15-64. eRatio of the number of dependents (including children) to the number of household members with an individual source of income. Notice also that some corrections of the household categorization based on the main source of income have been introduced by GUS in 1997, and of income-expenditure concepts in 1998.(see: Published HBS 1998 and 1999, Methodological Notes.) Source: HBS 1994-2001, own computation. 365 Table A19. Selected economic indicators, 1994 - 2001 Indicator 1994 1995 1996 1997 1998 1999 2000 2001 GDP growth rate 5.3 7.0 6.0 6.8 4.8 4.1 4.0 1.0 Inflation rate [CPI] 32.2 27.8 19.9 14.9 11.8 7.3 10.1 5.5 Regist unemployment rate* 16.6 15.2 14.3 11.6 9.6 11.6 13.6 15.9 Unemployed not entitled to benefit 49.9 41.1 48.1 69.5 77.1 76.4 79.7 80.0 BAEL (LFS) unemployment rate** 13.9 13.1 11.5 10.2 10.6 15.3 16.0 18.5 BAEL (LFS) employment rate** 51.0 50.7 51.2 51.5 51.0 48.0 47.4 45.5 Average gross wage growth 1.7 2.8 5.5 5.9 3.3 4.7 1.0 2.5 Average gross pension growth 4.1 3.2 2.1 4.6 2.2 3.9 -2.3 4.7 Note: * Registered unemployment, June 30 ** BAEL (Labor Force Survey, LFS): 1994-1998 in November; 1999-2001; in the fourth quarter Sources: GUS Statistical Yearbooks, Statistical Bulletins, and Webpage (various editions). GUS, 2002, Concise Statistical Yearbook of Poland 2002, table 12(97). GUS, 2002, Labour Force Survey in Poland in the years 1992-2001. MoL, 2001, Basic Statistical Data on Social Policy. 366 Table A20. Medium poverty by various household categories 1994- 2001 WB0 headcounts HEADCOUNTS (%) 1994 1995 1996 1997 1998 1999 2000 2001 Poland - total 18.0 16.9 14.6 13.6 12.6 13.5 14.5 14.8 By residence: Urban 14.8 14.3 11.8 10.8 9.3 9.6 11.5 11.6 500 000+ 9.3 9.2 6.1 6.4 4.5 4.0 5.4 6.4 200 - 500 000 13.3 12.9 11.5 10.1 8.5 9.3 10.2 9.8 100 - 200 000 12.9 12.3 10.8 10.7 8.7 9.3 11.8 12.8 20 - 100 000 16.9 16.1 13.2 12.2 10.1 10.3 11.6 11.6 20 000 - 18.9 18.0 15.9 13.6 13.6 14.1 17.2 16.8 Rural 23.1 20.9 18.8 18.0 17.8 19.5 19.1 19.8 By socio-economic group**: Worker 15.8 14.8 13.0 12.2 11.3 11.4 12.6 12.9 Worker-Farmer 17.5 17.9 15.5 13.4 14.4 16.4 14.4 16.7 Farmer 21.7 17.8 16.3 14.6 17.9 19.4 17.7 16.3 Pensioner (retiree) 17.6 16.5 13.6 12.5 7.0 7.8 9.2 9.1 Pensioner (disable) 17.2 18.5 21.5 21.0 Self-employed 7.4 7.5 6.1 7.7 7.3 6.6 7.3 8.5 Living on social transfersa 50.8 51.3 47.5 47.6 41.3 43.6 42.8 40.8 By unemployment status: Nof unemployed = 0 14.0 13.5 11.7 10.7 10.0 10.9 10.7 10.9 Nof unemployed > 0 34.2 31.9 28.5 29.6 28.7 30.9 32.1 31.2 nof unemployed = 1 31.3 29.4 25.7 26.5 25.0 26.7 27.4 26.5 nof unemployed = 2+ 47.8 43.6 43.1 45.2 51.6 53.0 51.6 50.4 By education** of the hh head: tertiary 1.8 1.8 2.0 1.6 1.3 1.4 1.2 1.6 secondary 10.1 9.2 7.4 6.8 5.8 6.7 7.3 7.6 basic vocational 21.4 20.6 18.1 16.3 16.0 16.5 17.5 18.8 primary 25.9 23.9 22.1 21.7 19.9 21.7 24.4 24.6 By family type: Single person 12.1 11.0 9.8 8.0 7.5 6.1 7.5 6.5 Couple, 0 childrenb 5.3 5.5 4.4 3.7 3.2 3.6 3.6 3.7 Couple, 1 child 8.4 8.2 6.3 6.3 5.4 5.8 7.4 7.0 Couple, 2 children 14.2 13.3 11.3 11.7 10.7 10.5 11.0 12.2 Couple, 3 children 25.4 23.8 21.4 20.8 19.9 21.7 20.1 20.5 Couple, 4+ children 46.1 42.2 37.8 36.6 35.7 36.2 41.4 41.8 Single motherc with children 20.2 21.1 19.9 18.5 17.1 22.6 20.6 All other households 22.0 20.2 17.9 15.6 14.9 16.4 17.1 17.9 Note: ** Based on the main source of income. Changes of classification in 1997. a.Mostly households living on unemployment or social benefits, having children. Living on temporary jobs are also included in this group. b Child - a dependent household member, age 24-. c Single fathers (less then 0.1 percent of the sample) are not displayed. Notice also that GUS has changed the way of data processing with regard to expenditure in 1998. Shaded - headcounts above the overall. Source: HBS 1994-2001, own computation. 