W O R L D B A N K R E G I O N A L A N D S E C T O R A L S T U D I E S Income Support for the Unemployed Issues and Options MILAN VODOPIVEC Income Support for the Unemployed Issues and Options WORLD BANK REGIONAL AND SECTORAL STUDIES Income Support for the Unemployed Issues and Options Milan Vodopivec THE WORLD BANK Washington, D.C. © 2004 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved. The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750- 4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street, NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. ISBN 0-8213-5761-1 Cover photo: Jure Erzen Library of Congress Cataloging-in-Publication Data has been applied for. Contents Preface vii Acknowledgments ix 1. Introduction 1 2. Conceptual Issues 5 Objectives of Income Support Programs for the Unemployed 7 Accounting for Interactions among Social Risk Management Mechanisms 8 Establishing Evaluation Criteria 14 Why Should the Public Sector Provide Unemployment Insurance? 17 Who Should Receive Unemployment Income Support? 19 Summary 21 3. Review of Income Support Programs for the Unemployed 25 Typology and Description of Main Income Support Programs 25 The Incidence of Income Support Programs for the Unemployed 52 Concluding Remarks 62 Annex 3.1 Determinants of Social Insurance Programs for the Unemployed 63 4. Performance of Income Support Programs: Theoretical Aspects and Empirical Evidence 65 Income Protection Effects 66 Efficiency Effects 81 Suitability to Confront Different Shocks 104 v vi Contents Resistance to Political Risk 111 Concluding Remarks 117 Annex 4.1 Efficiency Effects of Income Support Programs ...............118 5. Design and Implementation Criteria: Exploring Country-Specific Conditions 131 Interactions with Labor Market Institutions and Shocks 131 Administrative Capacity for Program Implementation 133 Characteristics of Unemployment 137 Size of the Informal Sector 138 Prevalence and Pattern of Interhousehold Transfers 139 Nonsocial Insurance and Self-Protection 141 The Nature of Shocks 145 Cultural and Political Factors 146 Concluding Remarks 151 6. Improving Income Support Programs for the Unemployed in Developing Countries 153 Choosing the Right Program 153 Designing Unemployment Insurance 177 Improving Income Protection of the Informal Sector 180 Concluding Remarks 183 References 187 Index 209 Preface In May 2003 I was asked to help the Sri Lankan government introduce an unemployment insurance program. The task was daunting: never had I encountered a country at such a low level of economic development, with so large an informal sector and such weak administrative capacity, embark upon such an effort. Yet as I learned, providing unemployment insurance was the only politically acceptable way for Sri Lankan policymakers to reform another badly performing labor market institution: severance pay. This is an example of the difficult choices and tradeoffs developing coun- tries face in providing income support against unemployment risk. Industrial countries offer such support primarily by providing social insurance. When is a developing country ready to introduce such a program? How should the blueprints for programs in place in industrial countries be modified to meet the circumstances of developing countries, including the different character- istics of unemployed workers and the abundant employment opportunities in the informal sector? How should the tradeoffs between different types of public income support programs be evaluated? What should the mix be between public and private risk management arrangements? How should the income protection and efficiency effects of different programs be balanced? Can income protection come only at the cost of efficiency? I tackle these questions by evaluating different unemployment support programs and assessing their applicability in developing and transition countries. I combine and reconcile theoretical insights with empirical evi- dence, paying due attention to country-specific circumstances. I identify the strengths and weaknesses of each program, as well as point out circum- stances that are conducive to the success of each. I hope that the book will provide useful information and advice for policymakers throughout devel- oping and transition countries and that it will arouse interest among acade- micians and others interested in income support for the unemployed. vii viii Preface Writing this book proved to be a fascinating undertaking, not least because the approaches countries have taken are so diverse and because new, innovative, and sometimes controversial programs (such as unem- ployment insurance savings accounts) are being introduced. With increas- ing recognition that protecting workers is more productive than protecting jobs, income support for the unemployed has gained in importance, and promising new, attractive solutions may be forthcoming. Acknowledgments This book is part of a larger effort of the Labor Market group of the Human Development Network to understand better and synthesize the effects of labor market policies and programs. The author has many intellectual debts to acknowledge. Special gratitude is owed to Robert Holzmann, who provided the impetus for this work, as well as invaluable general guidance; to the reviewers of the World Bank's Editorial Committee for rich, careful, and insightful comments; and to Dhushyanth Raju for excellent research assistance. I also benefited from written comments from Gordon Betcherman, Peter Fredriksson, Indermit Gill, Luis Guasch, Bertil Holmlund, and Jan van Ours, as well as from fruitful discussions with Alan Abrahart, Christine Allision, Mavricija Batic, Jacob Benus, Tito Boeri, Rocio Castro, Wendy Cunningham, Amit Dar, John Earle, Jude Esguerra, Cresenta Fernando, Indermit Gill, Margaret Grosch, John Haltiwanger, Hugo Hopenhayn, Jai-Joon Hur, Arvo Kuddo, Hartmut Lehman, William Maloney, John Micklewright, Gyula Nagy, Philip O'Keefe, Christopher O'Leary, Peter Orazem, Michael Orszag, Carmen Pages, Robert Palacios, Robert Pavosevich, Sonja Pirher, Martin Rama, Mansoora Rashid, Michelle Ribaud, Elizabeth Ruppert, Michal Rutkowski, Stefano Scarpetta, Kalanidhi Subbarao, Jan Svejnar, Hong Tan, Kathy Terrell, Wayne Vroman, David Warren, and Jungyoll Yun. ix 1 Introduction Unemployment is becoming an increasingly pressing problem in many parts of the world. Macroeconomic crises and increased globalization have put more workers at risk of job loss in Latin America and, more recently, in East Asia. In an effort to transform themselves into market economies, former socialist countries have faced the enormous task of efficiently reallocating workers and jobs across sectors and firms, which has led to unemployment and poverty of large proportions. In Europe economic growth has declined and unemployment risen since the 1970s. Particularly worrisome is the rise in the share of the long-term unemployed. Given these trends, the task of helping the unemployed has gained impor- tance and attracted the interest of policymakers. Because job loss entails the loss of income, providing effective income support is a prime concern and a necessary component of assistance to the unemployed, be it in the form of a pure transfer or through jobs created by public programs.1 But other aspects must also be considered. It is important to consider how to increase the employability of the unemployed, that is, their capacity to search for jobs and to match skills with vacancies. Adverse labor supply incentives created by income transfers also need to be studied and addressed. And the risk of unemployment needs to be reduced, both by designing appropriate income support programs and by increasing employment opportunities. While the task of increasing employment opportunities reaches far beyond labor market policies and programs, important links between an economy's capacity to create jobs job and income support programs--and social protection systems in general--should not be overlooked. Indeed, income support pro- grams for the unemployed should be developed in line with a broader concep- tual framework that lays out the complex linkages between institutions and policies in a systematic and comprehensive way. This book relies on such a com- prehensive framework, developed by the World Bank (2001a). By formulating various strategies to manage social risk, the framework provides the analytical foundations for formulating social protection approaches and policies. The purpose of this book is to provide guidelines for countries wishing to introduce or improve income support programs for the unemployed. 1 2 Income Support for the Unemployed: Issues and Options The book focuses on five such programs: unemployment insurance, unem- ployment assistance, unemployment insurance savings accounts (UISAs), sev- erance pay, and public works. To arrive at guidelines, the book summarizes the evidence on the performance of different types of programs in four areas: how well they protect the incomes of the unemployed, how they affect efficiency, how able they are to confront different types of shocks, and how resistant they are to political interference. Based on this evaluation, the suitability of particu- lar programs for developing and transition countries is assessed, taking account of countries' specific circumstances, chief among them labor market and other institutions, the capacity needed to administer income support pro- grams, the size of the informal sector, and the prevalence of private transfers. While the book provides a global perspective and reviews programs in countries around the world, its relevance and lessons are geared toward transition countries and middle-income developing countries. Although these countries may offer various income support programs (most often sev- erance pay and public works), the unemployed remain inadequately pro- tected: coverage is limited, programs may suffer from nonperformance, and the programs may impose heavy litigation costs and significant inefficien- cies. As these countries develop, the issue of improving public income sup- port programs--particularly by introducing social insurance--inevitably emerges. Indeed, many transition countries have recently introduced unem- ployment insurance programs, and some middle-income developing coun- tries are contemplating doing so. For these countries, this book will provide valuable guidance by evaluating the choices and tradeoffs between various income support programs. In many low-income countries, social insurance­ type programs are not a likely option in the foreseeable future. Some of the issues addressed in the book will thus not be relevant for these countries. The book's main conclusions can be summarized as follows: · Unemployment insurance enables a high degree of consumption smooth- ing, performs well under various types of shocks, and acts as an auto- matic macroeconomic stabilizer, but it creates reemployment disincentives and wage pressures, which increase the equilibrium unemployment rate and make unemployment persistent. Because its successful performance depends on conditions that are typically lack- ing in developing and transition countries, the case for introducing unemployment insurance in these countries is less compelling than it is in industrial countries. · Unemployment assistance allows more effective targeting than unem- ployment insurance, but it may not produce savings and it offers a lower level of protection for high-income workers, imposes larger administrative costs, and suffers from similar employment disincen- tives. Its applicability is thus limited, perhaps to countries with rela- tively developed administrative capacity and a small informal sector. · Unemployment insurance savings accounts are a potentially interesting option. By internalizing the costs of unemployment benefits, UISAs ameliorate the moral hazard inherent in traditional unemployment Introduction 3 insurance and thus improve reemployment incentives--an impor- tant advantage given the weak monitoring capacity of developing countries. By integrating the accounts with social insurance--thus circumventing the main weakness of not pooling risk among individuals--UISAs may improve work incentives while in principle offering income protection comparable to traditional unemployment insurance. UISAs also have the potential to attract informal sector workers. By allowing individuals to borrow from their accounts, some versions of integrated accounts introduce problems of their