Work and the Social Safety Net: Labor Activation in Europe and the United States 0190241594, 9780190241599

Citation preview

Advance Praise for Work and the Social Safety Net “Activation policies have spread over all the advanced capitalist democracies during the

last four decades as the main way to bring people back to the labour market. It was time to assess their performances and impacts. This book is the one to be read to know everything you wanted about activation, with the right country sample and the best possible authors.” —Bruno Palier, French National Centre for Scientific Research/CNRS Research Director at Sciences Po, Centre for European Studies and Comparative Politics

“Most people assume that European social welfare policies are more generous than those in the U.S. What hasn’t been recognized is that this generosity has been accompanied by increasing emphasis on connecting people to jobs. Meanwhile, progressives in the U.S. have recently been moving in the opposite direction—arguing for more unconditional assistance. This volume provides a nuanced and empirically grounded treatment of this debate. With chapters authored by some of the leading experts in the field, I recommend it to anyone interested in a comparative perspective on one of our deepest and most important contemporary policy debates.” —Isabel Sawhill, Senior fellow in Economic Studies, The Brookings Institution “Labor market policy and anti-poverty policy are often at loggerheads in most rich nations. Programs like the child allowance in rich countries are not tied to work, except for pending policy proposals for the refundable Child Tax Credit in in the USA. But as this timely volume discusses, there appears to be room to promote work and to provide income stability for all children. Readers will be amazed at the wide range of options for promoting stable work arrangements while also providing income security that are presented in this volume.” —Timothy Smeeding, Lee Rainwater Distinguished Professor of Public Affairs and Economics, University of Wisconsin–Madison “This book examines the functioning of welfare systems and active labour market policies and their interaction during economic downturns. While I do not share all its underlying assumptions, important policy lessons are drawn: for example, extending labour market service support into early periods of re-employment is a well-founded and highly promising proposal. This requires, however, resources that only few countries are ready to devote to such policies. An interesting proposal in this respect is to mandate increasing resources for active policies after recessions similar to improving generosity of benefit schemes during them. A differentiation that could have been more clearly made is between major child and family allowances which serve the objective of child well-being versus UI and on certain social assistance programs. Nevertheless, I learned a great deal about the programs reviewed and it gave me an opportunity to sharpen my own views.” —Georg Fischer, Vienna Institute for International Economic Studies; former Director for Social Affairs, European Commission “Since the 1990s, a combination of work requirements and generous social and economic supports has helped millions of Americans rise out of poverty and welfare dependency. By revealing how some European governments have embraced that approach, this volume may inform and inspire the ongoing efforts to promote work, responsibility, and independence among struggling Americans.” —Robert Doar, President of the American Enterprise Institute, former Commissioner of the New York City Human Resources Administration/Department of Social Services

International Policy Exchange Series Published in collaboration with the Center for International Policy Exchanges University of Maryland

Series Editors Douglas J. Besharov Neil Gilbert

Adjusting to a World in Motion: Trends in Global Migration and Migration Policy Edited by Douglas J. Besharov and Mark H. Lopez Caring for a Living: Migrant Women, Aging Citizens, and Italian Families Francesca Degiuli Child Welfare Removals by the State: A Cross-​Country Analysis of Decision-​Making Systems Edited by Kenneth Burns, Tarja Pösö, and Marit Skivenes Improving Public Services: International Experiences in Using Evaluation Tools to Measure Program Performance Edited by Douglas J. Besharov, Karen J. Baehler, and Jacob Alex Klerman Welfare, Work, and Poverty: Social Assistance in China Qin Gao Youth Labor in Transition: Inequalities, Mobility, and Policies in Europe Edited by Jacqueline O’Reilly, Janine Leschke, Renate Ortlieb, Martin Seeleib-​Kaiser, and Paola Villa Decent Incomes for All: Improving Policies in Europe Edited by Bea Cantillon, Tim Goedemé, and John Hills Social Exclusion in Cross National Perspective: Actors, Actions, and Impacts from Above and Below Edited by Robert J. Chaskin, Bong Joo Lee, and Surinder Jaswal The “Population Problem” in Pacific Asia Stuart Gietel-​Basten United States Income, Wealth, Consumption and Inequality Edited by Diana Furchtgott-​Roth Europe’s Income, Wealth, Consumption and Inequality Edited by Georg Fischer and Robert Strauss The World Politics of Social Investment: Volume 1 Edited by Julian Garritzman, Bruno Palier, and Silja Hausermann The World Politics of Social Investment: Volume 2 Edited by Julan Garritzman, Bruno Palier, and Silja Hausermann Work and the Social Safety Net: Labor Activation in Europe and the United States Edited by Douglas J. Besharov and Douglas M. Call

WORK AND THE SOCIAL SAFETY NET Labor Activation in Europe and the United States Edited by DOUGLAS J. BESHAROV AND DOUGLAS M. CALL

Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and certain other countries. Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America. © Oxford University Press 2023 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by license, or under terms agreed with the appropriate reproduction rights organization. Inquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above. You must not circulate this work in any other form and you must impose this same condition on any acquirer. CIP data is on file at the Library of Congress ISBN 978–​0–​19–​024159–​9 DOI: 10.1093/​oso/​9780190241599.001.0001 9 8 7 6 5 4 3 2 1 Printed by Sheridan Books, Inc., United States of America

CON TEN TS

List of Illustrations About the Authors Contributors

vii ix xiii

1 Introduction: European and US Experiences with Labor Activation Douglas J. Besharov and Douglas M. Call 2 Activation and Employment Support Policies for Stronger and Fairer Labor Markets Stefano Scarpetta 3 Early Activation in European Union Unemployment Insurance Programs Márton Csillag and Anna Adamecz-​Völgyi 4 Unemployment Insurance After the Great Recession Jacob Alex Klerman

v

1

25

38

61

vi

vi Contents 5 Activation in Eight European Social Assistance Programs Ivar Lødemel and Amílcar Moreira

91

6 Less Activation in US Social Assistance Programs? Matt Weidinger

122

7 Five Decades of Disability Benefit Policies in Five OECD Countries Duncan McVicar, Roger Wilkins, and Nicolas R. Ziebarth

150

8 Lessons for US Disability Policy from Other OECD Countries Richard V. Burkhauser and Mary C. Daly

183

9 Activation in Public Employment Services in Europe J. Timo Weishaupt, Henning Jørgensen, and Alexander Nunn

204

10 Workforce Development Services in the United States Carolyn J. Heinrich

242

11 Whither Activation Policies? Reflections for the Future John P. Martin

283

12 Lessons for Labor Policy in the Aftermath of the Great Recession Edward Montgomery

306

Index

335

L IST OF ILLUS TR ATIONS

FIGURES 5.1 Importance of variables in determining clustering 6.1 AFDC/​TANF caseload, 1959–​2020 6.2 Labor force participation rates for selected groups of working-​age women, 1990–​2018 6.3 Mean wage and salary income for working-​age mothers, 1990–​2018 6.4 Child poverty rates using two measures of poverty, 1990–​2020 6.5 National average TANF work participation rate for all families, FY2002–​FY2020 6.6 TANF work participation rate, FY 2002–​FY2016 7.1 Disability recipiency rates across countries 7.2 Disability recipiency rates by country 7.3 Development of new private disability insurance policies in Germany 8.1 Sources of SSDI prevalence rate growth, 1980–​2011 8.2 United States SSDI recipients per 100, Ages 16–​64, 1970–​2018 8.3 Normalized growth in SSDI prevalence by age, 1980–​2011

vii

103 131 132 134 135 139 140 152 155 173 188 190 192

vi

viii  LIST OF ILLUSTRATIONS 8.4 8.5 10.1 12.1 12.2 12.3 12.4 12.5 12.6 12.7

Past SSA projections of SSDI prevalence rate, 1980–​2020 Disability receipiency rates across countries, 1970–​2018 Public spending on labor market programs, 2000–​2018, for OECD countries United States employment-​to-​population ratio United States labor underutilization rate United States long-​term unemployed Selected unemployment rates during Great Recession Food stamp recipients Recipients of other assistance programs Long-​term unemployment remains high

193 194 245 307 308 308 309 316 316 322

TABLES 3.1 The evidence on the effectiveness and cost efficiency of different early interventions 54 5.1 Cross-​tabulating welfare-​state and minimum income typologies100 5.2 Models of activation in Europe (two-​step cluster analysis) 102 5.A1 Description of minimum income schemes 114 7.1 Disability recipiency rates for five OECD countries 153 7.A1 Data description and sources 178 8.1 Disability recipiency rates for six OECD countries 195 8.A1 Summary of disability insurance data availability across countries 201 9.1 Work Programme payment structure (summary maximum year 1 payments) in 2014 220 10.1 Summary of findings from key studies on workforce development program/​active labor market policy effectiveness 250 12.1 Economic indicators in selected OECD countries 328

ABOU T THE A UTHO RS

Douglas J. Besharov is a professor at the University of Maryland’s School of Public Policy, where he teaches courses on poverty, welfare, children and families, policy analysis and logic models, program evaluation, and performance management. He is also a senior fellow at the Atlantic Council, where he leads a program on international policy exchanges. Between 1985 and 2009, he was also a resident scholar at the American Enterprise Institute for Public Policy Research in Washington, DC. Between 1975 and 1979, he was the first director of the US National Center on Child Abuse and Neglect. In 2008, he was President of the Association for Public Policy Analysis and Management (APPAM) and, subsequently, APPAM’s International Conference Coordinator. Among his eighteen books are Counting the Poor: New Thinking About European Poverty Measures and Lessons for the United States (Oxford University Press, 2012), with Ken Couch; and Poverty, Welfare, and Public Policy: Readings from Journal of Policy Analysis and Management (John Wiley and Sons, 2010), with Douglas M. Call. He has written more than 250 articles and has contributed to the Los Angeles Times, New York Times, Wall Street Journal, and Washington Post. He is the director of the University of Maryland’s Welfare Reform Academy and its Center for International Policy Exchanges. Together with Neil Gilbert of the University of California (Berkeley), Professor Besharov is co-​Editor-​in-​Chief of the Oxford University Press Library on International Social Policy. Douglas M. Call is the deputy director of UMD’s Program for International Policy Exchanges and its Welfare Reform Academy. He has coauthored (with Professor Besharov) articles for the Journal of Policy Analysis and Management, the Policy ix

x

x  About the Authors Studies Journal, and The Wilson Quarterly and a number of book chapters. With Professor Besharov, he was the co-​editor of Poverty, Welfare, and Public Policy, the third volume in the Journal of Policy Analysis and Management’s “Classics” series and the “Cohabitation” section of the Oxford International Handbook of Family Policy. Call is also a lecturer at the UMD School of Public Policy, teaching courses on program evaluation, poverty measurement and alleviation, and the capstone course for students in the social policy concentration. Anna Adamecz-​Völgyi has been a research fellow of the Institute of Economics CERS in Budapest and the UCL Social Research Institute in London. She earned her PhD in economics at the Central European University. Her research looks at labor market inequalities, the effects of educational interventions, and intergenerational educational mobility. She has published in the Scandinavian Journal of Public Health, the Journal of Population Economics, the Oxford Review of Education, and the Economics of Education Review. Richard V. Burkhauser is the Emeritus Sarah Gibson Blanding Professor of Public Policy in the Department of Policy Analysis and Management at Cornell University. Previously, Burkhauser held tenured professor positions in the Economic Department at Vanderbilt University and Syracuse University. Between September 2017 and May 2019, he was a member of President Trump’s Council of Economic Advisers. His professional career has focused on how public policies affect the employment and well-​being of vulnerable populations. Márton Csillag, an empirical labor economist, is a senior researcher at the Budapest Institute for Policy Analysis. His recent research focuses on the evaluation of labor market and social programs for the long-​term unemployed and on the incentive effects of sickness and unemployment insurance benefits. He also regularly works as a senior expert for the European Network of Public Employment Services and is a research fellow at the Institute of Economics, Center for Economic and Regional Studies. He earned his PhD in economics at Toulouse University (France) and worked at KU Leuven, Maastricht University, and University College London before joining the Budapest Institute. Mary C. Daly is the president and CEO of the Federal Reserve Bank of San Francisco. Daly began her career with the San Francisco Fed in 1996, as an economist specializing in labor market dynamics and economic inequality. She went on to become the Bank’s Executive Vice President and Director of Research. As a participant on the Federal Open Market Committee, she helps set American monetary policy that promotes a healthy and stable economy. Daly is committed to making the San Francisco Fed a more community-​engaged bank that is transparent and responsive to the people it serves. She has focused much of her professional career on crafting public policy that can improve the economic quality of life for all Americans. Carolyn J. Heinrich is the Patricia and Rodes Hart Professor of Public Policy, Education, Economics and Health Policy at Vanderbilt University. Her research