367 Table A21. Medium poverty by various household categories 1994- 2001 WB0 relative headcounts RELATIVE HEADCOUNTS 1994 1995 1996 1997 1998 1999 2000 2001 Poland - total 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 By residence: Urban 0.82 0.84 0.81 0.80 0.74 0.71 0.79 0.78 500 000+ 0.52 0.54 0.42 0.47 0.36 0.30 0.37 0.43 200 - 500 000 0.74 0.76 0.79 0.74 0.68 0.69 0.70 0.66 100 - 200 000 0.72 0.72 0.74 0.79 0.69 0.69 0.82 0.86 20 - 100 000 0.59 0.63 0.70 0.69 0.63 0.66 0.76 0.74 20 000 - 1.13 1.12 1.17 1.22 1.12 1.10 1.11 1.12 Rural 1.59 1.52 1.40 1.39 1.55 1.47 1.36 1.38 By socio-economic group**: Worker 0.80 0.85 0.92 0.90 0.80 0.83 0.91 0.92 Worker-Farmer 1.35 1.27 1.21 1.15 1.45 1.34 1.25 1.22 Farmer 1.57 1.34 1.07 1.16 1.68 1.50 1.24 1.23 Pensioner (retiree) 0.56 0.56 0.58 0.54 0.62 0.64 0.61 0.54 Pensioner (disable/survivor) 0.46 0.46 0.47 0.46 Self-employed 1.28 1.25 1.22 1.24 1.25 1.29 1.22 1.24 Living on social transfersa 3.22 3.31 3.24 3.45 2.73 2.79 2.77 2.85 By unemployment status: Nof unemployed = 0 0.78 0.80 0.80 0.79 0.80 0.81 0.74 0.74 Nof unemployed > 0 1.90 1.88 1.95 2.18 2.28 2.29 2.22 2.11 nof unemployed = 1 1.74 1.74 1.75 1.95 1.99 1.98 1.89 1.78 nof unemployed = 2+ 2.66 2.58 2.95 3.32 4.10 3.93 3.56 3.39 By education** of the hh head: tertiary 0.10 0.11 0.13 0.11 0.10 0.11 0.09 0.11 secondary 0.56 0.54 0.50 0.50 0.46 0.50 0.50 0.51 basic vocational 1.19 1.22 1.24 1.20 1.27 1.22 1.21 1.27 primary 1.44 1.41 1.51 1.60 1.58 1.61 1.68 1.66 By family type: Single person 0.67 0.65 0.67 0.59 0.60 0.45 0.51 0.44 Couple, 0 childrenb 0.30 0.32 0.30 0.28 0.25 0.26 0.25 0.25 Couple, 1 child 0.46 0.49 0.43 0.47 0.43 0.43 0.51 0.47 Couple, 2 children 0.79 0.79 0.77 0.86 0.85 0.78 0.76 0.82 Couple, 3 children 1.41 1.41 1.46 1.53 1.58 1.61 1.39 1.38 Couple, 4+ children 2.57 2.49 2.58 2.69 2.84 2.69 2.85 2.82 Single motherc with children 1.13 1.25 1.46 1.47 1.27 1.56 1.39 All other households 1.23 1.19 1.22 1.15 1.18 1.22 1.18 1.20 Note: a.Mostly households living on unemployment or social benefits, having children. Living on temporary jobs are also included in this group. b Child - a dependent household member, age 24-. c Single fathers (less then 0.1 percent of the sample) are not displayed. See also methodological sheets. Shaded - headcounts above the overall. Source: HBS 1994-2001, own computation. 368 Table A22. Hard Poverty by various household categories 1994- 2001 WB0 headcounts HEADCOUNTS (%) 1994 1995 1996 1997 1998 1999 2000 2001 Poland - total 10.5 9.6 8.2 8.0 6.8 7.7 8.6 8.9 By residence: Urban 8.5 8.0 6.5 6.3 5.2 5.2 6.8 6.7 500 000+ 5.0 4.8 3.1 3.6 2.4 2.4 2.9 3.3 200 - 500 000 8.3 7.3 6.0 5.7 4.1 5.4 5.5 5.7 100 - 200 000 7.1 6.7 5.9 6.5 4.6 4.9 7.3 7.4 20 - 100 000 9.6 9.6 7.3 7.2 6.2 5.4 6.8 6.4 20 000 - 11.2 9.5 9.5 7.7 7.4 7.8 11.0 10.5 Rural 13.7 11.9 10.5 10.7 9.3 11.6 11.2 12.1 By socio-economic group**: Worker 8.7 8.1 7.3 7.0 6.0 6.0 7.1 7.6 Worker-Farmer 8.7 9.1 8.3 7.2 7.1 8.9 6.9 9.4 Farmer 12.2 9.5 8.0 7.4 8.5 11.1 10.0 8.8 Pensioner (retiree) 10.6 9.2 7.3 7.4 3.6 4.4 5.7 5.1 