own, however, creating incentives to withdraw from the formal sector in order to avoid repaying the debt. Because the program has been largely untested, further investigation of its effects and design param- eters, including piloting of the program, is needed. · Severance pay is an important program, because it is available in many developing and transition countries, but it offers few advantages. To improve its protection and efficiency effects, countries may consider streamlining severance pay programs and reducing their costs if they are too generous. A more radical reform could introduce prefunding, to improve nonperformance and reduce labor market rigidities. · Public works programs are effective in reaching the poor, have good targeting properties and substantial capacity to redistribute income from the rich to the poor, are able to attract informal sector workers and provide flexible and fast response to shocks, and are administra- tively less demanding than other public income support programs. Despite weaknesses--high nonwage costs, the likely countercyclical pattern of funding, and, in some countries, stigmatization of partici- pants--public works programs are suitable for developing countries, particularly as a complementary program. The book is organized as follows. Chapter 2 examines the conceptual issues that arise in evaluating and designing income support programs for the unemployed. It describes their objectives and discusses the need to evaluate them in a broader, social risk management framework that allows for various interactions. Based on this framework, two sets of evaluation criteria are developed, which are used to assess the applicability of various programs to developing and transition countries. Performance criteria evaluate the income protection and efficiency effects of these programs; design and implementa- tion criteria evaluate how different programs fit the specific circumstances of developing and transition countries. Chapter 2 also explains why unemploy- ment insurance is usually provided as part of social insurance and addresses definitional and measurement issues of unemployment. It notes the historic conditions responsible for the emergence of unemployment and draws impli- cations about the suitability of transferring unemployment support programs from industrial countries to developing and transition countries. Chapter 3 describes key income support programs for the unemployed and describes their stylized design features. It emphasizes the richness of the approaches and the complexity of the programs, highlighting important 4 Income Support for the Unemployed: Issues and Options features that should be considered when introducing or improving such pro- grams. It also analyzes the incidence of the main types of income support pro- grams for the unemployed throughout the world. It shows that industrial countries base their income protection on social insurance but also provide other unemployment support programs, while some low-income countries have no special programs for the unemployed. The chapter also shows that parameters of a particular income support program differ sharply across countries, contribut- ing to differences in coverage and the degree of protection provided. Chapter 4 evaluates the performance of the selected income support pro- grams, based on a review of theoretical predictions and empirical evidence. It focuses not only on income protection effects but also on efficiency effects and the ability of such programs to confront various types of economic shocks and resist political risk. It shows that, through various channels and due to complex interactions, these programs generate a wide variety of effects, some intended and some unintended, some anticipated and some unanticipated. It also shows that no program outperforms all others: one program may offer superior insurance properties, but it may create labor market disincentives and generate an "entitlement mentality" that makes reform of the program difficult. When introducing or improving income support programs, countries should also account for country-specific circumstances, because they cru- cially determine how well a program fits a particular country. Chapter 5 is therefore devoted to the circumstances prevailing in developing and transi- tion countries. It reviews their key institutional and labor market features, as well as the capacity needed to administer such programs. It shows that these countries may deviate significantly from the typical labor market and other institutional features under which income support programs function in industrial countries and that developing countries may also lack the capa- bilities necessary for the smooth and effective administration of some pro- grams. Chapter 5 also shows that the desirability of alternative income support programs depends crucially on the interactions of these programs with other social risk management mechanisms and that these interactions may be very different in developing and transition countries. Building on the discussion of previous chapters, Chapter 6 evaluates alter- native income support programs and provides tentative guidelines for improv- ing these programs in developing and transition countries. Drawing on the previously established performance effects, the chapter describes the strengths and weaknesses of each type of income support program and points to country- specific circumstances that are particularly conducive to the performance of each. It also examines some important design features of unemployment insur- ance and options for improving income support for informal sector workers. Note 1. An important recent contribution to the analysis of policy options for income support for the unemployed, focusing on the Latin American context, is by de Ferranti and others (2000). 2 Conceptual Issues Faced with the risk of unemployment, individuals choose among a variety of risk management mechanisms. Some try to get a good education or enter jobs that are known to be stable, in order to reduce the risk of becoming unemployed. Others may accumulate real or financial assets or participate in unemployment insurance programs, in order to have financial means at hand if unemployment occurs. Yet others may rely on private transfers of cash, food, and clothing; draw down financial or real assets; participate in public works or public training; or receive social assistance, in order to cush- ion the loss of earnings associated with job loss. How should the desirability of such mechanisms be judged from the viewpoint of society? Successful smoothing of consumption is important, but so are other considerations. Do public programs displace other mecha- nisms, formal or informal? Do they affect job search effort and the type of post-unemployment job? How successful are they in reaching the hardest- hit segments of the population and the very poor? What are the tradeoffs between pure income transfers and programs that combine transfers with other requirements--and opportunities--such as public works or training? How can the loss of human capital associated with prolonged unemploy- ment spells be prevented? This chapter provides the conceptual underpinnings of the approach the book takes in evaluating income support programs for the unemployed and describes the rationale for public interventions. The starting point--and the recurrent theme of the book--is the recognition that public income support programs for the unemployed are just a subset of social risk management mechanisms. It is of utmost importance, therefore, to look at the system of social risk management in its entirety, in order to consider links among its components and assess the repercussions of introducing new public pro- grams on other mechanisms. The richness of the mechanisms and strategies available to individuals, families, and communities is staggering. To be tractable, theoretical models must thus focus on specific aspects of income support. The robustness of their conclusions must therefore be verified to insure that they are applicable under the circumstances prevailing in a 5 6 Income Support for the Unemployed: Issues and Options particular country. So, too, must empirical findings obtained in industrial countries be checked to make sure that they hold under the circumstances prevailing in developing and transition countries. Because of the complexity of interactions and the lack of generality of theoretical results, as well as the dearth of empirical studies on income sup- port programs in developing and transition countries, the book developed two sets of criteria with which to judge the desirability of income support programs in a particular country. The first set consists of performance crite- ria, the second of design and implementation criteria. Performance criteria relate to the effects of income support programs predicted by theoretical models and validated by empirical studies. These criteria evaluate programs not only based on how effectively they provide income protection but also based on how they affect efficiency, how suitable they are to confront differ- ent types of shocks, and how resilient they are to political risk. When transferring the experience of one country to another, policymak- ers need to take account of "initial conditions"--the particular features of the country in question. This is particularly important given that theoretical models are built on many simplifying assumptions and that empirical stud- ies have been conducted largely on industrial countries, where conditions may be very different from those in developing and transition countries. The second set of criteria, the design and implementation criteria, therefore include country-specific features, such as the nature of labor market institu- tions, the administrative capacity of the country to deliver specific pro- grams, the characteristics of the unemployed, and the size of the informal sector. The degree of informality of the economy determines how many people can take advantage of formal sector programs (such as unemploy- ment insurance and severance pay); the administrative capacity of the econ- omy is a good predictor of the success of programs that require strong monitoring or information capacity. This chapter also provides the rationale for public provision of unem- ployment insurance as part of social insurance, while continuing to apply the social risk management framework. Markets provide insurance only under certain circumstances--most notably, in the absence of informational asym- metries and the presence of certain characteristics of the risk--circumstances that are violated in the case of unemployment insurance. Hence there is room for government interventions, which usually consist of mandating membership or providing insurance through the public sector. The chapter also examines the interrelation between unemployment insurance and development of the labor market, emphasizing that unemployment is a special social construct that emerges only in industrial societies and that its emergence, in turn, creates demand for social insurance. The chapter is organized as follows. After describing the objectives of income support programs for the unemployed--compensating for lost wages, stabilizing employment, helping the unemployed obtain better jobs more quickly--the chapter places income support programs in the context of social risk management. The interaction with other risk management Conceptual Issues 7 mechanisms--informal and formal, private and public--has an important bearing on the success of these programs and hence on their desirability. Although theoretical modeling of such interactions provides significant insights, the fact that it simplifies these interactions and imposes other con- straints means that can provide only limited recommendations on policy. The suitability of alternative income support programs is therefore evaluated using an approach that combines performance evaluation criteria (stressing income protection and efficiency aspects, as well as suitability to confront shocks and resistance to political interference) with design and implementa- tion criteria (stressing countries' specific conditions). (In subsequent chapters, these analytical instruments are used to evaluate various income support pro- grams and assess their desirability for developing and transition countries.) The chapter concludes by describing the rationale for public provision of unemployment insurance and discussing the appropriateness of income support programs for developing and transition countries. Objectives of Income Support Programs for the Unemployed From the viewpoint of the individual, the objective of income support programs for the unemployed is precisely that implied by their name-- compensation for income loss due to unemployment. Such programs pro- vide income protection to workers by partly replacing their lost wages with social insurance payments (under unemployment insurance and unemploy- ment assistance as well as under some variants of unemployment insurance savings accounts), lump-sum payments upon separation (severance pay), or wages obtained from publicly generated employment (public works). Such payments help cushion the reduction in consumption in the wake of job loss--that is, they smooth consumption--and may also help alleviate poverty. To varying degrees, other objectives also figure prominently, particularly if these programs are viewed from a broader, societal perspective. These additional objectives, which relate to both efficiency and fairness, include the following: · Unemployment insurance: Stabilizing employment and preventing unemployment by discouraging layoffs and restoring purchasing power, encouraging workers to acquire new skills, providing work- ers with the means to conduct more extensive and more effective job searches, redistributing income from high- to low-wage earners, and increasing output and efficiency (by promoting restructuring of enterprises or encouraging workers to accept higher wage and thus higher-productivity, but also riskier, jobs, for example). · Unemployment assistance: Similar to the objectives of unemployment insurance, but because unemployment assistance is a means-tested program, the emphasis is on poverty reduction rather than consump- tion smoothing. 