x  i

About the Authors  xi focuses on education, workforce development, health and social welfare policy, program evaluation, and public management. She works directly with federal, state, and local governments in her research to improve policy design and program effectiveness and also collaborates with nongovernmental organizations to improve the impacts of economic and social protection investments in middle-​income and developing countries. She received the David N. Kershaw Award for distinguished contributions to the field of public policy analysis and management in 2004 and was elected to the National Academy of Public Administration in 2011. Henning Jørgensen is a professor in political science and public administration at Aalborg University in Denmark and the former director of the research institutes CARMA (Aalborg) and ETUI (Brussels). His publications include labor market policy, policy formation and implementation, public sector reforms, and street-​level bureaucracy. Jacob Alex Klerman is a senior fellow with Abt Associates, where he leads large randomized trials of social programs. His recent evaluation efforts have concerned job search assistance (for welfare recipiences and unemployment insurance claimants), job training, and nutrition assistance. Ivar Lødemel is professor of comparative social policy and founding research director of the Research Group for Inclusive Social Welfare Policies (GiV) at Oslo Metropolitan University. His main current areas of research are cross-​Atlantic comparative studies of activation programs and global studies about the relationship between poverty and shame. John P. Martin is a chair of the Irish government’s Labour Market Advisory Council and a member of the National Statistics Board of Ireland. He worked for many years at the Organisation for Economic Co-​operation and Development (OECD) in Paris in both the Economics Department and the Employment, Labour and Social Affairs Directorate, where he was Director from 2000 to 2013. He was the founding editor of the OECD Employment Outlook and also edited the OECD Economic Outlook. He has published extensively in international trade and labor economics. Duncan McVicar is a professor of economics in the Queen’s Management School at Queen’s University Belfast. His research focuses on labor economics and the economics of social policy. Edward Montgomery is the president of Western Michigan University. Previously, he served as the executive director of the White House Council for Auto Communities and Workers under President Obama and as the Deputy Secretary of the US Department of Labor under President Clinton. A labor economist, he has held faculty appointments at Georgetown University, Carnegie Mellon University, and the University of Maryland is a research associate at the National Bureau of Economic Research. Amílcar Moreira holds a PhD in social and political sciences from the University of Bath and is currently a researcher at the Research Centre in Economic and

xi

xii  About the Authors Organizational Sociology (SOCIUS) and a guest assistant professor at the Lisbon School of Economics & Management, University of Lisbon. Between 2017 and 2021, he was also a member of the Board of the European Sociological Association—​ Research Network on Ageing. He has published on such issues as the demand for older workers, intergenerational conflicts over labor market regulation, austerity-​ led labor market reforms in the south of Europe, activation policies for minimum income recipients, and, more recently, on the social policy responses to COVID-​19. Alexander Nunn is a professor of global political economy and the head of the Centre for Social, Cultural and Legal Research at the University of Derby, in the United Kingdom. His research focuses on the production, reproduction, and governance of inequalities, including through labor market policy. Stefano Scarpetta is the director for employment, labor, and social affairs of the OECD since 2013. He leads the work of the Organisation in a number of broad areas including employment, labor, migration, health, adult skills, gender, and tackling inequalities. He represents the OECD in the G20, G7, and many other international fora on these topics. He has published extensively in academic journals and is the author of various books. Matt Weidinger is a senior fellow and Rowe Scholar in poverty studies at the American Enterprise Institute (AEI), where his work is focused on safety-​net policies including cash welfare. Before joining AEI, Mr. Weidinger served as the deputy staff director of the US House Committee on Ways and Means and as the long-​time staff director of its subcommittee with jurisdiction over TANF and other safety-​net programs. He was a primary staff author of the landmark 1996 welfare reform law. J. Timo Weishaupt is a professor of sociology with a focus on social policy at the University of Göttingen, Germany. He received his PhD from the University of Wisconsin, Madison, in 2008. Before joining the faculty in Göttingen in 2015, he held a position as junior (assistant) professor at the University of Mannheim, Germany. His research interests include labor market policy and governance, public employment services, EU social policy, industrial relations, and, most recently, homelessness and social exclusion. Roger Wilkins is the deputy director of the Melbourne Institute of Applied Economic & Social Research at the University of Melbourne. He is also the deputy director (research) of the HILDA Survey, Australia’s nationally representative longitudinal household study. His research interests include the nature, causes, and consequences of labor market outcomes; the distribution and dynamics of individuals’ economic well-​being; and the incidence and determinants of poverty, social exclusion, and welfare dependence. Nicolas R. Ziebarth is an associate professor in the Economics Department and the Jeb E. Brooks School of Public Policy at Cornell University; he is also an associate director of the Cornell Institute for Healthy Futures. His research focuses on the interaction of social insurance with labor markets and population health.

CON TRI BUTOR S

Anna Adamecz-​Völgyi

Research Associate Institute of Economics Centre for Economic and Regional Studies Budapest, Hungary

Douglas J. Besharov

Deputy Director Center for International Policy Exchanges, School of Public Policy University of Maryland College Park, MD, USA

Mary C. Daly

Carolyn J. Heinrich

Emeritus Sarah Gibson Blanding Professor of Policy Analysis Department of Policy Analysis and Management Cornell University Ithaca, NY, USA

Douglas M. Call

Senior Researcher Budapest Institute for Policy Analysis Budapest, Hungary

President and CEO Federal Reserve Bank of San Francisco San Francisco, CA, USA

Professor School of Public Policy University of Maryland College Park, MD, USA

Richard V. Burkhauser

Márton Csillag

Patricia and Rodes Hart Professor of Public Policy, Education and Economics and University Distinguished Professor Leadership, Policy and Organizations Vanderbilt University Nashville, TN, USA

Henning Jørgensen

Professor Department of Politics and Society Aalborg University Aalborg, Denmark

xiii

xvi

xiv CONTRIBUTORS Jacob Alex Klerman

Stefano Scarpetta

Ivar Lødemel

Matt Weidinger

Senior Fellow Department of Social and Economic Policy Abt Associates Rockville, MD, USA

Professor Oslo Metropolitan University Oslo, Norway

John P. Martin

Chair Irish Government's Labour Market Advisory Council Dublin, Ireland

Duncan McVicar

Professor of Economics Queen’s Management School Queen’s University Belfast Belfast, UK

Edward Montgomery

President Western Michigan University Kalamazoo, MI, USA

Amílcar Moreira

Guest Assistant Professor of Economic Sociology Lisbon School of Economics and Management University of Lisbon Lisbon, Portugal

Alexander Nunn

Professor of Global Political Economy and Director of the Social, Cultural and Legal Research Centre Department of Social Sciences University of Derby Derby, UK

Director Directorate for Employment, Labour and Social Affairs OECD Paris, France

Senior Fellow Department of Poverty Studies American Enterprise Institute Washington, DC, USA

J. Timo Weishaupt

Professor Institute of Sociology University of Göttingen Göttingen, Germany

Roger Wilkins

Professorial Fellow Melbourne Institute of Applied Economic and Social Research University of Melbourne Melbourne, Australia

Nicolas R. Ziebarth

Associate Professor Brooks School of Public Policy and Department of Economics Cornell University Ithaca, NY, USA

 1

1    INTRODUCTION EUROPEAN AND US EXPERIENCES WITH LABOR ACTIVATION Douglas J. Besharov and Douglas M. Call

F

or many decades, American liberals have pointed to Europe’s social welfare systems as a model for the United States. As Senator Bernie Sanders famously said: “I think we should look to countries like Denmark, like Sweden and Norway, and learn what they have accomplished for their working people” (Moody 2016). In early 2021, the Democrats had an opportunity to act on their wishes. In the midst of the COVID-​19 pandemic, they gained control of both houses of Congress as well as the presidency. With the country newly resensitized to poverty and racial and income inequality by the starkly disparate impact of the virus, and emboldened by the political opening it provided, Democrats embarked on an ambitious effort to address many long-​standing social and economic problems. A key element was a revamped and greatly expanded federal Child Tax Credit (CTC) that now gives between $3,000 and $3,600 per child (depending on the child’s age and family income) to the parents of about 96% of American children. Most important for our purposes, it removed the requirement that parents have earnings in order to receive the payments. Severing this connection to work essentially reversed a key tenet of American welfare policy. Sanders and other progressives, however, have missed the big changes in European social safety-​net programs that began in the 1990s. Undoubtedly, European countries have more generous social benefits than the United States, for sure, but, to preserve the viability of their programs in face of changing economies, demographics, and social attitudes, many have added programmatic Douglas J. Besharov and Douglas M. Call, Introduction In: Work and the Social Safety Net. Edited by: Douglas J. Besharov and Douglas M. Call, Oxford University Press. © Oxford University Press 2023. DOI: 10.1093/​oso/​9780190241599.003.0001

2

2  Work and the Social Safety Net elements designed to discourage long-​term recipiency and encourage work and other work-​related behaviors. The overarching concept is “labor activation,” and tracing US and European experiences is the theme of this volume. By describing how European countries seek to activate recipients of their safety-​net programs without compromising basic protections, we hope that this volume will help illuminate the US debate about the future of social assistance and the importance of work and work-​related activities. Separate sections on unemployment insurance, social assistance, disability programs, and workforce development services are designed to provide useful examples for reconciling the provision of financial aid while encouraging participation in the labor force. Each section has a chapter on European programs and their activation efforts, followed by one on US equivalents. The volume concludes with two chapters on the wider context of enhanced labor market polices to support and extend the effectiveness of activation efforts.

SEVERING THE CONNECTION OF CASH AID TO WORK IN THE UNITED STATES Permanently transforming the CTC into an unconditional cash grant would make it the largest US cash assistance program. Add the pre-​2021 credit’s underlying a ten-​year cost of very roughly $1.2 trillion, and the total cost of the CTC would be more than $2.6 trillion, of which about $1.2 trillion would be cash payments to families with no tax liabilities (Crandall-​Hollick, Carter, and Boyle 2021). That’s about the same as the ten-​year cost (in 2018) of all federal aid to education (K-​12 through college) (US Department of Education, National Center for Education Statistics 2020a, 2020b). Moreover, because the CTC sets no requirements for receiving payments, it effectively undoes the current welfare system’s job-​seeking or other work-​related requirements. (All dollar figures are in 2020 dollars, unless otherwise indicated.)

US Welfare Reform (1996) In 1996, under the leadership of President Bill Clinton and House Speaker Newt Gingrich, America’s welfare system was dramatically revamped. The AFDC program (enacted in 1935 as Aid to Dependent Children) and its entitlement to welfare benefits was replaced by the Temporary Assistance for Needy Families (TANF) program, which ended the entitlement to welfare benefits. In its stead, TANF imposes, through the states, a bevy of activities on recipients that are best characterized as “work-​related,” including job search requirements, physicals to determine ability to work, work-​prep, and sometimes job training, coupled with partial or full benefit sanctions for failure to comply with the requirements.

3

Experiences with Labor Activation  3 Parallel programs, primarily the Earned Income Tax Credit and child care, support parents who get jobs.1 Implementation of TANF’s provisions has always been quite uneven, as described by Matt Weidinger of the American Enterprise Institute (Chapter 6, this volume), but, together with a strong economy and work-​support programs, TANF is widely credited with having helped reduce welfare caseloads by about 78% while raising the labor force participation rate of never-​married mothers and lowering their poverty rate and the teen birth rate. Although millions of low-​skilled single mothers entered the labor force and became better off in the years after TANF’s passage, these are, of course, average effects. Advocates for the poor have never liked TANF, in part because it “forces” many mothers into “bad” jobs and prevents many needful families from getting the help they need. Some have surely been left behind. Hence, support for dropping the earnings requirement from the CTC is at least partly based on pent-​up hostility toward welfare reform and the perception that it left behind many single mothers with children without adequate funds to live a decent life. (Another important factor was the decades-​long drumbeat about rising income inequality, even though objective research is much more ambiguous; see Furchtgott-​Roth 2020.) Welfare reform, however, remains popular among voters. Hence, rather than attempting to repeal TANF, Democrats upped the CTC’s value and removed its work requires, effectively creating a substitute welfare program. In fact, the revamped CTC provides roughly the same amount of money as TANF in direct, monthly checks.2 As the Washington Post headlined, “Goodbye, Clinton Welfare Reform. Hello, Child Tax Credit” (Lane 2021).