Pensioner (disable/survivor) 9.6 11.3 13.3 12.8 Self-employed 3.6 4.4 2.2 4.2 3.6 3.7 4.1 5.2 Living on social transfersa 37.8 35.2 32.8 35.5 29.9 31.5 32.1 27.7 By unemployment status: Nof unemployed = 0 7.8 7.2 6.2 5.9 5.1 6.0 5.8 6.2 Nof unemployed > 0 21.6 19.8 17.6 19.6 17.4 19.2 21.6 20.1 nof unemployed = 1 19.0 18.2 15.1 17.2 14.2 16.0 17.3 16.5 nof unemployed = 2+ 33.4 27.6 30.3 31.8 36.6 36.1 39.0 34.8 By education** of the hh head: tertiary 0.9 0.9 0.8 0.8 0.4 0.5 0.7 0.8 secondary 5.0 4.3 3.8 3.3 2.8 3.1 3.8 3.8 basic vocational 12.5 11.6 10.1 9.5 8.6 9.6 10.1 11.0 primary 16.1 14.5 12.7 13.9 11.3 13.4 15.4 16.0 By family type: Single person 6.7 6.1 5.1 4.5 4.2 3.2 4.4 3.9 Couple, 0 childrenb 2.8 2.8 2.0 1.9 1.4 1.8 1.9 1.9 Couple, 1 child 4.8 4.2 3.3 3.3 2.5 3.0 3.8 4.4 Couple, 2 children 7.7 7.2 5.6 6.3 5.5 5.5 6.5 6.7 Couple, 3 children 14.7 13.2 11.8 12.5 10.6 13.1 12.0 12.1 Couple, 4+ children 31.3 27.3 24.4 24.9 23.9 23.1 28.6 30.7 Single motherc with children 12.0 13.7 13.5 10.7 10.4 13.5 12.8 All other households 12.9 11.2 10.0 9.1 7.7 9.4 9.8 10.3 Note: ** Based on the main source of income. Changes of classification in 1997. a.Mostly households living on unemployment or social benefits, having children. Living on temporary jobs are also included in this group. b Child - a dependent household member, age 24-. c Single fathers (less then 0.1 percent of the sample) are not displayed. Notice correction of expenditure registration by in 1998. See also methodological sheets. Shaded - headcounts above the overall. Source: HBS 1994-2001, own computation. 369 Table A23. Hard Poverty by various household categories 1994- 2001 WB0 relative headcounts RELATIVE HEADCOUNTS 1994 1995 1996 1997 1998 1999 2000 2001 Poland - total 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 By residence: Urban 0.81 0.83 0.80 0.78 0.76 0.68 0.80 0.76 500 000+ 0.47 0.50 0.38 0.45 0.36 0.30 0.34 0.37 200 - 500 000 0.79 0.76 0.74 0.71 0.61 0.70 0.65 0.64 100 - 200 000 0.68 0.70 0.72 0.82 0.68 0.63 0.85 0.84 20 - 100 000 0.91 1.00 0.90 0.90 0.92 0.70 0.79 0.72 20 000 - 1.06 1.00 1.16 0.97 1.09 1.00 1.28 1.19 Rural 1.30 1.25 1.29 1.34 1.37 1.49 1.31 1.37 By socio-economic group**: Worker 0.83 0.85 0.89 0.87 0.89 0.78 0.83 0.86 Worker-Farmer 0.82 0.95 1.02 0.90 1.05 1.15 0.80 1.06 Farmer 1.16 0.99 0.99 0.92 1.26 1.43 1.16 0.99 Pensioner (retiree) 1.01 0.96 0.90 0.93 0.53 0.57 0.66 0.58 Pensioner (disable/survivor) 1.42 1.46 1.55 1.44 Self-employed 0.34 0.46 0.27 0.52 0.53 0.47 0.48 0.59 Living on social transfersa 3.59 3.67 4.02 4.45 4.42 4.07 3.74 3.13 By unemployment status: Nof unemployed = 0 0.75 0.75 0.76 0.74 0.75 0.78 0.68 0.70 Nof unemployed > 0 2.05 2.07 2.15 2.45 2.57 2.49 2.52 2.27 nof unemployed = 1 1.81 1.90 1.85 2.15 2.10 2.07 2.02 1.86 nof unemployed = 2+ 3.18 2.88 3.71 3.97 5.41 4.67 4.55 3.92 By education** of the hh head: tertiary 0.08 0.09 0.10 0.10 0.06 0.06 0.08 0.09 secondary 0.47 0.44 0.46 0.41 0.41 0.40 0.45 0.43 basic vocational 1.19 1.21 1.24 1.19 1.27 1.24 1.18 1.25 primary 1.53 1.52 1.56 1.74 1.66 1.73 1.80 1.81 By family type: Single person 0.63 0.63 0.63 0.56 0.62 0.41 0.51 0.44 Couple, 0 childrenb 0.26 0.29 0.24 0.24 0.21 0.23 0.22 0.22 Couple, 1 child 0.46 0.44 0.41 0.41 0.37 0.39 0.45 0.50 Couple, 2 children 0.73 0.76 0.69 0.79 0.82 0.71 0.75 0.76 Couple, 3 children 1.39 1.38 1.45 1.56 1.56 1.69 1.39 1.36 Couple, 4+ children 2.98 2.85 3.00 3.12 3.53 2.98 3.33 3.47 Single motherc with children 1.14 1.43 1.69 1.58 1.34 1.58 1.44 All other households 1.23 1.17 1.23 1.13 1.13 1.22 1.14 1.16 Note: a.Mostly households living on unemployment or social benefits, having children. Living on temporary jobs are also included in this group. b Child - a dependent household member, age 24-. c Single fathers (less then 0.1 percent of the sample) are not displayed. See also methodological sheets. Shaded - headcounts above the overall. Source: HBS 1994-2001, own computation. 