8 Income Support for the Unemployed: Issues and Options · Unemployment insurance savings accounts: Similar to the objectives of unemployment insurance. Because the program design encourages work incentives, the relative importance attached to these objectives may be different, however. In particular, the program leaves less scope for income redistribution. · Severance pay: Stabilizing employment and preventing unemployment by discouraging layoffs. In voluntary schemes, the objective of the firm is to encourage long-term relationships in order to retain valu- able workers, as well as to reduce transactions costs due to the loss of skills and firm-specific knowledge generated by labor turnover. · Public works: Providing goods and services for the poor, maintaining and creating human capital, countering the psychological effects of unemployment, reducing income inequalities and countering social exclusion, and fostering community development and empowerment. The debate preceding the introduction of unemployment insurance in the United States in the 1930s illustrates the opposing views about the objec- tives of unemployment insurance. The view that eventually prevailed emphasized consumption smoothing (Blaustein 1993). This view attached great importance to the principle that compensation be provided as a matter of right and not as a means-tested benefit. But two other important schools of thought were prominent. One emphasized preventing unemployment by discouraging layoffs. It claimed that employers should be held responsible for unemployment of their workers by bearing the costs of unemployment compensation, through "experience rating." This view has retained ground in the United States to the present day. The other school of thought pro- moted faster reemployment, claiming that unemployment insurance pro- grams should not only provide resources to the unemployed but also help them find jobs in other ways. Some of the stated objectives of these programs may fail to materialize-- and may indeed backfire. For example, the stated objective of severance pay of stabilizing employment and preventing unemployment should be con- trasted with its widely established effects on employment: while the program does stabilize jobs of incumbent workers, it reduces overall employment rates (see chapter 4). Similarly, although public works are sometimes justified on the grounds that they help the unemployed counter the psychological effects of unemployment, these programs often stigmatize participants, thus reducing their job prospects and hurting their self-image. These examples highlight the importance of accounting for the complex interactions typically at work in providing public income support to the unemployed. Accounting for Interactions among Social Risk Management Mechanisms The risk of unemployment is best studied in the context of social risk management, the framework that recognizes that many risk management Conceptual Issues 9 mechanisms--informal and formal, private and public--exist and interact in many important ways. Recognizing and appropriately accounting for such interactions is critical to selecting and designing the most appropriate public income support programs. As shown below, formal treatment of interactions provides significant insights yet of necessity abstracts from many interactions and simplifies institutional features. Social Risk Management Framework Public income support programs are a subset of risk management mecha- nisms available to the unemployed. These mechanisms can be divided into three categories: those that reduce the risk of unemployment (reduce the probability of becoming unemployed or increase the probability of leaving unemployment if unemployed), those that mitigate that risk (reduce the impact of a future unemployment spell), and those applied in response to the undesirable event (coping mechanisms).1 Within all three categories, informal and formal mechanisms are usually available. Formal mechanisms include both market-based and public mechanisms (table 2.1). The social risk management framework enables one to position public income support programs for the unemployed in the context of other mech- anisms that may be used to manage the risk of unemployment. Viewing the program in its entirety and considering links among various components is useful when evaluating the performance of public programs or assessing the effects of their introduction. Among the diverse interactions that must be considered are the following: · General equilibrium and growth interactions. The introduction of unem- ployment insurance may encourage the emergence or expansion of more risky jobs or industries, which conceivably could increase effi- ciency (Acemoglu and Shimer 1999, 2000). Income support programs may also have dynamic effects, including effects produced by the interaction of unemployment benefits with adverse shocks to the economy. As discussed in chapter 4, these interactions typically slow the reduction of unemployment to the shock and thus contribute to the persistence of unemployment. In contrast, in their role as an auto- matic macroeconomic stabilizer, unemployment benefits soften the impact of adverse shocks on GDP--but they also restrain expansion when the economy starts growing again. · Interactions with private risk management mechanisms. There is a great variety of such interactions, many having important long-term conse- quences. For example, participation in a public income support pro- gram may increase investments in human capital (such as enabling better education of family members and improving health) and reduce the use of undesirable coping strategies (such as selling cattle essential for providing important nutrients). But public programs-- as well as private transfers received by participants--may reduce participants' incentives to save or to enroll in training. Table 2.1. Formal and Informal Income Support Mechanisms for the Unemployed Informal Formal mechanisms Strategy mechanisms Market based Public Risk · Less risky · Training · Sound macroeco- reduction production · Financial nomic policies · Migration market literacy · Public health policy · Proper feeding · Company- · Labor market and weaning based and policies (including practices market-driven employment · Maintainence of labor standards protection good health policies, such as severance pay) Risk mitigation Portfolio · Multiple jobs · Investment · Multipillar diversification · Investment in multiple pension systems in human, financial assets · Asset transfers physical, and · Microfinance · Protection of real assets programs property rights · Investment in · Support for social capital extending financial (rituals, recipro- markets to the poor cal gift-giving) Insurance · Savings (in form · Old-age · Unemployment of financial or annuities insurance other assets) · Disability, · Unemployment · Marriage, family accident, and assistance · Community other personal · Unemployment arrangements insurance insurance savings · Sharecrop · Crop, fire, and accounts tenancy other damage · Pensions insurance (including early retirement), disability, and sickness insurance Risk coping · Selling of real · Sale of financial · Social assistance assets assets · General subsidies · Borrowing from · Commercial · Active labor neighbors loans market programs · Intracommunity (job-search transfers and assistance, training, charity employment subsi- · Child labor dies, public works) · Dis-saving in · Social funds human capital Source: Adapted from World Bank (2001a). Note: Major public mechanisms used to manage the risk of unemployment are shown in bold. 10 Conceptual Issues 11 · Interactions with other policies and programs. Important links also exist among public policies and programs themselves. The financing of social insurance typically requires contributions from both employ- ers and employees, thus creating a wedge between the wage received and labor costs. To the extent that this wedge reduces labor demand, the introduction of unemployment insurance contributes to higher unemployment. (The increase in the reservation wage in response to higher unemployment benefits has the same effect.) Similarly, increas- ing the generosity of severance pay is likely to reduce employment rates and slow labor market flows, particularly from employment to unemployment but also from nonemployment to employment. Thus the introduction of a risk mitigation mechanism (such as social insur- ance) or a risk reduction mechanism (such as severance pay) may increase the unemployment rate or negatively affect the job creation capacity of an economy, thereby worsening the effectiveness of other risk management mechanisms. The social risk management framework thus makes clear not only that there are multiple mechanisms for dealing with the risk of unemployment but also that there are complex links and interactions among them. Recognizing-- and appropriately accounting for--such interactions is critical if policymakers are to choose and design the most appropriate public income support pro- grams. For example, when a low-income country is considering improving its public income support for the unemployed, it must take diverse and far- reaching implications into account. These range from the impact on individu- als' self-protection mechanisms (including changes in the intensity of job search and the propensity to enroll in training) to the effects on labor market outcomes (including the unemployment rate and the intensity of labor market flows) to the capacity to administer a particular program (box 2.1). Guidance and Limitations of Formal Modeling One rich source of policy advice is economic theory, which provides insights into the working of income support programs and the implications of such programs for economic policy (see chapter 4). Ehrlic and Becker (1972) propose a comprehensive insurance theory (box 2.2).2 They conclude that relatively rare and large losses are better insured through market insurance, while relatively frequent and moderate losses are better insured though self-insurance. Extending the comprehensive insurance approach, Gill and Ilahi (2000) provide the following insights about the rationale of and guidelines for social policies: · The lack of insurance offered by the market provides a rationale for its public provision. · One of the best means of self-protecting is investing in human capital, but doing so provides poor collateral. By subsidizing the acquisition 12 Income Support for the Unemployed: Issues and Options Box 2.1. Recognizing Interactions among Different Risk Management Mechanisms In dealing with the risk of unemployment, Filipino workers employed in the formal sector rely on