Many Years of Political Conflict The CTC did not emerge out of thin air. Its initial version was passed nearly twenty-​five years ago, as part of a major reform of the tax system in the Taxpayer Relief Act of 1997. At the time, and for many years afterward, it had strong Republican support. Over the years, its coverage was repeatedly expanded and size of payments increased.

1. TANF’s other main element, the 5-​year time limit on benefit receipt, receives much less attention, probably because it seems so harsh and, in practice, is essentially unenforced as states are allowed to exempt 20% of their TANF caseload from the time limits and can also easily work around it (e.g., through separate state programs and child-​only cases so that families can continue receiving benefits beyond the putative time limit) (Falk 2017). 2. For example, a single mother with two children (one under age six and one age six or older) would receive $6,600 ($3,600 for the child under age six and $3,000 for the child age six or older) annually from the CTC. That would be more than 20% larger than the 2019 national average TANF payment for families of the same size, which was about $5,400 (and about 20% smaller that the average annual AFDC grant in 1966, which was about $8,300; see US Department of Health and Human Services 2014; US Department of Health and Human Services 2020.)

4

4  Work and the Social Safety Net The 2021 expansion raised the maximum CTC payment from $2,000 to $3,600 for each child under 6 and from $2,000 to $3,000 for each child ages 6 to 17. The full credit went to individuals with incomes as high as $75,000 and to married couples with incomes as high as $150,000, before being phased down to $2,000 for individuals with incomes up to $200,000 and married couples with incomes up to $400,000 (above which the credit declined on a sliding scale, phasing out completely at $240,000 and $440,000, respectively). There was no limit to the number of children who can be claimed. Even more striking than the increases in payment amounts is that they were made fully refundable; that is, payments are made even to those with zero earnings and, hence, no tax liability. (Parents with less than $2,500 in earnings previously received zero payments.) The changes to the CTC have been widely praised by the liberal commentariat as a new system of more generous, no-​strings income support for low-​income Americans. In fact, many believe that it portends a much more generous universal basic income. For the same reasons, most conservative pundits have been sharply critical. The CTC expansion was part of the $1.9 trillion American Rescue Plan (ARP), the Biden Administration’s first COVID-​19 legislation passed in March 2021. The ARP about doubled the size of the CTC (from about $118 billion to about $223 billion a year; Crandall-​Hollich et al. 2021). If projected over a ten-​year period, the credit’s additional cost is estimated to exceed $1.4 trillion, with some estimates as high $1.6 trillion (depending on population growth) (Crandall-​ Hollich et al. 2021). The CTC was expanded for only one year because including its real ten-​year cost in the March legislation would have ballooned its price tag by 75% (to $3.3 trillion), which was deemed a political nonstarter. Leading Democrats, though, always vowed to extend the expanded CTC beyond one year, and President Biden included a four-​year extension (again avoiding its full cost) in his original $3.5 trillion social infrastructure bill, originally the American Families Plan (Executive Office of the President 2021). When the bill was trimmed to a supposed $1.75 trillion or so, the CTC extension was again shaved to only one year so that it would fit within the bill’s reduced cost. (Other programs, such as child care subsidies and parental leave, were likewise given shortened lives.) This is a common budget gimmick in the United States. The idea of these one-​ year extensions is that, over time, public support will grow (as was the case for the Affordable Care Act), and then Democrats (or the Republicans should they take control) will then extend the programs. Either way, the process is sure to be highly contentious, especially if the Democrats do not add to their small majorities. After all, institutionalizing the expansions would reverse twenty-​five years of post-​welfare-​reform policy toward low-​income Americans. However, even this slimmed down, one-​year extension failed passage in 2021/​ 2022. Nevertheless, the CTC retains strong support among progressives, and an

5

Experiences with Labor Activation  5 expanded CTC is likely to be on the front burner of social/​political debate for many years to come. (Most recently, a Republican expansion was proposed by Senators Mitt Romney (R-​Utah), Richard Burr (R-​NC), and Steve Daines (R-​ MT). Even if it is not made permanent, though, the sharp debate is unlikely to end. This political contest will center on progressive extolling the CTC’s claimed effects on child well-​being and conservatives bemoaning its feared effects on work effort. Expect warring academic and think tank studies supporting one or the other position. This volume seeks to add another dimension to the debate: the almost three decades of experience, in both the United States and Europe, with “labor activation,” the process through which recipients of safety-​net benefits are encouraged to seek work.

EUROPEAN EXPERIENCES US welfare reform, culminating in the 1996 enactment of TANF, was a highly visible example of embedding activation efforts in safety-​net programs. Although the United States did not go further in encouraging work and now seems to be pulling back from the essentials of welfare reform, many European countries pursued similar policies and extended them beyond the US model. (In doing so, most countries applied the full ambit of positive as well as negative work “incentives.”)

The Origins of European Labor Activation Beginning in the 1970s, many European countries experienced troubling declines in employment and labor force participation (similar to those in the United States), with increasing numbers of people receiving government benefits. Across the original 15 members of the European Union (EU-​15),3 between 1970 and 1982, the percentage of the working age population that was employed fell from about 61% to about 57.8% (before beginning a slow increase as more women entered the labor force). For men alone, the decline was much longer and steeper, from about 83.7% in 1970 to about 70.5% in 1994 (Organisation for Economic Cooperation and Development [OECD] Stats Database, 2022). Total labor force participation (of all individuals of working age) increased from about 62.1% to about 67.6% during this period, but only because more women were entering the labor force. For men along, the labor force participation rate declined from about 85.2% to about 78.5 percent. At the same time, in most countries, new highs were reached in the percent of the population receiving government benefits from

3. Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, and the United Kingdom.

6

6  Work and the Social Safety Net unemployment, disability, and welfare/​social assistance programs (OECD Stats Database, 2022) 2021). In response, in the 1990s and early 2000s, many European countries adopted policy reforms aimed at “activating” those recipients apparently able to work; that is, requiring them to actively seek employment or engage in other specified work-​related or job training activities. Neil Gilbert of the University of California (Berkeley) observes how, in that period, European “policies for social protection of the unemployed originally formulated to provide income maintenance in times of duress were recast into schemes for getting people back to work as soon as possible” (Gilbert 2004, 61–​62). Stefano Scarpetta of the Organisation for Economic Cooperation and Development (OECD; see Chapter 2, this volume) describes the framework for such efforts as focusing on “three key goals.” (i) fostering the motivation of job seekers to actively pursue employment (e.g., work incentives, job-​search requirements, and benefit sanctions); (ii) improving their employability (e.g., training and employment rehabilitation); and (iii) expanding their opportunities to be placed and retained in appropriate jobs (e.g., labor market intermediation and programs that support labor demand through wage subsidies or direct job creation). By the end of the first decade of the twenty-​first century, various European nations had made a series of programmatic and administrative changes aimed at activating recipients. As described in Box 1.1, the most common were tighter eligibility conditions, job search requirements, time limits on benefit receipt, benefit step-​downs, earnings disregards and incentives, and enhanced job readiness assistance, job training, and case monitoring. The most common administrative changes were consolidating the administration of services, decentralizing administrative or programmatic authority, synchronizing benefits across safety-​net programs, and outsourcing services (see Chapter 5, this volume). These changes seem to have made a difference, according to numerous studies (many of which are described in this volume). A 2015 report by the Organisation for Economic Cooperation and Development (OECD), for example, concluded that, in countries that had adopted “strong activation approaches” (Australia, Austria, Denmark, New Zealand, Norway, Switzerland, and the United Kingdom), unemployment rates increased, but they stayed “below former peaks” (OECD, 2015). The trend is similar for employment-​to-​population ratios and labor force participation rates. Between 1994 and 2008, the employment-​to-​population ratio in the EU-​15 increased dramatically from about 60.0% to about 67.5 percent. The rate subsequently fell during the economic crisis and in 2014 was about 65.9 percent, or about 9.8% higher than in 1994. Similarly, the overall labor force participation of the EU-​15 continued to rise (even during the economic downturn) up to about 73.8 by 2014. As with the 1970–​994 period, the increase in the labor force participation rate was driven by more women joining

Experiences with Labor Activation  7

Box 1.1  Recent European Elements of Activation Common programmatic changes include • Tighter initial eligibility conditions, which narrowed the pool of individuals who could obtain benefits by, for example, increasing the number of months an individual must contribute to a UI program before becoming eligible • Job search requirements, which require recipients to search for employment as a condition of benefit receipt and, in many jurisdictions, to accept a “suitable job,” when offered • Time limits on benefit receipt, which reduce the period of time that recipients can receive benefits • Benefit step-​downs, which reduce the amount of benefits at designated time intervals or for particular groups of recipients (such as youth) • Earnings disregards and incentives, which do not count a portion of recipient earnings in the determination of eligibility or benefit amount or which provide increased benefits to recipients who find employment • Job readiness assistance, which provides services to recipients aimed at improving employability (such as meeting with job counselors, taking classes on resume writing, and practicing interviews for job) • Job training, which provides recipients who might need new or additional skills with job training services or vouchers for services to improve their employability (often targeted to identified needs) • Enhanced case monitoring, which increases the frequency and intensity of case worker–​recipient engagement on various activation activities Common administrative changes include • Consolidating the administration of services, which, for example, combines the activation and job training services for UI, social assistance, and disability programs into one office (and can include combining national and local offices) • Decentralizing administrative or programmatic authority, which provides local jurisdictions with the flexibility to configure programs to address local needs or resources • Synchronizing benefits across safety-​net programs, which transfers recipients who time-​out of one program (such as UI) to another (such as social assistance) so that recipients do not lose benefits during a transition period • Outsourcing services, which authorizes local jurisdictions to contract with private providers of services, often using results-​based or performance-​based contracting to improve outcomes

the labor force. However, unlike the 1970–​1994 period, the labor force participation rate for men slightly increased, from about 78.5% to about 79.4% (OECD Stats Database, 2022). The same OECD report also cited national examples of activation programs that lowered the receipt of unemployment benefits and increased employment during times of high unemployment (OECD 2015). For example, although

8

8  Work and the Social Safety Net caseloads began to decline before Germany’s reform of its unemployment assistance and welfare/​social assistance programs (the so-​called Hartz reforms) and before the Netherlands’s reform of its disability program, the declines in both countries persisted and did not revert upward to previous levels with the onset of the Great Recession in 2007. In other programs (such as Denmark’s welfare/​social assistance and unemployment program and the Netherlands’s welfare/​social assistance program), the declines occurred significantly after passage, possibly because of a difference between passage and effective dates or the time it took for the changes to be implemented. Moreover, some countries, such as Germany and the United Kingdom, experienced substantial increases in their male labor force participation rate over the past ten years. It is worth noting that such activation policies often seem to enjoy reasonable and sometimes broad political acceptance from the left and right—​if not initially, then over time. In Germany, for example, although the Hartz reforms were initially enacted by the Schroder government, they were continued by the Merkel government. In the United Kingdom, changes instituted by the Cameron government “built on work commissioned by the previous Labour Government and embodied a high degree of continuity with existing policy” (Haddon 2012, 1). Perhaps significantly, in both countries, the changes seem to have maintained the essentials of that nation’s safety net.