370 Table A24. Cross-Section vs. Panel Samples, 1994-2000 WB0 Living Standards Distributions 1994 1995 1996 1997 1998 1999 2000 Mean cross-section 237 305 380 454 514 553 604 panel 233 302 371 441 493 537 590 Median cross-section 211 273 339 403 460 492 534 panel 209 271 334 398 447 483 536 Std.dev. cross-section 122 150 190 254 257 287 323 panel 112 144 182 215 236 265 315 Gini's coeff. cross-section 0.2790 0.2747 0.2778 0.2844 0.2795 0.2872 0.2900 panel 0.2476 0.2429 0.2441 0.2458 0.2451 0.2543 0.2524 Medium poverty headcount cross-section 18.0% 16.9% 14.6% 13.6% 12.6% 13.5% 14.5% panel 18.9% 16.3% 15.3% 13.2% 14.1% 14.8% 14.4% Hard poverty headcount cross-section 10.5% 9.6% 8.2% 8.0% 6.8% 7.7% 8.6% panel 10.8% 8.9% 8.2% 7.8% 8.1% 8.8% 8.4% Note: WB0 LS measure, individual level, nominal terms. Source: HBS 1994-2000 and HBS panels 1994-1996 and 1997-2000, own calculations. 371 Table A25. Poverty Temporal Incidence, 1994-2000 1994-1996 1997-2000 Hard poverty Medium poverty Hard poverty Medium poverty 0 yrs 3 yrs 0 yrs 3 yrs 0 yrs 4 yrs 0 yrs 4 yrs Family Type Single 86.2% 2.2% 78.9% 2.5% 88.5% 0.4% 81.6% 1.0% Couple 0 children 95.1% 0.0% 89.2% 0.7% 94.6% 0.0% 90.2% 0.4% Couple 1 children 92.2% 0.6% 85.2% 1.6% 92.7% 0.6% 86.2% 1.7% Couple 2 children 83.2% 1.5% 74.6% 5.1% 86.2% 1.4% 77.1% 2.4% Couple 3 children 73.9% 3.0% 59.7% 10.8% 69.0% 3.7% 57.5% 10.0% Couple 4+ children 63.0% 13.7% 43.3% 24.3% 54.7% 6.8% 39.3% 16.1% Single parent 77.9% 2.9% 62.7% 7.3% 71.1% 4.6% 60.7% 5.5% Other 80.5% 3.4% 67.0% 7.6% 78.8% 1.2% 67.1% 3.6% Total 82.5% 2.8% 71.3% 6.7% 81.6% 1.6% 71.8% 4.0% Socio-Economic Category of Household Worker/Employee 82.6% 2.9% 72.2% 6.9% 83.5% 1.8% 74.3% 3.7% Worker-farmer 85.0% 3.2% 68.8% 6.0% 83.0% 1.3% 70.4% 3.5% Farmer 82.3% 1.3% 70.1% 6.0% 75.6% 0.4% 60.0% 4.1% Self-employed 84.3% 2.0% 72.9% 4.6% 92.2% 0.0% 87.7% 1.8% Retired/Disabled 91.6% 0.3% 86.3% 1.4% 83.4% 0.5% 73.6% 2.3% Living on soc. tr. 49.3% 14.3% 35.3% 30.0% 37.6% 11.9% 30.3% 21.5% Total 82.5% 2.8% 71.3% 6.7% 81.6% 1.6% 71.8% 4.0% Level of Education of Household Head (selected) Tertiary 96.7% 0.0% 95.1% 0.0% 98.5% 0.0% 95.6% 0.0% Secondary 89.2% 1.2% 81.9% 4.0% 88.0% 1.1% 81.0% 2.0% Vocational 79.3% 3.5% 66.1% 8.5% 77.4% 1.4% 66.7% 4.6% Elementary 75.5% 4.5% 60.1% 9.2% 71.7% 3.5% 60.0% 7.6% Total 82.5% 2.8% 71.3% 6.7% 81.6% 1.6% 71.8% 4.0% Size of Place of Living (in thousands) 500+ 89.7% 0.9% 82.9% 2.8% 95.3% 0.0% 91.0% 0.3% 100-500 84.1% 1.8% 75.8% 4.6% 89.4% 1.2% 83.9% 2.4% 20-100 83.3% 2.1% 73.6% 6.6% 84.8% 2.1% 74.9% 3.6% -20 81.3% 3.8% 68.8% 8.9% 78.9% 2.0% 69.8% 3.7% Rural 79.8% 3.8% 65.8% 8.1% 73.9% 1.9% 60.9% 5.9% Total 82.5% 2.8% 71.3% 6.7% 81.6% 1.6% 71.8% 4.0% Note: WBO LS measure, individual level. Source: HBS panels 1994-1996 and 1997-2000, own calculations. 