severance pay (although it may be difficult to obtain). These workers are better off than informal sector workers, who have little or no protection. It is therefore not surprising that Filipino workers have relied heavily on informal mechanisms to manage unemployment risk, many of which are costly, inefficient, and above all inadequate. How can public policy be improved to help Filipino workers deal more effectively with unemployment? The social risk management framework sug- gests that the answer should rest on the following considerations, among others: · How well does the program fit into existing risk management mecha- nisms? Would its introduction disrupt existing self-protection mecha- nisms or displace existing coping mechanisms (such as a program of private transfers, especially for the nonpoor population) that may be more efficient than public programs? Is the program well attuned to the prevailing norms and culture? Can existing institutions be "upgraded" to provide better protection and increased coverage? · How do the likely beneficiaries compare with other population groups? What are the likely effects of the program on income redistribution and poverty reduction? · Is the program compatible with other public support mechanisms and policies? Above all, does it promote labor reallocation and job creation as sources of productivity growth? · How well does the program respond to the country's income shocks, such as economic recessions, structural imbalances caused by liberal- ization and globalization, and shocks caused by natural calamities? · Is the program well attuned to local circumstances, so that it functions well? Can it be supported with the existing administrative capacity? Are there mechanisms that allow participants to be selected effectively? Source: Esguerra, Ogawa, and Vodopivec (2001). of human capital, the government can mitigate the resulting tendency to underinvest. · Private agents may self-insure using "bad" instruments (using cattle or land as a medium of precautionary saving, for example) because "good" instruments (such as diversified financial assets) are not available. · The government should foster the development of insurance and financial markets. Strengthening of financial markets should be a cen- tral component of social policy, because it can foster self-insurance, market insurance, and self-protection (through prudential regulation of capital markets, for example). Conceptual Issues 13 Box 2.2. Insights from the Theory of Comprehensive Insurance Ehrlic and Becker (1972) model a utility-maximizing individual deciding how much to spend on three risk management instruments: market insurance, self- insurance, and self-protection. (Individuals insure by transferring income from the good to the bad state; they self-protect by taking actions that reduce the probability of the bad state). They draw several important conclusions: · Market insurance and self-insurance are substitutes, as are self- insurance and self-protection; market insurance and self-protection may be substitutes or complements. An important implication is that the existence of market or social insurance may not reduce self- protection and thus produce a moral hazard problem. If, for example, more intense job search is rewarded by lowering the unemployment insurance premium, moral hazard is not inevitable. (Note that this assumes that self-protection behavior is detectable.) · Relatively rare and large losses are better insured through market insurance; relatively frequent and moderate losses are better insured though self-insurance. This conclusion follows from the fact that while the price of market insurance is lower if the probability of loss is lower, the (shadow) price of self-insurance (for example, the cost of precau- tionary saving) does not vary with the likelihood of loss. · Individuals enjoy higher welfare when all three instruments are avail- able than when one of them is missing. · Introducing a social safety net reduces self-protection but not neces- sarily self-insurance. The richness and complexity of the issues that arise in the context of social risk management means that formal modeling has only a limited abil- ity to capture all relevant aspects. Three problems in particular stand out. First, formal models simplify interactions among risk management mecha- nisms and abstract from general equilibrium effects. They focus on limited interactions and thus generate results that apply under specific circum- stances. Once richer possibilities of interactions are introduced, they may recast the results in a different light. For example, recent research has shown that allowing interactions between unemployment benefits and the deter- mination of wages--an issue that previous modeling had not taken into account--has important implications for the optimal time schedule of unem- ployment benefits (Cahuc and Lehmann 2000). Allowing for possibilities of private transfers and the existence of individual savings also affects income support modeling, yet these issues have only recently been considered in formal analysis. Second, formal modeling imposes restrictive assumptions about the behavior and characteristics of economic agents and their knowledge, sacri- ficing generality for analytical tractability. For example, theoretical models of labor reallocation that explicitly treat job creation and destruction processes 14 Income Support for the Unemployed: Issues and Options are often analytically intractable, forcing researchers to use calibration models that yield solutions under less general conditions.3 Simplifications include using specific functional forms of utility and other important economic vari- ables; assuming that individuals are homogeneous or allowing heterogeneity in only one trait; and (unreasonably) assuming what kind of information is available to certain agents (for example, that the government or insurer is able to observe abilities or self-protecting activities of insured workers). Third, formal modeling disregards important institutional features, including features of income support programs themselves. Atkinson and Micklewright (1991, p. 1706) complain that "the great generality of research reaching conclusions about unemployment compensation has paid scant attention to the institutional details, and some elements have been almost totally ignored.... The importance of the institutional aspects is a matter on which we would like to insist." This argument applies with particular force once policy implications of theoretical models are applied to developing countries or OECD-style programs are transferred to developing countries; in doing so, important institutional features and other social risk manage- ment mechanisms may be ignored or neglected. In particular, the whole notion of unemployment in developing countries is very different than in industrial countries. Self-protection as well as informal insurance mecha- nisms are more prevalent, and initial conditions--administrative capacity, existence of financial markets, prevailing culture--are very different. To summarize, theoretical modeling offers important, increasingly complex, and refined insights into the working of public income support programs. But such models can provide only partial guidance for policymakers--particularly in crafting policy for developing and transition countries, where the implicit assumptions (such as the existence of suitable institutional capacity and strict enforcement of law and order) often do not apply. Moreover, theoretical modeling of income support programs rarely incorporates long-term considerations such as investment in education or health, which are especially important in developing countries. Establishing Evaluation Criteria Given the complexity of interactions, the lack of generality of theoretical results, and the dearth of empirical studies on income support programs in developing and transition countries, how can income support programs for the unemployed be evaluated? Two sets of criteria are used here. The first are performance criteria, which assess how well a program works (how well it protects incomes and what other effects it may have). The second are design and implementation criteria, which assess how well a program fits a particular country (how suitable it is given country-specific conditions). This approach combines and reconciles theoretical insights about the performance of such programs with the empirical evidence and country- specific conditions that affect the applicability of programs to developing and transition countries. Conceptual Issues 15 The need for a synthetic approach that provides global perspectives and broad comparisons stems from the pressing and difficult choices faced by developing and transition countries. As shown in chapter 3, almost all indus- trial countries have unemployment insurance programs, while most other countries do not. As the economies of these countries develop, the question of improving public income support programs will inevitably emerge. Indeed, many countries (for example, all European transition economies, as well as China and Turkey), have recently introduced unemployment insur- ance programs, and some developing countries, including low-income ones (the Philippines, Sri Lanka, Thailand) are contemplating doing so. For these countries, evaluating the choices and tradeoffs between diverse income support programs--which do not lend themselves easily to rigorous modeling in a unified framework--is crucial. The basis for distinction and the underlying principle of determination of the two groups of criteria are based on the usual role of variables that repre- sent the criteria in economic investigations. Economic models often treat and formally specify the effects of an income support program (the "dependent" variables) as a function of its design features and initial (country-specific) conditions (sometimes called the "control" variables). Dependent variables of such economic models are assigned the role of performance criteria. These dependent variables typically reflect the effects of the program on income protection, income redistribution, and efficiency. In contrast, the control variables--a subset of right-hand-side variables that account for differences across countries when evaluating programs' cross-country performance-- are assigned the role of design and implementation criteria. Performance Evaluation Criteria The performance of income support programs for the unemployed is mea- sured by their success in achieving their objectives, as well as by other intended and unintended effects and features. Four subsets of performance criteria are assessed: protection criteria, efficiency criteria, suitability to con- front shocks, and resistance to political risk. Coverage and adequacy of support are examined to determine protec- tion effects. The effects of income support programs on income redistribu- tion are also analyzed, as these programs--intentionally or not--often involve substantial redistribution. A variety of measures are examined to assess efficiency effects, including the intensity of job-search effort, post- unemployment wages, equilibrium labor market outcomes, and effects on programs on output and growth. Suitability to confront economic shocks looks at how suitable different programs are for, say, a country that is fre- quently plagued by sudden, regionally concentrated shocks due to natural disasters or by long-lasting, covariant shocks. Resistance to political risk emphasizes political economy considerations in providing income support, such as which circumstances are conducive to the introduction of various programs and how to raise support for reforms to improve such programs. 