Activation in Periods of High Unemployment During the Great Recession of 2008–​2009,4 many questioned whether activation policies should be paused or at least slackened because of the persistently high unemployment. Was it an unfair waste of effort to put recipients through the stress of looking for work at a time of substantially reduced job opportunities? Or was the need to move safety-​net recipients toward work greater than ever, and might doing so help speed a recovery? This issue was the subject of a 2011 conference at the Paris headquarters of the OECD titled “Labor Activation in Times of High Unemployment.” Co-​ sponsored by the University of Maryland (UMD) and the OECD, the conference explored the effectiveness of European and US activation efforts during severe economic slowdowns (such as the one from which the developed world was then emerging). The conference planners started with the assumption that activation efforts should be paused: after all, “There are no jobs.” The message coming from the research papers presented at the UMD/​OECD meeting, however, was more nuanced: Although jobs are often scarce during recessions, encouraging reasonable job-​seeking is an important discipline for recipients (and programs!) and has at least a modest stimulative effect. This is

4. According to the National Bureau of Economic Research (2021), the Great Recession in the United States technically began in December 2007.

9

Experiences with Labor Activation  9 especially true when the downturn affects only some sectors or geographic areas. Hence, even in the midst of high unemployment, most conference participants were supportive of cautiously balanced activation. Scarpetta (Chapter 2, this volume) writes that, even during the Great Recession: “careful evaluations of active labor market policies (ALMPs) suggest that well-​designed and targeted measures can foster the employability of job seekers and their employment in a cost-​effective manner.” Also, in Chapter 11 of this volume, John P. Martin of University College Dublin writes of “public spending on ALMPs having helped to cut unemployment rates,” but later adds that “activation strategies have performed well during the Great Recession and subsequent recovery in some, but not all, countries.” In fact, as Edward Montgomery of Western Michigan University points out in Chapter 12 of this volume, “The positive impact of active labor market policies in reducing unemployment has led many European countries to extend those efforts to reach harder to employ populations such as disability recipients, early retirees, and single parents on social assistance.” Throughout the chapters, a cross-​cutting theme is the different ways policies seek to balance providing protection from undue financial hardship with decreasing dependency and, hence, encouraging economic recovery. Caution is needed because, as a World Bank report warns, “Well-​designed policies can have a positive impact on employment outcomes for participants, but . . . many existing policies have in fact failed to prove effective or cost efficient” (Immervoll 2012, 8). The chapters in this volume describe the different ways in which European countries have sought to strike this balance. They illustrate how such efforts can be approached reasonably and not punitively and with due regard to the underlying weakness of the labor market and the stresses of joblessness. Among other things, this means a greater emphasis on human capital and skills development and a greater recognition of varying job opportunities by geographic area and skill set.

This Volume With the enactment of the unconditional CTC, it became apparent that the experience during the Great Recession told a valuable story about efforts to “activate” the recipients of safety-​net programs while preserving a strong social safety-​net, even in times of high unemployment and even as safety-​net programs were being greatly expanded in response. Therefore, we asked the participants in the UMD/​ OECD meeting to revise and update their papers so that they can inform a post–​ COVID-​19 world. Hence, this volume. As readers will note, many contributors touch on the applicability of this past research to post-​COVID labor markets and policy. As a number note, because of the lockdowns and wide-​scale illness, this was a much deeper recession. Montgomery (Chapter 12, this volume) writes:

10

10  Work and the Social Safety Net The severity and speed of the economic contraction [the pandemic] generated was greater than the Great Recession as numerous countries from China to the United States to the EU imposed lockdowns and adopted various public health measures to slow the spread of the virus. The result of these measures was an unprecedented contraction in both labor supply and demand. Nevertheless, many preexisting policies (discussed in this volume) did much to lessen the damage. For example, according to Scarpetta (Chapter 2, this volume), “Preliminary evidence suggests that the rapid and widespread use of job retention schemes during the COVID-​19 crisis could have contributed to save more than 20 million jobs across the OECD.” In addition, some policies that were introduced in the wake of the Great Recession were enhanced or expanded during the pandemic. Martin (Chapter 11, this volume) and Timo Weishaupt of the University of Goettingen, Henning Jorgensen of Aalborg University, and Alexander Nunn of Leeds Metropolitan University (Chapter 9, this volume) all cite European Public Employment Services rapidly expanding their e-​service capabilities during the pandemic, such as providing video-​based job counseling. At the same time, more fundamental forces may have been unleashed. As Jacob Klerman of Abt Associates notes: “Recovery from the pandemic recession may suggest that different forces are now at work at least in the United States. Some have speculated that the pandemic recession has caused some workers to rethink their willingness to work for low wages. . . . It is simply too soon to distinguish short-​term reactions as the pandemic recession recedes from longer-​term trends.” The scope of the volume is broader than just cash welfare and the CTC. The volume opens with a chapter that sets the stage for examining European labor market polices broadly and another on work incentives and the proper limits of government that seeks to explicate the differences between the perspectives of classical liberalism and modern progressivism. Those chapters are followed by specific sections on unemployment insurance, social assistance, disability programs, and workforce development services. Each chapter has a section on European programs and activation efforts, followed by one on the US equivalents. The volume concludes with two chapters on the wider context of enhanced labor market polices to support and extend the effectiveness of activation efforts. We hope that these chapters will expand thinking about the state’s role in broad labor market issues.

Broad Labor Market Policies In “Activation and Employment Support Policies for Stronger and Fairer Labor Markets,” Scarpetta (Chapter 2) first reviews the implementation of activation

1 1

Experiences with Labor Activation  11 policies by many OECD countries, tracing some of the major themes from the volume’s chapters. In doing so, Scarpetta places activation within the context of the OECD’s comprehensive approach to the concept, that is: “Keeping motivation of job seekers high, while ensuring that work pays. . . . This involves a fine balance between ensuring that unemployment and other benefits protect workers . . ., thus enabling them to smooth consumption and engage in job search, and making sure incentives of benefit recipients to take up a new job remain strong.” Assessing initial activation efforts (made before the Great Recession), Scarpetta concludes that the “available evidence suggests that, in particular, effective implementation of strict job-​search requirements and mandatory participation in PES activities enhance outflows from unemployment benefits.” These efforts, of course, were made during a period of strong economic growth for most OECD countries. The Great Recession, however, “exposed activation strategies to a hard test” of maintaining activation and the full range of ALMP services in a time of increased demand and constrained budgets. Many countries temporarily extended the length of unemployment benefits, but they still maintained such activation requirements as requiring that recipients engage in job search, job readiness activities, and job training. As mentioned above, Scarpetta concludes that “well-​ designed and targeted measures can foster the employability of job seekers and their employment in a cost-​effective manner.” Overall, however, spending on activation programs did not keep pace with the increased number of job seekers, and the spending per unemployed declined. As a result, many countries with strict activation policies faced “capacity constraints,” although some were better prepared and were more successful. Finally, Scarpetta identifies a number of “demand-​side” activation policies that countries used, many of which are beyond the purview of the other chapters. These include “finance for business start-​ups, targeted wage subsidies (including short-​time working schemes), and direct job creation programs.”

Unemployment Benefits Unemployment insurance (UI) is designed to support those who have lost jobs while they seek new employment. “In the absence of UI,” Klerman (Chapter 4, this volume) writes, “the unemployed would have to take (nearly) the first job offered. With UI, they can search more carefully for a job, yielding better job matches (i.e., higher wages and longer job tenure).” However, more generous benefits and longer benefit durations can also create disincentives to seek employment. As such, UI represents an implicit tradeoff between cushioning job loss and encouraging job seeking. UI provides financial assistance to individuals who have lost a job and, in most countries, have also contributed to the unemployment insurance system

12

12  Work and the Social Safety Net while employed. (In the pandemic, many countries relaxed this requirement for individuals employed in the “gig” economy.) Some countries have additional eligibility criteria, such as recipients having been laid off for economic reasons and not “for cause.” Countries vary in the amount of benefits, the maximum period of receipt, required job search or job readiness activities, and the requirements for accepting an offer of a “suitable job.” In “Early Activation in European Union Unemployment Insurance Programs,” Márton Csillag of the Budapest Institute for Policy Analysis and Anna Adamecz-​ Völygi of the Hungarian Academy of Sciences (Chapter 3), review the European research of the effect of the provision of activation services to those who have recently lost employment. Such programs include caseworker meetings, job search counseling, information, nudging, and expanded job search criteria. According to the authors, the type of interventions with the strongest evidence base and that seem to have the largest effect on re-​employment is intensive individual job counseling that takes place soon after an individual has lost employment or been informed of pending unemployment. The authors explain that one reason that these programs may be successful is that they generate both ex ante and ex post effects: ex ante effects because the perceived “costs of being unemployed go up, or, the utility derived from being unemployed goes down” and ex post effects because the programs provided job connections and improved job search techniques and understanding of re-​employment chances. These interventions appear to be effective for all skill levels of job seekers. Also effective are mandatory work experience programs, although these programs appear to be successful because of ex ante effects. The authors write, “Early (mandatory) assignment to active measures seems to speed up job finding rates to some extent because job seekers do not value these programs and may increase their search effort rather than enroll in these programs.” The authors caution, however, that these interventions appear to be more effective during stronger economic periods. Much less effective are programs that merely provide information or send reminders (“nudges”) to job seekers regarding their obligations to search for employment. According to the authors, the research indicates that these interventions are effective only if they are provided immediately upon unemployment. They also find little evidence to support the effectiveness of job training programs for newly unemployed job seekers early in their unemployment spells when compared to providing such programs after six months of unemployment. The authors conclude by discussing the broader effects of early activation programs on the labor market. In some circumstances, job seekers receiving activation services take jobs that would have ordinarily gone to others who did not receive such services (i.e., “general equilibrium effects”) thereby lengthening their unemployment spells and generating no net employment effect for the labor market. As the authors note, an unresolved question is the extent to which and under what circumstances such general equilibrium effects may moderate

1 3

Experiences with Labor Activation  13 the overall positive re-​employment effects, especially “in regions and times when the labor market is weak.” These findings are echoed in the United States. In “Unemployment Insurance After the Great Recession,” Klerman (Chapter 4) summarizes the evolution of the US system and then applies research findings to identify changes that could encourage recipients to seek employment. Klerman says that “spareness” is the “most salient feature” of the US unemployment insurance program: the United States ranks last in the OECD in both the replacement rate of benefits and the maximum period for benefits (although maximum period of receipt was temporarily extended during the Great Recession and for the COVID-​19 pandemic). Another “salient feature,” according to Klerman, is that US programs have a “minimal activation component.” Although UI recipients are required by law to actively search for employment, in practice, “this requirement has traditionally been only very weakly enforced.” (Some European programs are more demanding.) In sum, the weight of the research is that “higher UI benefits and longer potential UI durations lead to longer UI spells” and that these higher benefits and longer durations have disincentive effects on recipients taking available jobs. There is disagreement, however, on the size of these effects. Significantly, spell length appears much more sensitive to specific activation elements (such as required job searches, required check-​ins with caseworkers, and job search assistance [including resume preparation services]) coupled with penalties for failing to comply. Klerman points to “several promising and complementary strategies to lower UI durations while still providing funds for consumption during unemployment.” These include “mandatory maintenance of a job search log with some level of employer job contacts and job offers received,” “mandatory participation in low-​intensity job search assistance,” and a “broadened definition of a ‘suitable job’ and a requirement to accept suitable job offers.”