372 Table A26. Poverty Mobility, 1994-2000Table 1994-1996 1997-2000 Hard poverty Medium poverty Hard poverty Medium poverty In Out In Out In Out In Out Family Type Single 3.9% 4.5% 6.8% 7.5% 0.7% 4.3% 0.9% 7.4% Couple 0 children 1.3% 2.4% 2.3% 4.3% 0.8% 1.9% 1.3% 2.9% Couple 1 children 1.0% 4.1% 2.4% 6.7% 0.7% 1.9% 1.0% 3.6% Couple 2 children 3.6% 7.0% 4.6% 9.6% 0.7% 3.3% 2.9% 3.7% Couple 3 children 6.4% 10.4% 8.0% 13.6% 3.4% 4.9% 7.6% 5.4% Couple 4+ children 10.1% 9.4% 10.6% 14.4% 8.9% 5.5% 10.3% 7.1% Single parent 5.4% 7.3% 5.5% 15.0% 1.9% 4.7% 9.5% 3.9% Other 4.8% 8.0% 7.6% 10.1% 2.5% 2.8% 4.0% 4.9% Total 4.2% 6.9% 6.0% 9.6% 2.1% 3.2% 3.8% 4.6% Socio-Economic Category of Household Worker/Employee 3.3% 7.9% 4.3% 10.6% 1.3% 2.8% 2.9% 3.2% Worker-farmer 4.2% 5.3% 8.7% 8.4% 2.2% 1.9% 3.5% 5.4% Farmer 5.8% 5.5% 7.2% 8.7% 5.5% 3.0% 9.7% 2.7% Self-employed 3.9% 6.1% 7.3% 9.3% 1.0% 2.1% 2.0% 3.8% Retired/Disabled 0.9% 3.4% 3.1% 3.8% 2.5% 3.3% 2.4% 6.3% Living on soc. tr. 18.6% 11.6% 13.2% 13.2% 4.7% 11.5% 17.1% 10.0% Total 4.2% 6.9% 6.0% 9.6% 2.1% 3.2% 3.8% 4.6% Level of Education of Household Head (selected) Tertiary 0.2% 1.7% 0.4% 3.1% 0.0% 1.1% 0.0% 1.8% Secondary 2.6% 4.7% 3.4% 6.5% 1.3% 2.8% 1.5% 4.3% Vocational 4.7% 8.0% 6.6% 10.9% 2.7% 3.2% 5.1% 4.3% Elementary 6.4% 9.4% 9.7% 12.7% 3.0% 4.9% 4.5% 6.6% Total 4.2% 6.9% 6.0% 9.6% 2.1% 3.2% 3.8% 4.6% Size of Place of Living (in thousands) 500+ 1.9% 4.9% 4.2% 5.7% 0.0% 2.5% 0.9% 2.7% 100-500 4.7% 6.8% 5.1% 9.4% 1.1% 2.2% 1.9% 2.6% 20-100 3.2% 7.5% 5.8% 8.5% 0.3% 2.0% 2.9% 4.1% -20 3.3% 7.3% 5.2% 10.9% 2.8% 3.0% 4.4% 4.7% Rural 5.3% 7.1% 7.2% 10.8% 3.7% 4.3% 5.7% 6.1% Total 4.2% 6.9% 6.0% 9.6% 2.1% 3.2% 3.8% 4.6% Note: WB0 LS measure, individual level. Source: HBS panels 1994-1996 and 1997-2000, own calculations. 373 Fig. A1. Medium poverty by land area, Poland 2001 32.0 27.0 ] %[s 22.0 CSO CSO reg ountcdaeH 17.0 WB0 WB2 12.0 7.0 2.0 1.01 - 1.99 2.00 - 4.99 5.00 - 6.99 7.00 - 9.99 10.00 - 14.99 15.00 + Land area (hectars) Source: HBS 2001, own computations. Fig. A2. Hard poverty by land area, Poland 2001 22.0 20.0 18.0 ] %[ 16.0 s 14.0 CSO CSO reg ountcdaeH 12.0 WB0 10.0 WB2 8.0 6.0 4.0 2.0 1.01 - 1.99 2.00 - 4.99 5.00 - 6.99 7.00 - 9.99 10.00 - 14.99 15.00 + Land area (hectars) Source: HBS 2001, own computations. 374 Fig. A3. Medium poverty (WB2) by age and gender Poland 2001 28.0 23.0 ] %[ stnu 18.0 co 13.0 eadH 8.0 3.0 0 - 4 5 - 9 10 -14 15 - 19 20 - 24 25 - 29 30 - 34 35 - 39 40 - 44 45 - 49 50 - 54 55 - 59 60 - 65 65 - 69 70 - 74 75 + Age brackets Male Female Source: HBS 2001, own computations. Fig. A4. Hard poverty (WB2) by age and gender Poland 2001 19.0 17.0 ] 15.0 %[ stnu 13.0 11.0 co 9.0 eadH 7.0 5.0 3.0 1.0 5 - 9 10 -14 15 - 19 20 - 24 25 - 29 30 - 34 35 - 39 40 - 44 45 - 49 50 - 54 55 - 59 60 - 65 65 - 69 70 - 74 75 + Total Age brackets Male Female Source: HBS 2001, own computations. 375 Fig. A5. Medium poverty:worker vs. farmer households, Poland 1994 - 2001 25.0 0B Wrof 20.0 stnu Worker 15.0 Farmer co 10.0 eadH 5.0 1994 1995 1996 1997 1998 1999 2000 2001 Source: HBS 2001, own computations. Fig. A6. Medium poverty by unemployment status, Poland 1994 - 2001 ] 35.0 %[ 0B 30.0 Wrofstnu 25.0 20.0 15.0 co eadH 10.0 5.0 1994 1995 1996 1997 1998 1999 2000 2001 Hlds w ithout unemployed Hhlds w ith unemployed Source: HBS 2001, own computations. Fig. A7. Medium poverty: rural vs. urban, Poland 1994 - 2001 ] 25.0 [% 0B 20.0 Wr fo Urban tsnuocdaeH 15.0 Rural 10.0 5.0 1994 1995 1996 1997 1998 1999 2000 2001 Source: HBS 2001, own computations. 