16 Income Support for the Unemployed: Issues and Options Suitability to confront shocks and resistance to political risk are singled out in order to emphasize their role as indicators or predictors of both effi- ciency and protection performance. If a program is ill-suited for a particular shock or political risk, it means that it does not offer good protection or involves large efficiency costs. These additional criteria could have been ignored and the effects evaluated by them considered under the respective rubrics of protection and efficiency. They are treated separately in order to emphasize their importance. Design and Implementation Criteria Programs do not operate in a vacuum: country-specific circumstances affect their performance and thus determine their suitability to meet the specific needs of a country. In addition to performance criteria, which eval- uate income support programs based on their performance under typical conditions (conditions that usually prevail in industrial economies), our approach introduces design and implementation criteria, which reflect specific features of the country under investigation. These features include the following: · interactions with labor market institutions and shocks · administrative capacity for program implementation · characteristics of the unemployed · size of the informal sector · prevalence and pattern of interhousehold transfers · ability of individuals to self-insure and self-protect · nature of shocks · cultural and political factors To illustrate: in a country in which certain regions are often affected by natural disasters, income support programs must be flexible and capable of being quickly deployed in affected areas. Where the informal sector is large, programs must be accessible to the self-employed and other informal sector workers. When a program is chosen, a country's administrative capacity has to be taken into account. Unemployment insurance or assistance requires monitoring of recipients (to ensure compliance to continuing eligibility rules); unemployment assistance relies on means testing. The performance of such programs depends crucially on the administrative capacity to pro- vide quality monitoring and testing. Interactions with other programs and policies are also important. For example, to avoid incentive incompatibility, unemployment insurance savings accounts need to be harmonized with old-age income support programs to preclude scenarios in which the accounts are depleted in anticipation of forbearance and generosity on the part of the pension system. Introducing or increasing the generosity of unemployment benefits may have different effects in an economy with dif- ferent levels of centralization and coordination of wage bargaining: under fragmented and uncoordinated bargaining, unemployment benefits are Conceptual Issues 17 more likely to increase wage pressures and hence the equilibrium unem- ployment rate than under alternative arrangements (see chapter 4). Chapter 5 is devoted to the implications of country-specific circumstances for the choice and design of income support programs. Why Should the Public Sector Provide Unemployment Insurance? Some studies cast doubt on the welfare benefits of public programs, because such programs may displace old mechanisms for dealing with unemploy- ment risk. While precise conditions for the introduction of public programs are difficult to pinpoint, experience shows that these programs can provide important benefits. Informal insurance mechanisms may often be ineffec- tive, because the loss of employment is too large a shock and may occur too frequently. As Murdoch (1999) points out, informal insurance (such as re- ciprocal and other forms of transfers) tends to be least effective when insur- ance is most needed. Moreover, in poor countries the beneficiaries of private transfers are the elderly; keeping more income for themselves would enable the young--who are often also poor--to obtain more education. Providing formal income support may also improve fairness (many informal insur- ance mechanisms discriminate against women, for example). If informal insurance mechanisms are not satisfactory and unemploy- ment insurance is desirable, why does the market not provide it? Unem- ployment insurance is typically provided publicly as a part of social insurance primarily because of strong information asymmetries--moral hazard and adverse selection problems--that prevent markets from provid- ing such insurance. Moral hazard arises because unemployment insurance reduces self-protection; adverse selection arises because information prob- lems prevent insurers from charging bad risks higher premiums. Public pro- grams are also able to pool resources across larger groups than private insurance could. Correcting for market failures calls not only for regulation-- obligatory membership to avoid adverse selection--but also for public pro- vision, to improve the monitoring capacity and financial sustainability of the program. How does unemployment insurance work? By pooling risk across indi- viduals, insurance shifts resources from a good to a bad state (compensating unemployed workers by collecting contributions from employed workers, for example). By offering a certain outcome, insurance can increase welfare. For a risk-averse individual, expected utility from uncertain outcomes falls short of the utility provided by insurance that eliminates uncertainty and provides certain, unconditional income, as long as the price of the insurance falls within certain limits.4 Individuals are therefore prepared to pay to reduce the uncertainty, even when expected payout by the insurance com- pany is less than the premium paid. In addition to the welfare gains produced by reducing uncertainty, insur- ance can increase efficiency, for several reasons. First, insurance can stimu- late the emergence of riskier but more productive jobs and industries 18 Income Support for the Unemployed: Issues and Options (Acemoglu and Shimer 1999, 2000). Second, insuring transient shocks that reduce individual consumption below a threshold needed to retain produc- tivity can prevent "dynamic poverty traps" that lead to chronic poverty (Ravallion 2003). Third, uninsured risk reduces efficiency through costly production and portfolio choices, such as using outdated but less risky pro- duction technologies or holding livestock as a form of precautionary savings (see the example of Indian farmers in Rosenzweig and Wolpin 1993). Fourth, uninsured risk can adversely affect human capital accumulation, as children drop out of school in the wake of an income shock, for example. Markets may fail to provide insurance, or they may provide it ineffi- ciently. For the market to efficiently provide insurance, the following condi- tions must prevail:5 · Risk pooling among a sufficiently large number of individuals must be possible. Such pooling translates uncertain individual outcomes into relatively certain aggregate outcomes, thanks to the law of large numbers. · The risks of the individuals insured must be noncorrelated. If the dis- tribution of risk is not correlated, in each period there are a pre- dictable number of individuals who suffer losses and those who do not, enabling a smooth operation of the scheme. · The probability of the unfavorable event occurring must not be great (otherwise, insurance companies would be unable to survive). · The probability of an unfavorable event must be known or pre- dictable. This allows the insurance company to set the insurance premium. · The problems arising from asymmetric information (adverse selec- tion and moral hazard problems) must not be serious.6 As the existence of numerous insurance markets suggests, markets often overcome the special problems connected with providing insurance. This is not the case with unemployment insurance: in virtually all countries it is publicly provided as a part of "social insurance," which usually includes old-age, health, and disability insurance as well. Why is this the case? First, informational problems--adverse selection and moral hazard-- have a strong effect on unemployment insurance. Adverse selection arises because insurers cannot determine the unemployment risk with sufficient accuracy (work history is an imperfect and in some cases nonexistent pre- dictor). Because insurance premiums do not reflect an individual's risk, low- risk individuals have an incentive not to insure. An even more difficult problem is created by moral hazard: insured workers may reduce their work effort when employed and their job-search effort when unemployed and receiving unemployment benefits, without the knowledge of the insurance company. Second, unemployment risks are covariant and thus cannot be diversi- fied by private insurance. A severe recession may dramatically increase the number of claimants and jeopardize the existence of private insurers. Conceptual Issues 19 Third, the chance of (re)employment of some groups of unemployed-- such as workers close to retirement age living in areas with high unemploy- ment--may be so low that private insurance would not be possible. Fourth, society can adjust unemployment insurance premiums in order to promote equity (keeping in mind that unemployment insurance premi- ums cannot be actuarially set because of information problems). For these reasons, unemployment insurance is typically provided as part of social insurance--a program that mimics market insurance arrangements but deviates from actuarial principles. Workers and their employers pay earnings-related contributions that, upon separation, entitle workers to unemployment benefits according to predetermined eligibility conditions. To deal with informational problems, social insurance employs two mecha- nisms. To correct for adverse selection, membership is compulsory (if not economywide, then within certain sectors). To address the moral hazard problem (a challenging task not only for private but also for public agen- cies), in addition to directly monitoring claimant's behavior, it relies on information from other public programs when checking eligibility status of benefit claimants--an option unavailable to private providers. Moreover, to correct for covariance of unemployment risk, the government provides financial backing for the program (when the program's current obligations exceed its current contributions and accumulated reserves). Who Should Receive Unemployment Income Support? Unemployment is a special social construct that emerges only in industrial societies. Because unemployment arises from the development of the labor market, income support programs for the unemployed need to be attuned to labor market conditions and the characteristics of unemployed workers. As Atkinson (1995a) persuasively claims, social insurance is a response by the modern, industrial society to the changing nature of the labor market, above all to the development of a modern employment relationship. In an urban, industrial society, employment becomes discretionary: workers either work or do not work. This development has important implications for unemployment. If workers cannot find a job in wage employment, they are unable to resort to self- or home-production, because they are divorced from ownership of the means of production.7 Similarly, older and less pro- ductive workers in industrial societies stop working altogether once their productivity drops substantially. In contrast, in traditional societies workers continue to be economically active as long as they produce something valu- able. The emergence of unemployment and the changing nature of retire- ment create the need to insure against nonemployment.8 In other words, social insurance is based on the concepts of unemployment and retirement as specific social constructs. That the notion of unemployment differs substantially in developing and industrial countries is reflected in measurement problems. The stan- dard International Labour Organization (ILO) definition of unemployment 20 Income Support for the Unemployed: Issues and Options Box 2.3. ILO Definition of Unemployment According to the International Labour Organization definition of unemploy- ment, the unemployed include all people above a specified age who, over a specified reference period, are: · "without work," that is, are not in paid employment or self-employ- ment · "currently available for work," that is, are available for paid employ- ment or self-employment during the reference period · "seeking work," that is, are taking specific steps in a specified recent period to seek paid employment or self-employment. The specified steps may include registering at a public or private employment exchange; applying for jobs with employers; checking at work sites, farms, factory gates, markets, or other assembly places; placing or answering newspaper advertisements; seeking assistance from friends or relatives; looking for land, building, machinery, or equipment to establish an enterprise; arranging for financial resources; and applying for permits and licenses. Source: Resolution I of the 13th International Conference of Labour Statisticians, Geneva, October 1982. applied to developing countries classifies some unemployed as employed or inactive and some as inactive (box 2.3). Many workers in developing