Social Assistance (“Cash Welfare”) Social assistance programs (called “cash welfare” in the United States, and, in Europe, sometimes also called “minimum income programs” or “programs of last resort”) provide cash and sometimes cash-​like benefits to individuals and families that are unable to support themselves and that do not qualify for unemployment or disability benefits. Eligibility for these programs is means-​tested (often using a “needs standard” or “standard budget”) and may be further limited by program-​relevant characteristic (such as the elderly or single mothers). Typically, they provide lower benefit amounts than unemployment benefits and, until recent decades, were not time-​limited. In “Activation in Eight European Social Assistance Programs,” Ivar Lødemel of Oslo Metropolitan University and Amílcar Moreira of the University of Lisbon

14

14  Work and the Social Safety Net (Chapter 5) describe how, starting in the 1990s, eight European countries modified their social assistance programs from simply providing benefits to actively encouraging recipients to seek employment. The authors focus on what they describe as the “second wave” of activation reforms that took place in the first decade of the twentieth-​first century. They identify two major trends. The first was the strengthening of work requirements, which included assessing recipients’ ability to work; requiring recipients who can work to participate in work-​related activities for a given number of hours per month; making it more difficult for recipients to refuse job offers; and increasing financial incentives to work (including tax credits and the reduction of benefits after a certain amount of time receiving minimum income). The second major trend was improving the delivery of activation services, which included consolidating services into one-​stop career centers, devolving the administration of programs to sub-​national jurisdictions, and the outsourcing of some services to the private sector. During the Great Recession, however, according to Lødemel and Moreira, many of these countries reduced the size of benefits, shrank the size of their activation programs, and “for those who were still offered participation in programs, the positive trend toward more individualized and tailored programs, seen in the years leading up to the crisis, was reversed.” In “Less Activation in US Social Assistance Programs?” Matt Weidinger of the American Enterprise Institute (Chapter 6) traces the transformation of US cash welfare policy from the Aid Assistance to Families with Dependent Children (AFDC) program to the TANF program. He summarizes TANF’s activation provisions and their implementation since its 1996 enactment within the context of broader programmatic and political developments that first supported activation and then led to a shift away from it. In the 1980s, as state welfare caseloads continued their decades-​long rise, political and public pressure grew to “reform welfare.” The initial response was the Family Support Act of 1988, which, as Weidinger describes, introduced a number of voluntary measures meant to encourage individuals to move from welfare to work, including job search assistance, education, and job training. It also provided temporary health benefits and guaranteed childcare for those who left welfare for employment. Welfare caseloads, however, continued to rise, and many Republican and some Democratic leaders turned their attention to mandatory activation programs. By the mid-​ 1990s, President Bill Clinton’s campaign to reform welfare, coupled with Republican victories in Congress, overcame traditional Democrat opposition to mandatory work programs, leading to TANF’s 1996 passage, which, according to Weidinger, “created the strongest work requirements ever included in federal law.” TANF required states to engage a specified percentage of welfare recipients in work (or work-​related) activities and allowed states to reduce welfare recipients’ benefits for failure to comply with the requirements.

1 5

Experiences with Labor Activation  15 States that failed to meet these required work participation rates were subject to a reduction in federal funds. TANF also established a five-​year lifetime limit on federal benefit receipt. Following TANF’s passage, welfare caseloads fell dramatically almost everywhere, and, at the same time, single-​mother employment rates rose and child poverty fell. These positive developments are usually partially credited to TANF, but, of course, other factors were also at work. As Weidinger notes, “The strong economy in the late 1990s and policies designed to make work pay, including expansions to the EITC and the introduction of the Child Tax Credit [CTC], contributed to this improving work and earnings picture.” TANF contains a “caseload reduction credit” under which the share of adult recipients expected to participate in work and activities is reduced by the percentage of a state’s caseload decline. Because of the rapid decline in their caseloads, within a few years, most states no longer faced meaningful work participation requirements and thus no longer had an incentive to engage their remaining caseloads in work-​related activities. Moreover, in many places, so few families remained on assistance that TANF became almost a phantom of its former self. As Democrats regained political leverage at the end of the 2000s, their dissatisfaction with TANF grew, fed by what they perceived as TANF’s failure to respond to the Great Recession. (During the recession, TANF caseloads barely increased despite much higher unemployment and the creation of a $5 billion TANF Emergency Contingency Fund.) In contrast, UI and Supplemental Nutrition Assistance Program (SNAP) caseloads and spending increased substantially. In the years following, attempts to revitalize TANF’s work requirements failed, and TANF was continued through a series of short-​term extensions. As TANF reauthorization efforts stalled, meaningful work requirements remained limited. According to Weidinger, “States continued to claim caseload reduction credits and, when necessary, tap other loopholes. . . . to allow them to satisfy the work participation requirements without engaging a significant share of adults in work activities.” Hence, both parties instead focused their efforts on expanding other safety net programs, including the EITC, SNAP, and the CTC. In the 2020 election, Democrats gained control of the presidency and both houses of Congress, and they quickly took action to sever the connection between cash benefits and work (or work-​related activities). In March 2021, for example, they temporarily expanded and revamped the CTC by increasing the value of the credit and eliminating any work requirements to receive the credit. As of November 2021, the House was considering legislation that would temporarily continue that expansion while permanently repealing the work requirements for collecting CTC benefits. Weidinger concludes that the expanded and revamped CTC, if made permanent, “would effectively replace TANF and its expectation of work activity with a new federal benefit paid to almost all parents regardless of their participation in work, training, or other constructive activity.” Although a major policy shift,

16

16  Work and the Social Safety Net Weidinger notes (and this recounting demonstrates) that this change is actually part of a “decades-​long trend of the TANF program becoming a diminishing component of the American social welfare system” and its “growing irrelevance within that broader policy landscape.”

Disability Programs Disability programs provide assistance to recipients who have lost jobs or have reduced capacity to work because of illness or disability. Countries vary in eligibility rules (mainly if or how much workers have contributed as part of a disability or sickness insurance program), the degree to which employers are required to contribute to the programs, the permanence of the disability determination, and the expectation of work (based on work capability assessments). In “Five Decades of Disability Benefit Policies in Five OECD Countries,” Duncan McVicar of Queen’s University Belfast, Roger Wilkins of the University of Melbourne, and Nicholas R. Ziebarth of Cornell University (Chapter 7) explore the effect of changes in disability policies on the disability caseloads in Australia, Germany, the Netherlands, Sweden, and the United Kingdom. Each of these countries followed a similar pattern (which was different from that for social assistance: a relaxing of eligibility rules in good economic times when there are fewer applicants for disability benefits, followed by an increasing disability caseload during economic downturns, triggering policies that tightened eligibility and encouraged employment, and, then, declines in the growth of the disability caseload). (Countries differed on whether they applied the policy changes retrospectively.) At part of these changes, many countries added activation provisions for disability recipients. The most common were increasing the level of disability benefits for recipients who work, regularly reassessing recipients’ ability to work, requiring employers to bear the initial costs of disability payments, changing the definition of work capacity to be the ability “to carry out any work rather than work in their usual occupation,” and requiring recipients to participate in training or other workforce development programs for recipients deemed capable of working. According to McVicar, Wilkins, and Ziebarth, these policy changes appear to have slowed the increases of disability caseloads during the Great Recession. They point out, “One interpretation of the policy tightening is that recessions have been rendered less important determinants of benefit receipt. In particular, the economic downturn of the late 2000s did not produce a sharp rise in caseloads in any of the five countries examined here.” Similar themes are struck in “Lessons for US Disability Policy from Other OECD Countries,” by Richard V. Burkhauser of Cornell University and Mary Daly of the Federal Reserve Bank of San Francisco (Chapter 8). They document the changes in US disability policy that have contributed to the expansion in

1 7

Experiences with Labor Activation  17 the caseload of disability programs and then examine activation-​related policy reforms in other OECD countries that could provide guidance for the United States. According to Burkhauser and Daly, the number of US disability recipients increased by nearly six-​fold between 1970 and 2015. They classify the explanations for growth into two categories: (1) changes that were not directly related to disability programs, such as “the aging of the population, changes in the underlying severity of disability, the entry of women into the labor force, or the increase in the normal retirement age from 65 to 66 in the Social Security (OASI) retirement program,” and (2) policy changes to disability programs, including “the cyclicality of application rates, the growth in [Social Security Disability Insurance] SSDI benefits relative to wage earnings, and specific changes in rules and their interpretation and implementation over time.” For the latter category, the authors point to legislated changes that loosened eligibility requirements by “expanding the impairments with which a person could medically qualify for the SSDI program,” including those that were more difficult to measure such as mental illness and musculoskeletal conditions. These two categories represent “more than 50% of all newly enrolled beneficiaries.” Daly and Burkhauser then review the experience of other OECD countries in implementing activation-​related policy reforms in their disability programs, identifying four major lessons for the United States: (1) “disability does not mean incapacity,” (2) “incentives affect behavior,” (3) “early intervention reduces flows,” and (4) “hurdles to reform in the US are surmountable.”

Public Employment Services PES agencies (or “workforce development systems”) are the designated government agencies that provide workforce development services such as job training, job search, and job readiness assistance. Such services are important examples of activation efforts, and, therefore, PES agencies are increasingly seen as platforms for assisting the re-​employment of those receiving unemployment benefits and, to a lesser extent, social assistance and disability benefits. Therefore, this volume includes two chapters that describe efforts to improve PES agency functioning, in general and in regard to elements of activation. European research on PES programs is reviewed in “Activation in Public Employment Services in Europe,” in which Weishaupt, Jørgensen, and Nunn (Chapter 9) describe the changes to PES agencies in Denmark, Germany, and the United Kingdom intended to improve the delivery of services (especially as related to activation), analyze the available evidence on the effectiveness of these changes, and identify “common and diverging” elements of the reforms in the three countries. In all three countries, the PES acts as the primary mechanism for the delivery of workforce development–​related forms of activation (including, as mentioned

18

18  Work and the Social Safety Net above, job training, job search, and job readiness assistance). Although the countries differed in their exact approaches, the authors identify four common elements. The first major element is the increased application of New Public Management principles, especially a focus on outcomes (such as job placement, job retention, and wages) instead of just activities and outputs (such as the number of individuals served and the services delivered) and the comparisons of outcomes among similar sub-​national jurisdictions. Performance targets are set differently in each county. The United Kingdom uses a top-​down approach, with the national government setting targets, while Denmark allows municipalities to set their own targets. “In Germany, a middle road is taken with centrally set indicators,” with sub-​national jurisdictions given the authority to negotiate the targets. The second major element is the privatization of services. In Germany, recipients seeking job training are provided vouchers for services from private providers, although, according to the authors, the initial evidence suggests that these vouchers are underutilized. In the United Kingdom and Denmark, the management of one-​stop job centers has been contracted to private firms. According to the authors, because providers receive payments based on outcomes (such as job placements), there are concerns about “creaming” (when providers focus on serving clients with fewer barriers to employment) and “parking” (when service providers expend fewer resources on hard-​to-​serve clients). The third major element is decentralization, which transfers the administration of activation services from the national government to local jurisdictions. In theory, this “key feature” of PES reform gives local jurisdictions more latitude in providing the mix of services and approaches that better serve local recipients. Denmark and Germany have embraced decentralization to a greater degree than the United Kingdom. The fourth major element, especially in Denmark and the United Kingdom, is a shift from job training services (which are more long-​term) to focused, “work-​ first” policies. Germany, on the other hand, continues to emphasize more job training. In “Workforce Development Services in the United States,” Carolyn J. Heinrich of Vanderbilt University (Chapter 10) provides an overview of US workforce development policies and programs, their role in delivering services to UI and social assistance recipients, and evidence on their effectiveness. She also discusses the role of the federal government in improving employment outcomes of recipients, seeding innovative programs, and evaluating programs. To make a serious contribution to activation efforts for the unemployed and those on assistance,5 Heinrich believes, US workforce programs need to be

5. At least some UI and welfare agencies regularly refer recipients to local workforce development offices.

1 9

Experiences with Labor Activation  19 substantially expanded. Per capita federal spending on US workforce programs is substantially lower than in most European countries. Private investment in such programs does not make up for the difference; according to Heinrich, the United States ranks in the middle of European countries on private investment. She writes, “These patterns raise questions about the adequacy of US workforce investment.” Heinrich foresees budgets for workforce programs being constrained in the near future and therefore concludes that it is “increasingly important that spending is well-​targeted in terms of how and for whom it can be most effective, as well as in consideration of where skills shortages lie.” Unfortunately, the available research suggests that many services offered are of limited overall effectiveness. For example, although “vocational education” (or career and technical training) appears effective in increasing adult earnings, this is driven primarily by positive effects for women; there are few, if any positive effects for males and dislocated workers. Furthermore, job search assistance “is more likely to generate positive impacts in the short run that then fade in magnitude with time.” At the same time, there are some very encouraging findings in some of the latest evaluations of sectoral training programs, which focus on higher-​earning industries and occupations and provide soft skills training and training in certifiable skills that can be transferred to other jobs. Heinrich concludes with three “promising strategies” for improving workforce programs: (1) collaborate with employers on the planning and provision of education and training, especially for youth and through apprenticeship programs; (2) support innovative programs through a coordinated effort of evaluation, dissemination, and funding; and (3) promote performance measurement, emphasizing the European practice of “benchlearning” which “aims to facilitate indicator-​ based performance comparisons that support mutual learning and the identification of best practices for improving program efficiency and effectiveness.”