376 Fig. A8. Hard Poverty: worker vs. farmer households, Poland 1994 - 2001 ] 14.0 [% 12.0 WB0rof 10.0 8.0 tsnu 6.0 Worker cod Farmer 4.0 Hea 2.0 1994 1995 1996 1997 1998 1999 2000 2001 Source: HBS 2001, own computations. Fig. A9. Hard poverty by unemployment status, Poland 1994 - 2001 ] 22.0 %[ 20.0 0B 18.0 Wrof 16.0 14.0 s 12.0 10.0 ountcdaeH 8.0 6.0 4.0 1994 1995 1996 1997 1998 1999 2000 2001 Hhlds w ithout unemployed Hhlds w ith unemployed Source: HBS 2001, own computations. Fig. A10. Hard poverty: rural vs. urban, Poland 1994 - 2001 ] 16.0 %[ 0 14.0 WBrof 12.0 s 10.0 ountcdaeH 8.0 Urban 6.0 Rural 4.0 1994 1995 1996 1997 1998 1999 2000 2001 Source: HBS 2001, own computations. 377 Map A1 Source: HBS 2001, own computations. Map A2 Source: HBS 2001, own computations. 378 Map A3 Source: HBS 2001, own computations. Map A4 Source: HBS 2001, own computations. 379 Map A5 Source: HBS 2001, own computations. Map A6 Source: HBS 2001, own computations. 380 Table AP1. Poverty according to various LS concepts: an overview per capita approach "Medium Poverty" LS* concept Poverty measures GUS GUS reg WB0 WB2 Poverty line [PLN/month] ---> 277 277 249 287 Headcounts [%] persons H 15.14 14.25 14.85 14.31 households Hh 9.94 9.39 9.78 9.08 Poverty depth [%] P1 3.50 3.23 3.19 2.92 Poverty gap [%] PG 23.10 22.68 21.50 20.40 "Hard Poverty" LS* concept Poverty measures GUS GUS reg WB0 WB2 Poverty line [PLN/month] ---> 238 238 213 250 Headcounts [%] persons H 9.64 8.98 8.96 8.77 households Hh 6.07 5.64 5.68 5.32 Poverty depth [%] P1 2.06 1.88 1.75 1.66 Poverty gap [%] PG 21.37 20.95 19.50 18.89 "Extreme Poverty" LS* concept Poverty measures GUS GUS reg WB0 WB2 Poverty line [PLN/month] ---> 183 183 168 168 Headcounts [%] persons H 3.76 3.39 3.45 3.31 households Hh 2.22 2.00 2.03 1.85 Poverty depth [%] P1 0.70 0.62 0.61 0.55 Poverty gap [%] PG 18.52 18.43 17.55 16.57 Notes: * LS = Living Standard. For methodology - see Section1 (main text) and Methodological Appendix. Poverty lines for per acpita approach are equal to the 'equivalent' lines, times the ratio of tthe first (extreme poverty), second (hard poverty) or third (medium poverty) half-deciles of the appropriate consumption aggregates. PovLines in constant 1998 PPP$/month Medium poverty 136 136 122 141 Hard poverty 117 117 104 122 Extreme poverty 90 90 82 82 Exchange rate: 2.042 WB Staff [MS] Source: HBS 2001, own computation. 381 Table AP2. Poverty for various household categories per capita approach HARD POVERTY MEDIUM POVERTY HEADCOUNTS (%) GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Poverty line [PLN/month] 277 277 249 287 238 238 213 250 Poland - total 9.6 9.0 9.0 8.8 15.1 14.3 14.9 14.3 By residence: Urban 5.9 5.9 6.6 6.5 9.7 9.8 11.3 10.8 500 000+ 1.9 2.4 3.0 3.0 3.4 4.8 6.1 5.9 200 - 500 000 4.0 4.6 5.5 5.5 7.1 7.9 9.5 8.7 100 - 200 000 6.1 6.2 7.4 7.2 10.2 10.6 13.1 12.2 20 - 100 000 5.7 5.7 6.3 6.1 9.5 9.4 11.2 10.6 20 000 - 10.9 10.0 10.5 10.7 17.0 15.7 16.4 16.0 Rural 15.3 13.7 12.5 12.2 23.4 21.0 20.2 19.7 By socio-economic category*: Worker 7.7 7.6 8.2 7.8 12.1 11.9 13.4 13.1 Worker-Farmer 13.0 11.4 10.8 10.1 21.9 19.0 18.5 17.5 Farmer 15.1 12.1 9.6 10.2 25.4 21.4 18.5 17.6 Self-employed 5.4 5.1 5.2 4.8 9.4 9.1 9.1 7.7 Pensioner (retiree) 4.2 3.8 3.6 3.5 7.0 6.5 7.0 6.5 Pensioner (disable/survivor) 12.3 11.2 11.0 10.9 19.0 18.4 18.2 17.8 Living on social transfersa 31.0 29.7 29.1 30.3 41.4 39.8 41.2 40.8 By unemployment status: Nof unemployed = 0 7.0 6.4 6.4 6.1 11.6 10.8 11.2 10.6 Nof unemployed > 0 