countries who qualify as employed under the ILO definition are in fact underemployed, especially in rural areas.9 These workers may work fewer hours than they would like or work in low- productivity jobs and earn low wages. But because they are so poor that they cannot afford to be without a job, workers in developing countries are rarely unemployed. Some unemployed may be classified as inactive. People with a marginal attachment to the labor force--that is, people who are available for and desire work but who are not actively seeking it because they perceive, rightly or wrongly, that no jobs are available--are often considered econom- ically inactive when they should be more appropriately classified as unem- ployed (such workers are sometimes called "discouraged workers").10 Moreover, the conventional application of the term "actively seeking work" falters in light of the fact that much economic activity occurs through infor- mal employment or self-employment arrangements. Some employed workers may be classified as inactive. According to the ILO guidelines, an individual who works at least one hour in a week or who is temporarily absent from work (because of vacation or illness, for example) is employed. Those who are out of work but do not meet the ILO criteria of unemployment are classified as economically inactive. However, some Conceptual Issues 21 forms of informal economic activity--including home-based work, typically undertaken by women--may escape this definition of employment. Because such workers are not available for work, they do not qualify as unemployed. Acknowledging the possible restrictiveness and industrial country bias of its definition, the ILO advises to relax these clauses and formulate criteria suitable to the labor market characteristics of developing countries. Supple- mental measures of labor market slack, which may be very important in transition economies and developing countries, include measures of under- employment and discouragement. For example, throughout the 1990s the Philippines had about twice as many underemployed worker as unem- ployed workers (Esguerra, Ogawa, and Vodopivec 2001). In Bosnia and Herzegovina in 2001, there was one discouraged worker per every 2.5 unemployed workers (World Bank 2002). Reflecting historic conditions, the whole notion of unemployment therefore rests on the characteristics of the society and the labor market. In developing countries, where unemployment is rare because individu- als cannot afford to be unemployed, self-protection (risk reduction) emerges as the dominant strategy.11 It is therefore desirable that the introduction of public income protection only selectively displaces self-protective and self- insurance mechanisms--that is, that it displaces only expensive and counter- productive mechanisms. That means that underemployment also has to be taken into account. Indeed, the unemployed may not even be the most underprivileged group in the labor market. Summary A multitude of mechanisms are available to individuals, families, and com- munities for dealing with the risk of unemployment. When introducing or reforming public income support programs, policymakers should be con- cerned with the interactions of these programs with other mechanisms and institutions. Theoretical models offer increasingly refined and in-depth insights into the working of income support programs for the unemployed. But many aspects of these programs do not lend themselves to formal mod- eling, because the underlying theoretical models cannot be solved analyti- cally. To test the generality of the theoretical models, as well as to determine the effects that are theoretically ambiguous, empirical evidence is therefore called for. The approach adopted here for evaluating alternative income support programs for the unemployed uses two sets of criteria. One set evaluates the performance of these programs, stressing income protection and efficiency effects. The other set--design and implementation criteria-- recognizes wide differences across countries and assesses the specific features countries possess. This approach raises an important question: what can be learned from the experience of industrial countries that is of value for developing and transition countries, given the differences in institutional context and ability 22 Income Support for the Unemployed: Issues and Options to govern? Industrial countries' experience with income support programs can offer valuable lessons for developing and transition countries--provided that country-specific circumstances are appropriately accounted for. For example, severance pay was found to reduce labor market flows and employ- ment rates and to limit participation of marginal groups in OECD and Latin American countries (Heckman and Pages 2000). Some other findings from OECD countries, such as the "automatic stabilization" effects of unemploy- ment insurance and the effects of monitoring on the duration of the receipt of unemployment benefits, also offer important lessons for developing and tran- sition countries, once the scale of the program has been properly accounted for. The transferability of general equilibrium results--such as interactions between unemployment insurance systems and labor market institutions and policies, including collective bargaining--is more questionable. Industrial countries also offer valuable guidance on how to adapt income support systems when introducing them in developing countries. For exam- ple, monitoring of continuing eligibility for unemployment insurance-- which relies on a subjective assessment of recipients' behavior and their labor market status--has proved challenging even for OECD countries, prompting them to adopt various approaches and measures to improve it (see chapter 6). Given the weaker administrative capacity, larger informal sector, and different cultural norms in developing countries, programs there might do well to adapt the standard conditions of continuing eligibility (together with some other parameters of the program) and to drop the sub- jective assessment of recipients' behavior and status (perhaps as a tempo- rary measure). Unless such changes are made, monitoring is likely to be ineffective, costly, and prone to corruption (for the importance of adapting the lessons learned in OECD countries to the circumstances facing develop- ing and transition countries, see chapter 5). Notes 1. For a comprehensive framework of social protection based on social risk man- agement, see World Bank (2001a). 2. Besley and Coate (1995) provide another comprehensive approach, offering insights into the design and desirability of selected income maintenance programs, including negative income tax, wage subsidies, and workfare. 3. Davis and Haltiwanger (1999), Hopenhayn and Rogerson (1993), and Mortensen (1994) provide calibrations and simulations of the distortions in the mag- nitude of reallocation that can occur from various labor market interventions. 4. The net price for the insurance must be less than the difference between the expected income from uncertain outcomes and the certain income obtained by taking insurance (Barr 1990). 5. These are special circumstances that apply to insurance situations. In addi- tion, for market forces to produce efficient outcomes, there must be perfect informa- tion, perfect competition, and no market failures. Perfect information requires knowledge of prices, quality, and the future. Perfect competition in product and factor markets calls for economic agents to be price-takers and have equal power. Conceptual Issues 23 The main market failures are represented by the presence of public goods, external effects, and increasing returns to scale (Barr 1990). 6. According to Barr (2001), adverse selection arises when the insurer is imper- fectly informed about the individual's risk status (that is, when a poor risk can con- ceal that fact from the insurance company). Moral hazard arises when the insurer cannot monitor the insured's behavior. The insured can therefore influence the insur- ance company's loss without the knowledge of the insurance company, by affecting either the probability of the insured outcome or the size of the loss. 7. Unemployment is a relatively new phenomenon, emerging for the first time around 1895, in Great Britain, according to Atkinson (1995a). 8. Atkinson (1995a, p. 214) summarizes the origin of social insurance as follows: "The move from the traditional agricultural sector to modern industry may be char- acterized as that from a lower level of income to one where the wage was higher but subject to the `catastrophic' hazard of total unemployment. Unemployment emerges as part of the `modernization' of the economy." 9. According to the ILO, "underemployment exists when a person's employ- ment is inadequate in relation to specified norms of alternative employment account being taken of his or her occupational skill" (Resolution I of the 13th International Conference of Labour Statisticians, Geneva, October 1982). 10. For a discussion of the conceptual issues in defining discouragement and a review of its incidence in industrial countries, see OECD (1995b). 11. As Cox Edwards and Manning (2001, p. 346) note, "The transition from wide- spread underemployment to open unemployment is in part an income effect. As countries grow and household incomes rise, jobless workers are able to endure peri- ods without work while waiting for a job to open." 3 Review of Income Support Programs for the Unemployed Countries differ widely in the ways in which they provide income support for the unemployed. The social insurance program for the unemployed with the richest tradition--unemployment insurance--exists predominately in industrial countries. In most transition countries, unemployment insurance was introduced in the early 1990s. In developing countries, unemployment insurance is uncommon, and the unemployed often struggle to make ends meet. Faced with job loss, formal sector workers with permanent contracts may rely on severance pay. But these workers are often in the minority. Little protection is available to workers in nonregular employment and to those working in the informal sector; these workers must often rely on their own means to deal with unemployment. Among the few public programs they can sometimes turn to are training and, above all, public works. Govern- ments often provide public works programs in response to shocks that indi- viduals, households, and communities are ill-equipped to deal with, such as economic recessions and natural calamities. This chapter reviews how different countries provide income support for the unemployed. It begins with the typology of such programs, distinguish- ing between income maintenance and active programs. It then describes the features of the most important income support programs--participation rules, eligibility conditions, benefit levels, and sources of financing. It con- cludes by examining the incidence of these programs around the world. It examines the prevalence of various programs in individual countries, by themselves and in conjunction with other programs, and shows how their prevalence is related to the level of development, geographic position, and unemployment rate. Typology and Description of Main Income Support Programs Two main types of public income support programs assist the unemployed: income maintenance programs and active programs (table 3.1). Income maintenance programs determine benefit eligibility based on program 25 fing human the objective of ectly and overstaf dir ce income Main easing Smooth consumption Link support esourr management objectives employer Reduce without incr unemployment or book by both; social state be or omfr employer ough insurance to partly the pension and if tax the by unfunded thr or other contributors; have Financing additional by by chased Contributions employers, employees, often financing general evenuesr Paid (either funded eservesr contracts) Paid security sometimes financed employer evenues,r edits cr pur past no- work, on and to ch sear formalf sector Eligibility workers dismissal, job Conditional contributions, fault availability willingness and Laid-of sector Formal workers Unemployed payment the Duration for Limited One-time n.a. e Programs that past ar the time of of years rate. level usually than e over generally ement fair ar years ogram Support sometimes pr on tirer educed,r several Pensions. less Benefit centage at per Benefits a wages, declining Lump-sum payment, based