Complementary Active Labor Market Policies The last two chapters in this volume place activation efforts within the broader context of the full gamut of labor market policies and remind us of the importance of “active labor market policies,” both on their own account and in support of activation efforts. (This echoes the findings in the earlier chapters of how “mandatory employment services, earnings supplements, and welfare time limits” worked together to encourage recipients to leave welfare; Berlin 2002.) Traditionally, labor market policies have been divided between active labor market policies (ALMPs) and passive labor market policies (PLMPs). ALMPs are government programs that “actively” intervene in the labor market to help the unemployed find work. They include PES, job training programs, youth programs, short-​time work schemes, subsidized employment, employment incentives, direct job creation, and services for the disabled. PLMPs are government programs

20

20  Work and the Social Safety Net that provide income support to nonworking individuals, including unemployment benefits, disability benefits, early retirement, and social assistance.6 In “Whither Activation Policies? Reflections for the Future,” Martin (Chapter 11) assesses the successes and failures of ALMPs and activation strategies during the Great Recession and the subsequent recovery up to the COVID-​ 19 pandemic, and then identifies four major challenges for the future. “The vast majority of studies,” he concludes, point “to public spending on ALMPs having helped to cut unemployment rates.” As for activation policies, like Scarpetta, he says that they faced a “severe stress test” after the Great Recession and “performed well . . . in some, but not all, countries.” Some countries (such as Finland and Ireland) may have had activation policies but they were not implemented with the same vigor as in other countries and, hence, saw lesser effects. Other countries (such as Spain and Italy) had weak institutions and administrative frameworks that hampered the full implementation of activation efforts. Reflecting on the future, Martin identifies four “challenges” for activation and labor market policy generally: 1.  Extending activation to recipients of other welfare benefits. Originally, activation policies were focused on unemployment insurance recipients, but countries have recognized that activation policies can also be instrumental in assisting social assistance and disability recipients in moving back into the workforce. Such activation policies, especially when targeted at the disabled, however, have much less public support and “the disabled, in particular, have very active lobby groups in all countries, and these lobbies are very reticent about activation.” 2.  Supporting career progression. According to Martin, “Many benefit recipients are activated to take low-​wage jobs which do not offer great career prospects and which may not lift them and their families permanently out of poverty.” Because training programs are expensive (and still may not move recipients into better jobs), other possible policies include continued PES support for individuals after they have found employment and increased incentives for nongovernmental service providers if they place individuals in higher-​paying jobs or jobs with advancement possibilities. However, there is very limited evaluation evidence on what works with such employment retention and advancement programs and much more experimentation in this area is urgently needed.

6. As is apparent, this terminology has not kept up with recent programmatic changes that have added activation efforts to PLMPs. After all, with activation added, they hardly seem passive. As a result, some commentators tend to use the terms “activation services” and “ALMPs” interchangeably, while others try to clarify that PLMPs with work requirements or incentives are “active” elements of labor market policies.

2 1

Experiences with Labor Activation  21 3.  Contracting with private providers of reemployment service. Many countries have expanded the role of private providers, often focusing on specific groups such as the long-​term unemployed or youth. (At the extreme, Australia has completely outsourced these services.) Martin thinks this trend “seems likely to persist,” although it raises a number of questions including “optimal design and monitoring of contracts in order to secure the desired outcomes for the clients at an acceptable cost to the public purse.” 4.  The potential/​ growth of e-​ services. While a few countries (such as Estonia and the Netherlands) have transitioned activation services to online platforms in the years following the recovery from the Great Recession, this shift was given a major boost in many countries by the pandemic. This trend toward greater reliance on e-​services on the part of PES and private providers will persist once the pandemic has passed., Although such services are much cheaper to provide than those offered in person, Martin warns that “we do not know how effective such e-​ services are or for which clients they might work best.” The PES will have to strike the right balance between investing more in e-​services to cut costs and streamline services with the continued need to rely on face-​ to-​face contacts between caseworkers and those job seekers who lack digital skills and/​or access to broadband or who have special needs. This is an aspect of the delivery of activation which is in a state of flux, and it will be vital to monitor the spread of e-​services across countries, how different countries strike the balance highlighted above, and whether greater reliance on e-​services can deliver more cost-​effective outcomes for job seekers and society. Closing out the volume, “Lessons for Labor Policy in the Aftermath of the Great Recession” by Montgomery (Chapter 12) provides a detailed review of the research evidence concerning activation programs. As do the other authors in this volume, Montgomery interprets the evidence to be that, all things being equal, activation can lower unemployment rates (generally by shortening spells), and, on average, it raises later earnings, although he worries about ancillary effects, such as increased enrollment in disability programs. That is only the beginning of Montgomery’s analysis, however. Like other contributors, he finds that the jobs recipients get are not necessarily high-​ paying or ones with opportunities for advancement. He points out that worker protections are lower in the United States than in Europe and “the extent of institutionalized wage setting (by either unions or the government) is fairly limited. In the United States, hiring and firing cost are minimal.” Moreover, he sees a new economic reality in the United States of stagnant wage growth for the median worker, dramatic increases in inequality, and lower economic mobility than in many European countries.

2

22  Work and the Social Safety Net Montgomery therefore calls for an examination of the full range of programs that “help workers and families during dislocations.” For example, Montgomery notes that European countries spend more on labor activation programs as a percent of national income than does the United States and devote a higher percentage of spending on labor market programs to ALMPs (compared to PLMPs) than does the United States. (Private US employers provide roughly the same amount of training services as European ones, but this is, of course, almost entirely in-​service.) Although he endorses greater spending on workforce programs, he is, nevertheless, a realist about many of these programs, writing that: “Too many of our training programs, whether provided by the government or private sector vendors, yield little to no appreciative earning gain.” More broadly, Montgomery provides an agenda for enhanced US labor policies, quoted at length here because it seems like an apt finish to this chapter. The United States and Europe are still struggling with how to maintain our post-​World War II sense of social cohesion in which economic risks and rewards are broadly shared in a more global competitive environment where we must compete with nations paying far less and offering little in the way of a safety net. . . . It will require more targeted job training with stronger ties to employers, enhanced job search requirements and better assistance, investing in and reforming of the Employment Service with increased emphasis on performance, and increased temporary job creation and incentives for sharing. Perhaps an angry public fed up with what many perceive as inaction by the institutions designed to serve them will finally provide the impetus for policymakers to start moving forward on this agenda. Montgomery’s comments are particularly apt as the United States recovers from the COVID-​19 recession. *** This volume is not meant to provide a detailed blueprint for how to encourage work while enhancing the social safety net. Rather, it summarizes important aspects of European experiences as a way to help identify options for the United States in UI, income support, disability, job training, and various other programs (see Box 1.1). It is not that programs from other countries should simply be transplanted here. There are too many economic, social, and political differences for that to be possible, let alone wise. Instead, we hope that reviewing US and European experiences in tandem will help readers understand the nature and scope of the issues involved.

2 3

Experiences with Labor Activation  23

REFERENCES Berlin, Gordon. What Works in Welfare Reform: Evidence and Lessons to Guide TANF Reauthorization. New York: MDRC, 2002. Crandall-​Hollick, Margot, Jameson A. Carter, and Conor F. Boyle. The Child Tax Credit: The Impact of the American Rescue Plan Act (ARPA; P.L. 117-​2): Expansion on Income and Poverty. Washington, DC: Congressional Research Service, July 2021. Executive Office of the President. Fact Sheet: The American Families Plan. Washington, DC: Executive Office of the President, April 2021. Falk, Gene. The Temporary Assistance for Needy Families (TANF) Block Grant: A Primer on TANF Financing and Federal Requirements. Washington, DC: Congressional Research Service, December 2017. Furchtgott-​ Roth, Diana. Introduction and Summary. In United States Income, Wealth, Consumption, and Inequality, ed. Diana Furchtgott-​Roth. New York: Oxford University Press, 2020:1–​9. Gilbert, Neil. Transformation of the Welfare State. New York: Oxford University Press, 2004. Haddon, Catherine. Making Policy in Opposition: The Work Programme, 2007–​ 2010. London: Institute for Government, 2012. Immervoll, Herwig. Activation Policies in OECD Countries: An Overview of Current Approaches. Washington, DC: World Bank, 2012. Lane, Charles. Goodbye, Clinton Welfare Reform. Hello, Child Tax Credit. Washington Post. https://​www.was​hing​tonp​ost.com/​opini​ons/​good​bye-​clin​ ton-​welf​are-​ref​orm-​hello-​child-​tax-​cre​dit/​2021/​03/​09/​22628​de2-​80ed-​11eb-​ 81db-​b02​f039​8f49​a_​st​ory.html Moody, Chris. February 17, 2016. Bernie Sanders =​American Dream Is in Denmark. CNN, https://​www.cnn.com/​2016/​02/​17/​polit​ics/​ber​nie-​sand​ers-​ 2016-​denm​ark-​dem​ocra​tic-​social​ism/​index.html National Bureau of Economic Research. US Business Cycle Expansions and Contractions. https://​www.nber.org/​resea​rch/​data/​us-​busin​ess-​cycle-​exp​ansi​ ons-​and-​contr​acti​ons, 2021. OECD Employment Outlook 2015. Paris: Organisation for Economic Cooperation and Development, 2015. Organisation for Economic Cooperation and Development. Labor Force Participation Rate, https://​data.oecd.org/​emp/​lab​our-​force-​partic​ipat​ion-​ rate.htm, 2022. Organisation for Economic Cooperation and Development. OECD Stats Database, http://​stats.oecd.org/​#, 2022. US Department of Education, National Center for Education Statistics. Table 235.10: Revenues for Public Elementary and Secondary Schools, by Source of Funds: Selected Years, 1919–​20 through 2017–​18, In Digest of

24

24  Work and the Social Safety Net Education Statistics. Washington, DC: US Department of Education, National Center for Education Statistics, 2020a. US Department of Education, National Center for Education Statistics. Table 333.10: Total Revenue of Public Degree-​ Granting Postsecondary Institutions, by Source of Revenue and Level of Institution: Selected Years, 2007–​08 through 2018–​19. In Digest of Education Statistics. Washington, DC: US Department of Education, National Center for Education Statistics, 2020b. US Department of Health and Human Services, Administration for Children and Families. Characteristics and Financial Circumstances of TANF Recipients Fiscal Year (FY) 2019. Washington, DC: US Department of Health and Human Services, 2020. US Department of Health and Human Services, Administration for Children and Families. Welfare Indicators and Risk Factors: Thirteenth Report to Congress. Washington, DC: US Department of Health and Human Services, 2014.