20.6 19.7 19.6 19.6 29.9 28.5 30.0 29.7 nof unemployed = 1 16.6 15.9 16.1 15.8 25.2 23.9 25.5 24.5 nof unemployed = 2+ 36.7 34.7 33.1 34.7 48.6 46.9 48.2 49.9 By education** of the hh head: tertiary 0.7 0.7 0.8 0.5 1.4 1.4 1.6 1.5 secondary 3.8 3.6 4.0 3.3 7.2 6.7 7.6 6.7 basic vocational 12.6 11.9 11.9 11.7 19.9 18.7 19.9 18.8 primary 16.9 15.5 14.8 15.3 24.8 23.4 22.8 23.3 By family type: Single person 0.6 0.5 0.5 0.1 1.2 1.1 1.0 0.4 Couple, 0 childrenb 0.7 0.7 0.6 0.5 1.7 1.6 1.5 1.0 Couple, 1 child 3.4 3.0 3.3 2.6 5.5 5.4 5.9 5.1 Couple, 2 children 7.3 6.8 6.9 5.9 12.3 11.9 12.5 11.3 Couple, 3 children 16.0 15.1 15.2 14.9 25.2 24.3 24.5 23.9 Couple, 4+ children 40.5 38.6 38.1 40.2 52.7 50.5 52.0 54.3 Single motherc with children 13.2 12.9 13.9 13.6 21.0 20.7 21.3 21.5 All other households 11.1 10.2 9.9 10.1 18.0 16.5 17.3 16.8 Note: "Medium Poverty" uses social assistance threshold as poverty line. * Based on the main source of income. Changes of classification in 1997. a.Mostly households living on unemployment or social benefits, having children. Living on temporary jobs are also included in this group. b Child - a dependent household member, age 24-. c Single fathers (less then 0.1 percent of the sample) are not displayed. Source: HBS 2001, own computation 382 Table AP3. Relative poverty for various household categories per capita approach HARD POVERTY MEDIUM POVERTY REALTIVE HEADCOUNTS GUS GUS reg WB0 WB2 GUS GUS reg WB0 WB2 Poverty line [PLN/month] 277 277 249 287 238 238 213 250 Poland - total 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 By residence: Urban 0.61 0.65 0.74 0.74 0.64 0.69 0.76 0.75 500 000+ 0.20 0.26 0.33 0.34 0.22 0.33 0.41 0.41 200 - 500 000 0.42 0.51 0.62 0.62 0.47 0.55 0.64 0.61 100 - 200 000 0.64 0.69 0.83 0.82 0.68 0.75 0.88 0.85 20 - 100 000 0.59 0.63 0.70 0.69 0.63 0.66 0.76 0.74 20 000 - 1.13 1.12 1.17 1.22 1.12 1.10 1.11 1.12 Rural 1.59 1.52 1.40 1.39 1.55 1.47 1.36 1.38 By socio-economic category*: Worker 0.80 0.85 0.92 0.90 0.80 0.83 0.91 0.92 Worker-Farmer 1.35 1.27 1.21 1.15 1.45 1.34 1.25 1.22 Farmer 1.57 1.34 1.07 1.16 1.68 1.50 1.24 1.23 Self-employed 0.56 0.56 0.58 0.54 0.62 0.64 0.61 0.54 Pensioner (retiree) 0.43 0.42 0.40 0.40 0.46 0.46 0.47 0.46 Pensioner (disable/survivor) 1.28 1.25 1.22 1.24 1.25 1.29 1.22 1.24 Living on social transfersa 3.22 3.31 3.24 3.45 2.73 2.79 2.77 2.85 By unemployment status: Nof unemployed = 0 0.72 0.71 0.71 0.70 0.76 0.76 0.75 0.74 Nof unemployed > 0 2.14 2.20 2.18 2.23 1.97 2.00 2.02 2.07 nof unemployed = 1 1.72 1.78 1.80 1.80 1.66 1.67 1.71 1.71 nof unemployed = 2+ 3.80 3.87 3.70 3.96 3.21 3.29 3.25 3.49 By education** of the hh head: tertiary 0.08 0.08 0.09 0.06 0.09 0.10 0.11 0.11 secondary 0.40 0.40 0.44 0.38 0.47 0.47 0.51 0.47 basic vocational 1.31 1.32 1.33 1.33 1.31 1.31 1.34 1.32 primary 1.76 1.73 1.65 1.75 1.64 1.64 1.53 1.63 By family type: Single person 0.06 0.06 0.05 0.02 0.08 0.07 0.07 0.03 Couple, 0 childrenb 0.07 0.07 0.07 0.05 0.11 0.11 0.10 0.07 Couple, 1 child 0.35 0.34 0.37 0.30 0.37 0.38 0.40 0.36 Couple, 2 children 0.76 0.76 0.77 0.67 0.81 0.83 0.84 0.79 Couple, 3 children 1.66 1.68 1.70 1.69 1.66 1.71 1.65 1.67 Couple, 4+ children 4.20 4.30 4.25 4.58 3.48 3.54 3.50 3.80 Single motherc with children 1.37 1.43 1.55 1.55 1.39 1.45 1.44 1.50 All other households 1.15 1.13 1.11 1.15 1.19 1.15 1.16 1.17 Note: "Medium Poverty" uses social assistance threshold as poverty line. * Based on the main source of income. Changes of classification in 1997. a.Mostly households living on unemployment or social benefits, having children. Living on temporary jobs are also included in this group. b Child - a dependent household member, age 24-. c Single fathers (less then 0.1 percent of the sample) are not displayed. Source: HBS 2001, own computation. 