service Special grants rights earlier typically often actuarially Income programs of ograms ypesT pr pay ement maintenance 3.1. benefit tirer ableT ogram Pr Income Defined Unemployment insurance Severance Early 26 ) ingf of , ce, ong meet (str ogram (continued overstaf to esourr and, objectives insurance distorting and poverty means- pr income der between esortr or ficiency ovide ovide Reduce in human ef political Pr without incentives link benefits contributions) Reduce smooth consumption Pr tested support last in of if workers, accounts account ogram) has deposited pr (or evenuesr subsidy general sometimes ed) both om Employer/govern- ment Contributions employers, or individual and solidarity (funded Fr evenuesr contributions unemployment insurance expir General the funds public on of or sector (with limited individual's owing the Formal workers Conditional availability in account optional borr funding) Means-tested Means-tested stand- (if limited has or ed) Limited Limited Unlimited alone) (after unemployment insurance expir Unlimited that mass a flat a rate to to public ogram eshold family pays eshold family flat pr edundantr the ough under thr or thr or in thr fs as income income average average Special sheds labor sector layof Replacement same unemployment insurance opsT specific for income amount opsT specific for income amount ograms of pr (may type ograms pr savings some sector involvement contribution assistance enchment workers) Public etrr include active of Defined Unemployment insurance accounts Means-tested Unemployment assistance Social 27 and ces job in objective income goods, empower with esourr with income income Main ovide eation Pr support, services; communities Link support investment human Link support cr evenuesr evenuesr evenuesr Financing General General General on by ficials (self- based of categories Eligibility opriate deemed ogram unemployed Anyone selection wage) If appr pr Selected of limited limited Duration Usually Limited Usually flat level training or below- wage and elated Benefit age-r ) Usually market Stipend W (continued applicable. 3.1. programs works subsidies Not ableT ogram Pr Active Public rainingT age n.a. W 28 Review of Income Support Programs for the Unemployed 29 participation rules, including the payment of a premium, under unemploy- ment insurance. No offsetting services are performed in exchange for these transfers, although certain actions on the part of recipients are usually required, such as job seeking. Income maintenance programs can be divided into three subgroups: defined benefit programs, defined contribution pro- grams, and means-tested programs. Active programs require that beneficia- ries perform certain services or activities (such as participating in public works or training) in exchange for income support or subsidy. These pro- grams are included as income support because they provide income to par- ticipants; sometimes this goal is quite explicit in the program design. This chapter describes the features of the five programs that are the focus of this book: unemployment insurance, unemployment assistance, unem- ployment insurance savings accounts, severance pay, and public works. For the sake of completeness, it also describes some other programs that provide income support to unemployed workers. These programs are not mutually exclusive. In many countries unem- ployment insurance coexists with unemployment assistance, severance pay, and public works. Under particular circumstances, the differences between some programs also become blurred. For example, if unemployment insur- ance benefits are paid at a flat rate rather than related to earnings, particu- larly if they are set at a relatively low level, they do not differ much from those provided by unemployment assistance (although the beneficiaries may differ). The level of benefits under unemployment and social assistance may also be similar, but the initial and continuing eligibility criteria are dif- ferent. Moreover, some programs, particularly recently introduced programs, blend various traditional programs. The new Chilean unemployment insur- ance program, for example, combines individual savings accounts with social insurance and partly replaces the severance pay program. Unemployment Insurance Unemployment insurance is provided as part of social insurance, together with pension and health insurance. The program requires that workers, their employers, or both make contributions, which are used to finance unem- ployment benefits. While it mimics market insurance, unemployment insur- ance deviates from actuarial principles by charging premiums that do not reflect individual risks. In industrial countries, unemployment insurance is the most widely used income protection program for the unemployed (figure 3.1). It is typi- cally mandatory, with the few voluntary programs (in Denmark, Finland, and Sweden, for example) subsidized by the state and resembling the compulsory programs of industrial countries in both function and form (Holmlund 1998).1 Most mandatory programs cover the majority of employed people, irrespec- tive of occupation or industry. People who are not insured, such as recent uni- versity graduates, first-time job seekers, and the self-employed, are sometimes eligible for benefits, while casual workers and domestics are most often not 30 Income Support for the Unemployed: Issues and Options Figure 3.1. Incidence of Unemployment Support Programs, by GDP per Capita Groupings (Late 1990s) 100 90 80 70 60 50 Percent 40 30 20 10 0 Low Lower-middle Upper-middle High Unemployment Unemployment Unemployment insurance insurance assistance savings accounts Severance pay Public works Source: Table 3.6. (see table 3.2 for stylized features of unemployment insurance by groups of countries). A few programs, particularly those in developing countries, cover only workers in industry and commerce. In order to qualify for unemployment insurance, an individual must satisfy the minimum cov- ered employment or contribution requirement, the most common length being six months in the past year. The cause of dismissal may affect if and when the individual is entitled to benefits. A usual condition for maintain- ing entitlement to unemployment benefits is that applicants be capable of, available for, and looking for work. Noncompliance with other labor market requirements can also result in the permanent or temporary sus- pension of benefits. Benefits are usually a proportion of average earnings over some stipu- lated period of the recipient's most recent employment spell. Generally, the initial replacement rate is 40­75 percent of average earnings. In some transi- tion countries, benefits are paid as a flat rate (that is, all recipients receive the same amount). Wage or benefit ceilings are used to limit the range of the benefits; benefit floors, typically set at the minimum wage, are also some- times set. In addition to the basic benefit, some programs provide supple- ments for dependents. Benefits commonly decline over time and are limited in duration. However, extensions are sometimes given to people with long, ) by on g). , only only as In Italy States, few States; less of levied a e employers made covers state In vary some (continued e Iceland, made ar financing financed of omfr workers, contribute e arise. the only ar in Luxembour costs. oss higher (but rates only United ar United state rates acr most, rates cent). ce and or In in that the do the ograms the, Spain, subsidies. In per covers cent 8 workers contributions Sour pr and per equal and, deficits and as 3 e only employers workers state Most contributions employers with contribution employers. countries or (contributions by Italy contributions by ypicallyT ovides any Italy pr Japan, the administrative Contribution significantly countries. contribution than equirr much a to if og of or, fice ed. of work, ching eligibility to efuser fer with fense months). typically sear under of of period Recipients eportingr equirr ecipientr available, to as job ch of 1­4 conditions e comply willing fail ements. ograms able, well work. disqualified to sear employment typically Continuing Pr equirr e be and as for ar they training, unjustifiably suitable fail job equirr Severity determines disqualification (usually Regularly to is . last to All of Countries 6 in . year 24 e of of fice. fer person be least Iceland of Almost last at deny Iceland work efusalr of eligibility a, at past weeks in Portugal. cases quitting, in in equirr or job 10 in conditions and in usually esidencyr ements. Groups Initial qualify weeks days countries by oT must employed months Range: 52 540 months countries egistrationr employment France have equirr suitable all benefits voluntary misconduct, stoppage, a it 36 of 48 , 3­12 is such a limit some (60 in Belgium social years Programs, in age. duration entitlement. entitlement sometimes of usually countries France, Denmark, Norway in Benefit but longer in in in is factors within contribution of is months duration to period, and countries Benefit opean 32 much duration Most length unemployment insurance Maximum period months, Eur is months months months and Germany); benefit unlimited. duration elatedr as security payments certain service, Insurance for en. usually is flat- e ceilings, aiting is United the ar benefit W eland, to Belgium, benefits. ecipientsr average Ir Kingdom months. the childr Kingdom, cent some earnings or tax voluntary due period benefit rates In France, long-term of ecentr exception per However of and, additional United 90 range. United benefits. including Unemployment States, insurance of maximum additional rate. days. weeks­6 average spouses the of the cases waiting 6 ovide or of Level cent Notable with dismissal unemployment and limit 3­7 in by pr eplacementr and flat-rate or Germany with Denmark, countries, per is United into countries, benefits wages ovisions ovide centages the some Features Initial 40­75 earnings. Denmark, eplacementr on pr France, pr period countries, quitting misconduct, extended Belgium, Kingdom rate per workers Most Canada, Netherlands, and In unemployment transit assistance. of In. and, either all on Austria Stylized a sector entices exclude coverage Portugal, United domestic ovisions most of graduates. occupation exclude employees employees). Majority cover individuals g, appr 3.2. by Germany to ograms by, based pr ovincial the or pr workers. Japan, countries pr ed conditions. sector and exclude training self-employed, Canada for ableT fer ograms espective oups, eland, casual Coverage OECD Of countries. pr employed irr Austria, Luxembour extends and Many the generally gr other and public (voluntary exist government Ir Spain, States or 31 e 6 or to equirr omfr the state worker range (Slovenia) State common. finances In ograms. e orker Slovak needed) is financing insurance the pr W of rates cent range (Slovenia) state oups. ce countries contributions; equirr (the per Employer (when gr the all cent cent (Albania). special Republic Sour also financing 0.06 per per cent special Almost employer some contributions. contribution omfr 1 Latvia to Republic). contributions 0.06 per subsidies deficit In unemployment for Slovak finances if a or or work, ching eligibility ecipientr to to educed,r the fer (job of files typically available, or sear Recipients e if market training. the as ar efusesr job comply works or participation conditions e able, willing efuser go not labor ements ch, ograms be well work. disqualified public Continuing Pr equirr e to and as for ar they under Benefits postponed, terminated ecipientr suitable does with equirr fraudulent sear in training) a claim. be wage. fices Ukraine ranges months. at of all Latvia, must ineligible eligibility past 12 In by and e dismissed misconduct. ar e of conditions ement to level countries 3 ed minimum ar Initial Minimum employment equirr omfr most Registration employment equirr they countries. Romania, income below In workers if because for on Slovak ement graduates duration entitlement 6­24 some oatia, Russian the duration period, University periods. tirer is In (Azerbaijan, Cr Slovenia), employment, the extensions depending of age. near training shorter countries Benefit ovide Maximum duration months. countries Bulgaria, Poland, Federation, Republic, entitlement varies length contribution and/or and have entitlement Some pr people age. of cent Poland) over job to per floors wage) level cause to than Republic, the wage) for by on flat-rate benefit gyz 50­75 wages. national of benefits