 25

2 ACTIVATION AND EMPLOYMENT SUPPORT POLICIES FOR STRONGER AND FAIRER LABOR MARKETS Stefano Scarpetta

A

s countries around the world are struggling to recover from the COVID-​19 crisis, it is essential to develop and implement effective and well-​tailored policy interventions to provide support to job seekers and reduce the risk of dependency and vulnerability, but also to assist those at risk of displacement for the transition to a new job. (This overview also draws from other papers, e.g. by Immervoll and Scarpetta 2012; Organisation for Economic Cooperation and Development [OECD] 2015; and OECD 2021a.) Across the area covered by the Organisation for Economic Cooperation and Development (OECD), labor markets have been deeply affected by the COVID-​ 19 crisis, with large job losses in some and large reductions in hours worked in others that made extensive use of different forms of job retention schemes. At the end of 2020, around 22 million jobs had vanished in the OECD compared to 2019 and 114 million globally. In the OECD area, despite a gradual recovery, in June 2021, there were still more than 8 million more unemployed than before the crisis, and more than 14 million more inactive people. And while the OECD, like most organizations, has revised upward its employment projections for 2021 and 2022, OECD countries are not expected to regain the pre-​pandemic gross domestic product (GDP) per capita level before the end of 2022 and the employment rate by mid-​2023. Aggregate figures tell only part of the story, of course. At the peak of the first wave of COVID-​19, in the spring of 2020, hours worked declined by more Stefano Scarpetta, Activation and Employment Support Policies for Stronger and Fairer Labor Markets In: Work and the Social Safety Net. Edited by: Douglas J. Besharov and Douglas M. Call, Oxford University Press. © Oxford University Press 2023. DOI: 10.1093/​oso/​9780190241599.003.0002

26

26  Work and the Social Safety Net than 30% in Mexico, more than 20% in Spain, Italy, the United Kingdom, and France, but “only” by 10% in Japan, Korea, and Australia, countries that managed to avoid full lockdowns. Hours worked have since recovered across the board, but, in the first half of 2021, they were still 5% lower than the pre-​pandemic level. But the crisis also had more pervasive within-​country divides cutting across socioeconomic groups. It amplified existing inequalities in labor market outcomes, skills, and opportunities. It accelerated the digital transformation and automation, providing opportunities for many to continue work remotely, but also widening the gulf between workers. Teleworking became mainstream for many high-​skilled workers (almost 40% of them managed to shift to telework in the spring of 2020) but remained peripheral in many low-​skilled occupations (only 16% of them moved to telework). At the beginning of the crisis, low-​skilled workers were more likely to lose their jobs (more than half of the total change in hours worked among them was due to joblessness). High-​skilled workers were more likely to reduce their working time. Other vulnerable groups, such as workers with nonstandard contracts and youth, have been hit hard by job and earnings losses. The youth unemployment rose by twice as much as adult unemployment rate, and in low-​paying occupations, often with less stable labor contracts, one in ten jobs was destroyed across the OECD. Importantly, these groups are often weakly covered by earnings replacement benefits and job retention support, though many OECD governments tried to close these gaps with emergency measures. The COVID-​19 crisis risks leaving deep scars on those who lost their jobs at the beginning of the pandemic, particularly on vulnerable groups marginally attached to the labor market who face major or multiple employment obstacles (such as single parents with young children, people with disabilities, and low-​ qualified youth not in employment, education, or training [NEET]). These groups are often left outside the scope of active labor market policies (ALMPs). Where this is the case, ALMPs should be expanded to provide these vulnerable people with integrated, comprehensive, and individualized support in order to reconnect them with good opportunities in the labor market. More generally, ALMPs play a crucial role in fostering the resilience of a labor market in transformation. The COVID-​ 19 crisis, by increasing caseloads while generating a slump in job vacancies, has posed a significant challenge to public and private employment services. The good news is that about two-​thirds of OECD countries have increased their budgets for public employment services (PES) since the onset of the crisis (OECD 2021b). But increasing spending may not suffice. Among the countries that were more effective in responding to these challenges are those where the infrastructure to scale up support while ensuring quality employment and training services was already up and running. Building this infrastructure takes time and requires medium-​term planning. It will also require harnessing the potential of digital technologies to better identify and match the skill needs

2 7

Activation and Employment Support Policies  27 of firms and the skill sets of workers and to develop appropriate profiling tools without leaving the most difficult clients unserved. Many countries have made substantial progress in this respect during the crisis but additional investments in digital capacity and efficient internal processes are still necessary in many of them. This book provides important insights on the different activation strategies adopted by countries over the past decade and touches on the challenges posed by the COVID-​19 crisis. While activation strategies were generally launched in the OECD countries in the 1990s, with the aim of combating high and often persistent unemployment, their effectiveness in mobilizing the unemployed and other benefit recipients was shown, in particular, in the decade prior to the financial and economic crisis—​a period in which many countries enjoyed relatively strong economic growth and buoyant labor demand, and thus, when the number of job seekers was low relative to the number of vacancies. During this period, early activation measures focused on specific groups facing labor market difficulties (such as low-​skilled youth) in a context of relatively strong overall labor market performance. In the following decade, however, activation principles were extended to cover a wider range of population groups and, increasingly, in a coordinated manner across policy domains. With a rapidly growing number of job seekers and low labor demand, the financial and economic crisis first and then the COVID-​19 crisis exposed activation strategies to hard tests. Re-​employment services were scaled up, especially in the aftermath of the COVID-​19 crisis, to provide support to the greater number of job seekers, and activation procedures were adapted to ensure that job seekers received the appropriate supports. However, while resources for ALMPs increased in response to the crises, often more than in previous major downturns (OECD 2015), spending per unemployed generally declined. This meant fewer resources available to support job seekers at a time when their chances of quick reintegration were weaker, given the limited number of vacancies and high competition among the unemployed. Evidence of the evolution of spending per unemployed is not yet available for the aftermath of the COVID-​19 crisis, but the significant increase in financial resources bodes well for the availability of policies to provide adequate support. The main objective of the activation strategies is to strengthen employment and reduce dependency and vulnerability, but the actual structure and implementation capacity differ widely across countries. The OECD proposed a framework to characterize these strategies (OECD 2015), which focuses on three key goals: (1) fostering the motivation of job seekers to actively pursue employment (e.g., work incentives, job search requirements, and benefit sanctions), (2) improving their employability (e.g., training and employment rehabilitation), and (3) expanding their opportunities to be placed and retained in appropriate jobs (labor market intermediation and programs that support labor demand through wage subsidies or direct job creation).

28

28  Work and the Social Safety Net Keeping motivation of job seekers high while ensuring that work pays are key elements of activation strategies in most if not all countries. This involves a fine balance between ensuring that unemployment and other benefits protect workers against the loss of income from work, thus enabling them to smooth consumption and engage in job search, and making sure incentives of benefit recipients to take up a new job remain strong. No doubt, most jobless people are keen to (re)gain employment, but some may become discouraged with the prospect of finding a suitable job, especially after a long period of unfruitful searching. Simple economic models of labor supply decisions and job search consider out-​of-​work benefit levels as a main factor determining people’s decision to work and exert job search efforts. They establish a de facto wage floor. There are large differences in the level and duration of unemployment and other out-​ of-​work benefits across the OECD countries. In some (e.g., the United States and Canada), the maximum duration of benefits are also extended during periods of severe downturn and increases in unemployment. But the focus of activation strategies has also been concentrated on eligibility conditions. Some of these conditions exclude certain individuals from the group of potential benefit recipients altogether. These provisions serve as an initial “filter” that targets support measures to certain groups. In addition, those entitled to receive a benefit, in principle, often have to comply with specific behavioral requirements that are an integral part of activation strategies: namely, job search activities, participating in interviews and ALMPs, and accepting suitable job offers. These requirements tend to make continued benefit receipt costly for those who are not genuinely seeking to overcome benefit dependency. Well-​defined eligibility conditions can help to ease any tradeoffs between adequate out-​of-​work benefits and maintaining strong labor market performance. Eligibility criteria include waiting periods during which job seekers do not receive benefits, the definition of availability for work, active job search, and suitable work, with financial sanctions for noncompliance with these requirements. Activation strategies for recipients of inactive benefits often involve requiring the target groups to comply with these criteria or, in some cases, participate in training programs if they are not readily available for work. The OECD and the European Commission have collected qualitative information on the strictness of eligibility criteria for unemployment benefits for forty OECD and EU member countries (Venn 2012; Immervol and Knotz 2018). They suggest large differences in eligibility conditions for unemployment benefits: from being relatively lax in Sweden, Canada, Austria, and Finland, to being strict in Portugal, Romania, Slovakia, and Slovenia. Evidence suggests that a large number of reforms were enacted after the global financial crisis, leading to some convergence of policy rules across countries. However, overall measures of the strictness of eligibility criteria have remained relatively unchanged during the recent past (Immervol and Knotz 2018). But what really matters beyond formal strictness is the capacity to enforce the eligibility criteria: setting strict

2 9

Activation and Employment Support Policies  29 formal rules may produce little effect in practice if PES do not have the capacity to manage frequent contacts with benefit recipients or are unable to monitor job search activity in an effective way. For instance, Grubb (2000) argued that the formal strictness of legislation is often an unreliable guide to its actual impact on job search behavior: eligibility criteria can be formulated in very general terms and case managers may exercise considerable discretion when applying them to individual clients. During the COVID-​19 crisis, a number of countries have encouraged job seekers to use the confinement period in the best possible way while taking a more relaxed approach to job search monitoring during the lockdown. For example, obligations such as regular filing for benefits have been maintained in New York, Nevada, and other US states, and searching and applying for work was still required in Brussels (Belgium), Estonia, the Netherlands, and the United Kingdom. In some countries, this was also underpinned by proactive communication with clients to inform them that counseling interviews could be provided over the phone, online, or through video conferencing while face-​to-​face interviews were not possible (e.g., in the Flanders [Belgium] and Finland). However, such requirements could only be applied by PES with well-​developed digital services. However, even in countries that have maintained more stringent approaches, these requirements have often been eased and adjusted to reflect the new circumstances for job seekers with children at home due to (pre-​) school closures or for those in quarantine (e.g., in Austria, Brussels (Belgium), the Netherlands, United Kingdom). Many PES have, however, suspended job search requirements during the confinement periods and lifted sanctions for not demonstrating active job search (e.g., in France, Germany, Portugal, Slovenia, and Sweden) (OECD 2020). Available evidence suggests that, in particular, effective implementation of strict job search requirements and mandatory participation in PES activities enhance outflows from unemployment benefits. In a meta-​evaluation, Kluve (2010) found that programs described as “Services and Sanctions” are particularly effective in yielding positive employment effects. There is also some evidence that, where eligibility criteria are effectively implemented, unemployment benefits act as a positive incentive for participation in the labor market and in job search (OECD 2006, 2015). This is also consistent with the large (25–​50%) caseload reductions following welfare-​to-​work type reforms in some countries, including the United States, Denmark, Ireland, the Netherlands, and the United Kingdom. In most OECD countries, the benefit systems have rules defining the suitability of job offers, requirements to report on the outcomes of job search, the obligation to participate in ALMPs, and sanctions for non compliance with these rules. Langenbucher (2015) suggests dividing the criteria for suitable work into three categories: (1) required occupation mobility, (2) required geographical mobility, and (3) other valid reasons for refusing job offers. An interesting debate is whether continued availability for work during participation in ALMPs should

30

30  Work and the Social Safety Net be required; if taking a job in the market is indeed the priority, then this seems an important requirement. Whereas a clear majority of countries require continued availability for work during ALMP participation for most or all of their programs, only some of them, including, for example, Australia, Denmark, Germany, the Netherlands, New Zealand, Sweden, Switzerland, and the United Kingdom, require active job search during participation. There are also large cross-​country differences in the required reporting of independent job search activity, from fortnightly or weekly, as in Australia, Portugal, and the United Kingdom, to no reporting in countries like Chile, Czech Republic, Greece, Hungary, Italy, and Spain. The empirical evidence seems to suggest that job search monitoring can have a considerable impact on re-​employment rates (see Borland and Tseng 2007), but Van den Berg and Van der Klaauw (2006) suggest that, for relatively well-​qualified job seekers, such close monitoring may simply allow them to shift from informal to formal job search methods without a real increase in the exit rate to work. It is also important to make sure that obtaining a new job does not imply a significant decline in income and risk of moving into a low-​paying job. Efforts have been made in a number of OECD countries to combine strong financial work incentives with adequate support for those with a very low earning potential. For example, most countries operate gradual benefit phase-​outs for individuals who manage to earn only limited amounts (e.g., by working a few hours while looking for a higher-​paying job). These are important steps because they contribute to maintaining strong incentives to supplement benefit income with a small amount of earnings and to seek or maintain at least some link with the labor market. But, de facto, these gradual phase-​out mechanisms often do not provide a genuine incentive to increase employment incomes further. Indeed, there is the risk that steep benefit phase-​outs further up the earnings scale, combined with relatively high taxes or social contribution burdens for nonmarginal workers, imply that earning more “does not pay.” To try to address this problem, an increasing number of countries are considering employment-​conditional (or “in-​work”) benefits or tax credits that support the incomes of workers in nonmarginal employment (see Besharov and Call, Chapter 1, this volume, for a review of different schemes in the United States). While an obvious way to ensure that work does pay is to keep employees’ income tax and/​or contribution burdens low (as in most Anglo-​Saxon countries and Korea), in-​work benefits involve reducing tax burdens below zero for some groups (i.e., the benefit or tax credit exceeds tax/​ contribution burdens). Depending on how in-​work benefits are targeted, they can result in much-​improved incentives for nonmarginal employment for some groups. In some countries, “transitional” in-​work benefits are paid only for a limited period following new employment in a qualifying job. For job seekers facing particular difficulties in (re)gaining employment, it is also important to provide additional support (e.g., counseling and/​or training) to enhance their employability. This generally starts by understanding the