383 Table AP4. Medium Poverty in 1994 - 2001 according to various LS concepts per capita approach Poverty headcounts [%] 1994 1995 1996 1997 1998 1999 2000 2001 GUS 19.5 17.9 16.2 15.3 13.3 15.1 15.2 15.1 GUS reg 18.7 16.9 15.5 14.4 12.4 14.2 14.4 14.3 WB0 17.7 16.9 15.1 14.2 12.8 13.9 14.7 14.9 Poverty profile for WB0 aggregate By residence: Urban 14.0 13.5 11.6 11.0 9.1 9.5 11.1 11.3 500 000+ 7.9 8.4 5.6 6.1 4.2 3.9 5.0 6.1 200 - 500 000 11.8 11.5 10.8 10.0 8.1 9.6 9.5 9.5 100 - 200 000 12.4 11.6 10.1 10.7 9.0 9.0 11.7 13.1 20 - 100 000 16.3 15.6 13.3 12.4 9.9 10.1 11.2 11.2 20 000 - 18.8 17.4 16.2 14.6 13.5 13.8 17.0 16.4 Rural 23.7 21.8 20.2 19.4 18.6 20.7 20.1 20.2 By socio-economic group**: Worker 16.8 15.8 14.5 13.5 12.1 12.1 13.1 13.4 Worker-Farmer 19.2 20.3 17.9 15.6 16.8 19.4 16.3 18.5 Farmer 23.1 20.2 18.9 17.2 19.4 22.2 20.8 18.5 Pensioner (retiree) 14.3 13.3 10.1 10.1 4.8 5.8 7.1 7.0 Pensioner (disable/survivor) 14.2 15.2 18.8 18.2 Self-employed 8.1 7.7 7.1 8.8 8.0 7.1 8.7 9.1 Living on social transfersa 49.8 49.8 49.1 50.2 41.6 45.9 43.9 41.2 By unemployment status: Nof unemployed = 0 13.7 13.5 12.1 11.3 10.3 11.4 11.0 11.2 Nof unemployed > 0 34.2 31.7 29.1 30.3 28.3 30.6 31.9 30.0 nof unemployed = 1 31.5 29.2 26.2 27.6 24.8 26.7 27.0 25.5 nof unemployed = 2+ 46.8 43.5 44.2 43.9 50.0 50.8 51.6 48.2 By education** of the hh head: tertiary 1.9 2.0 2.1 1.7 1.5 1.3 1.3 1.6 secondary 10.2 9.2 7.9 7.7 6.1 6.9 7.4 7.6 basic vocational 22.3 21.9 20.2 18.4 17.6 18.4 18.9 19.9 primary 23.8 21.8 20.0 20.3 17.9 20.1 22.7 22.8 By family type: Single person 2.0 1.6 1.2 1.2 0.9 0.7 1.3 1.0 Couple, 0 childrenb 2.2 2.5 1.8 1.6 1.3 1.5 1.7 1.5 Couple, 1 child 6.5 6.5 5.0 5.1 4.0 4.4 5.6 5.9 Couple, 2 children 15.1 15.1 12.7 12.9 11.6 10.8 11.8 12.5 Couple, 3 children 32.1 28.7 27.4 26.4 25.6 26.5 25.1 24.5 Couple, 4+ children 56.2 52.5 49.1 50.9 45.0 47.2 52.4 52.0 Single motherc with children 20.0 21.5 14.1 20.3 18.7 16.8 22.2 21.3 All other households 21.5 19.65 17.6 15.6 14.57 16.64 16.6 17.3 Note: Poverty lines of 2001 have been kept constant in real terms over the whole period (deflator - annual CPI). They are equal to (PLN/month): GUS and GUS reg consum aggregates 113 144 173 199 222 238 263 277 WB0 consumption aggregate 101 130 156 179 200 214 236 249 ** Based on the main source of income. Changes of classification in 1997. a.Mostly households living on unemployment or social benefits, having children. Living on temporary jobs are also included in this group. cSingle fathers (less then 0.1 percent of the sample) are not b Child - a dependent household member, age 24-. displayed. For methodology (LS concepts and details) - see also Section 1 (main text) and Methodological Appendix, and methodological sheets. Notice also corrections of expenditure concepts have been introduced by GUS in 1998 (see: Published HBS 1998, Methodological Notes, p. XXV). Shaded - headcounts above the overall. Source: HBS 1994-2001, own computation. 384 wb13696 P:\POLAND\PREM\Living Standard Assessment\4RED\Part 4\FINALPLS Vol 2 Chap12-140301.doc March 24, 2004 11:15 AM 385