benefits. or first-time lower Albania, countries) eplacementr or Benefit (Albania, addition graduated flat-rate e Kyr Uzbekistan limited minimum (usually oportion average in be ar the Federation, Level e depends pr or or elated wage. usually average elated ar the in and income e wage). flat-rate of can eceiver that supplements ar egional,r countries Sometimes (transition ) ecentr ceilings loss. Russian ovide ovide Initial rates of Benefits (usually and local, average sometimes job Some pr (usually minimum instead earnings-r Earnings-r benefits time. seekers benefits minimum Azerbaijan, the Ukraine, pr dependants. Asia by Central for and workers e (continued and ograms esidencyr for ar oatia training also ope pr or 16­59 e Coverage Cr or ar of 16­54 3.2. In Eur Domestic employed ed). workers excluded. usually and ableT Coverage Eastern Majority cover (citizenship equirr age: men women. university graduates eligible. casual usually 32 ) 2 In range is cent. to business (continued employees a per rates cent both ogram employers, per and pr ough 0.65 with by contributing. of thr 0.75 the tax cent, Contribution omfr per employers usually Brazil financed mainly sales a not for willing and ecipientsr and does able, Continuing in looking gentina be e Ar gentina In enezuela,V work. must available, to eligibility Ar equirr job. In e of ar to 6­12 must other benefits. ecipientsr, or been for some dismissal period In if means applicants in self have employment. applicants misconduct, eceiver security to claimants other gentina Generally must employed months stipulated ecentr Brazil ineligible due and lack support household. Ar cannot social of Brazil months. and of In under entitlement depends durations length 3­12 cumstances. is the length eased cir gentina Ar the incr e Maximum period In Brazil, entitlement on employment. entitlement ar special In 30 50­60 e e ar wage. and ar wages. in minimum limits rates Brazil, minimum days average 3 in Brazil, benefit the is to days ecentr and 60 eplacementr of period enezuela.V cent maximum in gentina oportional aiting Income per Ar and pr W Barbados, days Caribbean was wage and the in gentina all and ovide enezuelaV ogram pr Ar employees Barbados, pr in legislation the be cover enezuelaV domestics excludes to but workers. sector America Brazil amendments gentina, oduced yet Brazil the Latin Ar and unemployment insurance. Unemployment insurance intr (last 1991), has implemented. and workers. excludes casual Barbados public and self-employed. 33 the In of of and cent 0.5 of social per finances 0.2 cost if local contribute (China), cent, cent financing on ovisions), contribute the by subsidies. other state of pr Republic per per and, workers ce 0.6­1.0 employers the the aiwanT 0.7 0.1 omfr funds. ovides In In contribute employers covers employers cent, wages transferring Sour depends pr ces employers cent, China ea, per cent. cent, state per oviding In contribute (rate government state Iran 3 deficit. Kor 0.5 per workers per contribute the administration, pr worker needed, esourr insurance ch be In and and not ced. eligibility job. is must a work job-sear enfor to for the conditions available, ement Continuing Recipients able, willing looking urkeyT equirr fectively ef the 2 and Iran in months ea, ea must 6 in In. cannot fice Kor Kor accept . at of eligibility the year to fer employment 1 months of aiwanT misconduct In of (China), conditions to of ed ement: 6 in ed. Initial Iran, involuntary due efusalr Insur equirr in China, Republic years (China). Republic aiwanT unemployment be unemployment be or suitable Registration employment equirr on ea: and (China): 6­50 on based (benefits length duration years, Kor evious special status. of pr in Iran: 1­2 days, claimant of aiwanT based Benefit of marital months; China: Republic 90­240 age length employment extended cases); 6 months, employment and of in up 50­60 bonus the e benefits Iran for in ar except Republic the rates wage. benefit of aiwanT period benefit flat-rate The wages, half and of cent unemployment dependent pays per eemploymentr aiting ea per a W days. Level minimum 10 leaves Kor eplacementr average a fers of 14 which the dependents. is of exhausting e ) ecentr ea ovides four claimant Income of China, below pr supplement to Kor if befor entitlement. Republic (China) the Iran. and other in (continued ea, and and by ea five firms Kuwait, Kor self- ed enterprises self- Coverage voluntarily China The Kor firms. than of ovide the permanent collective and of the 3.2. Iran, (China), all pr people, workers significantly cover sector (China) ed some fewer ableT fers Coverage Asia China, Republic aiwanT urkeyT ovisions. unemployment insurance. dif excludes employed, insur people pr excludes contract public and enterprises. Republic includes aiwanT excludes employed with workers. 34 In cent, cent. contribute South and state In contribute per per employees 1.5 2.5 the each deficit. employees cent, employees cent. Algeria per per In contribute employers Egypt 2.0 finances Africa employers 1 willing and able, be work. Must available, to ch ee of sear South in be cannot strike. Thr e benefits. in Africa, years must active misconduct, 3 employment; for weeks voluntary with of befor months. 13 weeks. to (2000a). ed ed 6 ent 52 South due Algeria: cover employer curr contributions. months equirr in, applying Egypt: Africa: past Unemployment be separation, or participation OECD is (1998); Egypt for based In on length. Roddis duration of entitlement 16­28 paid countries. is e Africa based ar and months, Algeria length South weeks. transition In 12­36 on employment. maximum duration weeks, contribution In benefits 26 Tzannatos of of under of and 7 cent In mean floor is 2002b); per Egypt is a minimum spousal Africa. included 60 e e in rate national 2002a, ar ar Africa. with ovides period and and national pr South (1999, wages rates South the aiting and OECD W of in earnings, of benefits eplacementr Algeria earnings Egypt average cent cent in per per members Replacement ecentr Administration e 45 Algeria average minimum 75 earnings. graduated allowances. days ar that Security and ovide only pr formal Algeria excludes highly Social Coverage casual workers. excludes countries oss In Egypt, covers om fr and and U.S. Africa acr salariedf Egypt. sector Africa ces: employees. fers ogram Sour ransitionT a Africa Algeria, South unemployment insurance. dif countries. pr laid-of workers sector public employees, workers, agricultural South domestics paid 35 36 Income Support for the Unemployed: Issues and Options continuous employment records or to those near early or regular retirement age. In many countries, it is also possible to move into means-tested unem- ployment assistance after exhausting unemployment insurance benefits. Unemployment insurance is typically financed through regular contri- butions based on wages by employers, workers, or both. The rate paid by employers is usually the same as or higher than that paid by workers. Some- times, however, workers are not required to make contributions. In the United States, employer contributions depend on the employer's layoff experience--employers who lay off workers more frequently and thus impose heavier financial burden on the program are assigned a higher rate (the practice is known as the "experience rating"). In some countries, the state subsidizes or finances program deficits that arise. Unemployment Assistance Unemployment assistance is a means-tested minimum income granted to working-age individuals who are unemployed and do not have the neces- sary financial resources to maintain a minimum standard of living for them- selves and their families. Australia provides means-tested unemployment benefits through two back-to-back programs. The Job Search Allowance is offered for a maximum duration of 12 months. If necessary, it is followed by the NewStart Allowance, which is offered indefinitely. Self-standing unemployment assistance programs currently exist in few countries, notably Australia, Hong Kong (China), and New Zealand among high-income countries and Serbia and Montenegro and Tunisia among middle-income countries. In all other countries that offer unemployment assistance, the program functions in tandem with unemployment insurance and is available upon expiration of eligibility for unemployment insurance to unemployed people who pass the means test (Vroman 2002). Like unemployment insurance, unemployment assistance programs require that applicants be capable of, available for, and looking for work (see table 3.3 for stylized features of unemployment assistance by groups of countries). Claims are reviewed at regular intervals to assess job-seeking intensity and to determine changes in household circumstances that may require a change in the benefit level. In some countries, eligibility for unem- ployment assistance is not conditional on previous employment or contri- bution history. However, in most countries, unemployment assistance is an extension of unemployment insurance offered to the long-term unemployed who have satisfied some minimum length of employment and do not have the economic means to support their households. The two primary groups that receive unemployment assistance are people who have exhausted their unemployment insurance entitlement and people who are ineligible for unemployment insurance because of insufficient employment records. Some countries penalize applicants whose unemployment was voluntary by limiting the length of unemployment assistance entitlement or extending the waiting period. ) tax ograms: financing pr financed tandem (continued in same of as general ce unemployment ough Sour Stand-alone government thr evenues.r Unemployment assistance with insurance: financing unemployment insurance. to be eceiptr eligibility usually must oughout available and of claimant seeking thr checks conditions e Eligibility and ograms unemployed, work, period Continuing Pr equirr be able for actively work. conditions satisfied the (periodic conducted). as of test, or job means such include year France such to if 8 Countries Generally years some limited to assets months the insurance 6-month have of and In a assistance due, satisfy suitable of eceding (6 in 10 contribution applicants and and employment pr in who of or has to countries, must assistance, benefits). of eligibility Exceptions. Australia years In extended is Groups (4 years Portugal only some people voluntary efusalr educedr conditions income state 5 assistance unemployment is e example, In to or by history in employment unemployment ar Initial ecipientsr, housing espective unemployment), employment unemployment). apply employment for for, unemployment their only benefits. period excludes and irr of ement). (household ed dispute, Netherlands benefits, years Austria, waiting ypicallyT fer eceding eceding test which family of contribution the employment unemployment), contributory pr (5 pr countries conditions unemployment ineligible (Germany equirr fer Programs, as available exhausted insurance unemployment labor of or weeks. as the not on for or long fulfilled. years), e (2 duration as ar include months), (5 Portugal exhausted longer have). Assistance In depends whether has benefits is (6­18 who Benefit Sweden and Unlimited, conditions Exceptions Netherlands Spain and months). duration age claimant unemployment insurance (duration those or of to level , in rate. granted in Unemployment benefit waiting flat marital Some which educedr Special older days of of e special people. applied ar benefit, number have e usually also not on income ar the no, omfr (3 days However period 5 and above Level for eshold younger Benefits ypicallyT ovisions ovisions eland, Features depends status dependents. countries thr levels, benefits eliminated. pr exist unemployed. Sometimes pr to Generally periods. waiting sometimes applicants transiting unemployment insurance Ir Sweden). of or (to half Stylized insurance New self- sector workers, about exhausted Only ograms. all of pr in countries and to 3.3. under have unemployment countries who ableT esent ograms). Coverage OECD Pr member those benefits unemployment pr Australia Zealand standing assistance vailableA espective unemployed irr occupation. 37 tax as ograms: ce financing pr financed tandem in same of sour general ce unemployment ough Sour Stand-alone government thr evenues.r Unemployment assistance with insurance: financing unemployment insurance. the be to fices be and and of job eligibility satisfy test. income job. in of oughout