3 1

Activation and Employment Support Policies  31 individual’s situation, aspirations, and employment barriers and creating an action plan involving gradual and incremental steps in the degree of support. Anecdotal evidence from case managers suggests that this approach can restore motivation and promote a return to work even after years of unemployment. But how to identify those at risk of prolonged spells of job search and exclusion? And how to identify suitable interventions to enhance their employability? Moreover, should PES concentrate their scarce resources to those who have become long-​ term unemployed or on those most at risk of becoming long-​term unemployed? Significant progress has been made by OECD countries in developing profiling tools to better tailor services to the different needs of job seekers (OECD 2018). These tools often involve an assessment of claimants’ probability of becoming long-​term unemployed based on their characteristics made at the start of the unemployment spell. Such an assessment helps to identify those in need of additional support at an early stage. Profiling procedures can also provide useful insights to caseworkers for developing individual action plans and personalized assistance for new clients. In practice, different types of profiling are often combined, and countries like Austria, Denmark, New Zealand, and Sweden use statistical profiling tools to support caseworker’s judgement (OECD 2018). For those job seekers needing to overcome specific employment barriers, access to well-​tailored active programs is key. They help job seekers acquire relevant new skills, gain work experience, and gain work motivation. Spending for active labor market programs varies significantly across OECD countries, as does its composition across different programs. Prior to the COVID-​19 crisis, spending as a share of GDP ranged from about 0.11% to 0.15% in the United States and Japan, respectively, on the one end of the spectrum, to around 1% in Finland and 1.3–​2% in Sweden and Denmark. Interestingly, differences across countries in the level of spending for active labor market programs are not closely related to the level of unemployment. This is indeed the case concerning spending on “passive” income support, as many job losers are entitled to unemployment benefits and other forms of income support, but not for active spending, which in most countries is at the discretion of governments. But do ALMPs really foster employability and better labor market outcomes? There exists a large evaluation literature for OECD countries and a number of high-​quality reviews or meta-​studies (e.g., Card et al. 2010). One general finding is that the impact varies substantially between different types of programs but, perhaps more surprisingly, also between different studies looking at the same policy measure. This reflects, in part, the fact that studies used different methodologies and outcome measures, but the heterogeneity of program outcomes remains even when attempting to control for outcome measures. Notwithstanding these limitations, careful evaluations of ALMPs suggest that well-​designed and targeted measures can foster the employability of job seekers and their employment in a cost-​effective manner. However, while some countries achieve low

32

32  Work and the Social Safety Net unemployment with relatively low expenditure on ALMPs, others spend more resources on ALMPs and manage to contain unemployment. When we look at the evaluation of specific interventions, there has been an evolution in the assessed impact and in the quality of the evaluation methodology. Early evaluations, in particular, often yielded disappointing results. This is the case, for example, of training programs: early evaluations often found a negative impact or only a small positive impact on net employment outcomes in the short run, largely due to so-​called lock-​in effects. To address this, at least partially, some countries retain job search requirements during participation in nonvocational courses. This arrangement might reduce the lock-​in effect, although it mainly rewards the final employment impact. Other evaluations have increasingly tracked employment outcomes for five years or more after entry to training programs, finding more reassuring evidence that they have a long-​ term positive impact on participants’ employment and earnings (see the meta-​ evaluation by Card et al. 2010). Good outcomes have been reported for some job search–​oriented shorter programs, but also for vocational training targeted on sectors in demand, longer vocational training, and training in workplaces. At the same time, there are signs that significant investment in general and classroom training pays off further into the long term, perhaps after ten years (OECD 2015). Ideally, programs are focused on identified employer needs, but there is also some evidence in favor of classroom and preparatory programs. It should also be stressed that most evaluations focus on individual programs rather than broader activation strategies. This implies that they cannot account for important interactions between policy areas (e.g., between a training program and the enforcement of requirements to participate in the program), which are indeed the essence of activation. Another important consideration when interpreting impact evaluations is that programs often impact outcomes other than employment. For instance, even with negative effects on longer-​term employment prospects, public works or other forms of direct job creation might be justified as emergency measures during major downturns with high unemployment. At the same time, from an equity perspective, displacement effects are not necessarily a primary concern if a program successfully strengthens employment among the target group. In general, however, there is still very limited information on the distributional effects of different policies and even less systematic evaluation of possible equity-​efficiency tradeoffs. So far, the focus of these considerations has been on supply-​side measures of activation strategies that are important to promote participation in the labor market and a move into employment. But this is only part of the story. Bringing people into employment also means expanding the set of available employment opportunities. This involves addressing demand-​side barriers through actively engaging and assisting employers in hiring and retaining workers, as well as in lifting the barriers to participation (e.g., by improving childcare arrangements and tackling health and other social problems). No doubt, sound macroeconomic

3 3

Activation and Employment Support Policies  33 conditions are essential to foster job creation, in general, and opportunities for job seekers, in particular. But PES also play an important role in this respect by ensuring that clients have access to a wide range of employment opportunities and promoting effective matching between job seekers and employers. Different ALMPs—​including programs to finance business start-​ups, targeted wage subsidies, and direct job creation programs—​also aim at directly increasing job opportunities for the unemployed. Moreover, for certain groups, activation policies need to be combined with childcare, adaptation of the workplace, or health measures to expand the set of available employment opportunities. Job brokering is crucial for matching job seekers with employers posting vacancies, and, in most OECD countries, PES are key actors in this process. There are, however, large differences in the involvement of PES in job matching (OECD 2015). Across the OECD in 2015, only about 10% of individuals who started their new job—​including those from out of the labor market and those changing their jobs—​indicated an involvement of PES, but this average hid large cross-​country variation: in the Nordic countries as well as Slovenia the share was 20% or more, while in Southern European countries the share was less than 4%. A different picture emerges when looking at the share of the unemployed who contacted the PES during the last four weeks to find work: here the share was about 67%, on average, in the OECD, ranging from almost 80% in the Nordic countries, Germany, and Austria, to less than 50% in the Southern EU countries. Beyond job brokering, activation policies may entail specific supports to enhance job opportunities of some groups of job seekers. This includes programs such as finance for business start-​ups, targeted wage subsidies (including short-​ time working schemes), and direct job creation programs, which aim at directly increasing job opportunities for the unemployed. Available evidence suggests that hiring subsidies paid to the employer—​which is the most common type of employment incentive—​tend to have a positive impact on future employment outcomes for their participants. This result, however, also reflects possible selection effects (individual effects are positive, but a hiring subsidy to a group of potential participants has no impact), deadweight effects (when hirings with a wage subsidy would have occurred without the wage subsidy), substitution effects (when hirings with a wage subsidy reduce unsubsidized hirings by the same employer), and displacement effects (when employers who do not use the subsidy reduce employment because they lose business to firms that avail of the subsidy). A specific reference should be made to short-​time working schemes (STWs). These schemes are aimed at preserving jobs at firms experiencing temporarily low demand by encouraging work-​sharing while also providing income support to workers whose hours are reduced due to a shortened workweek or temporary lay-​offs. These schemes were introduced, or significantly scaled up, during the financial and economic crisis in twenty-​five OECD countries as financially constrained firms were at risk of engaging in “excess” layoffs. The take-​up rate varied greatly across countries but, on average, across the OECD countries, it

34

34  Work and the Social Safety Net went from negligible in 2007, to more than 1% of dependent employment in 2009, or more than 4.5 million workers. At its peak, take-​up reached more than 7% of total employment in Belgium, around 4–​5% in Germany and Japan, and around 1–​2% in Austria, the Czech Republic, France, Ireland, Italy, the Netherlands, and Slovakia (Hijzen and Martin 2013). During the COVID-​19 crisis, the use of different forms of job retention schemes was one order of magnitude larger than during the financial crisis, covering in some countries even a third or more of total employees, around 60 million workers across the OECD. A number of studies have tried to assess the impact of STWs at the firm or aggregate level. Firm-​level studies generally suffer from selection bias because firms that participate in STW schemes tend to be less competitive than other firms that can serve as a control group. Aggregate approaches yielded some encouraging results but with important caveats. Early assessments of the widespread use of STWs during the financial crisis (e.g., Hijzen and Venn 2010; Boeri and Brucker 2011; Cahuc and Carcillo 2011) support the view that these schemes contributed to reduce job losses. However, Hijzen and Venn (2010) also suggested that the positive impact of STW was limited to workers with permanent contracts, thereby further increasing labor market segmentation between workers in regular jobs and workers in temporary jobs. Another study by Hijzen and Martin (2013) also looked at the effect of STW schemes over the course of the downturn and recovery and found that the timing of STW is crucial. In particular, the rapid and widespread use of job retention schemes during the COVID-​19 crisis is estimated to have contributed to save more than 20 million jobs across the OECD (OECD 2021a). Further analysis of this exceptional and unique episode is of course required, but this early evidence suggests that a rapid and widespread response, using tools which are well known by all key actors—​ employers and workers, but also PES—​can prove effective in limiting the impact of a major shock. The two final remarks of this chapter regard the fluctuation of resources for ALMPs and the need to evaluate individual programs. On the first point, it should be stressed that while spending on income support for the unemployed is strongly counter-​cyclical—​given the entitlement of many job losers to the unemployment benefits—​this is not the case for spending on active programs that, at the best, react only slowly and partially to large increases in the number of potential beneficiaries. The notable exceptions are the Nordic countries and, in particular, Denmark but also Switzerland. The lack of strong responsiveness means that, in most countries, spending per unemployed person tends to decline significantly as unemployment rises during cyclical downturns. This implies that it becomes more difficult to service job seekers effectively. During periods of extended labor market weakness, as in the weak recovery after the 2008–​2009 economic crisis, falling resources per job seeker were a major concern; at the time when independent job search was more difficult, job seekers depended more

3 5

Activation and Employment Support Policies  35 heavily on job search assistance and other labor market programs, but PES were facing daunting capacity and resource constraints. The more recent evidence of the COVID-​19 crisis suggests that most countries scaled up resources in the aftermath of the crisis and in 2021, often significantly. It is not clear at the time of writing whether spending for unemployed was maintained or even increased, but it is clear that ALMPs have been a key element of the overall policy response. More generally, the Danish model, in which participation in active programs is a duty but also a right for job seekers after a certain duration of unemployment, is interesting; this implies that the government has to put aside adequate resources in relation to the expected unemployment rate (and beneficiaries) in the annual budget. To strengthen the resilience of the labor market and reduce the risks of unemployment becoming entrenched after an economic downturn, governments could consider introducing an automatic adjustment in the resources available for activation measures. It is also important to have, as part of any effective activation strategy, a proper mechanism of evaluation of individual programs and, if possible, the strategy itself. As it is clear from the different contributions of this book and, indeed, the vast empirical literature, ALMPs operate in a complex and evolving environment that requires fine tuning, adaptation, and assessment of complementarities of individual measures. The implementation of activation strategies in OECD countries has generally involved significant changes in labor market policy institutions, legislation, and management principles, as well as in the design of specific programs. This has taken time and often required experimentation and testing. Evaluations should cover various aspects of the implementation of policies and programs and help to understand what effects the policies and programs had, for whom, and why. But we also need to make progress in systematic evaluations of policy packages that are the essence of activation strategies. Evaluations of such packages are complicated, and much of the literature focuses, instead, on the effectiveness of specific measures on individuals’ labor market outcomes, thereby failing to capture potentially sizeable interactions. To sustain and possibly scale-​up policy commitment to activation, systematic evaluations are needed that take into account possible synergies between individual policy elements and help to assess which individual programs might work best if used as a package in combination with other measures.

REFERENCES Boeri, T., and H. Bruecker. Short-​Time Work Benefits Revisited: Some Lessons from the Global Financial Crisis. Economic Policy 26, no. 68 (2011): 697–​765. Borland, J., and Y. P. Tseng. Does a Minimum Job Search Requirement Reduce Time on Unemployment Payments? Evidence from the Jobseeker Diary in Australia. Industrial and Labor Relations Review 60, no. 3 (2007): 355–​378.

36

36  Work and the Social Safety Net Cahuc, P., and S. Carcillo. Is Short-​Time Work a Good Method to Keep Unemployment Down? Nordic Economic Policy Review, 2011. doi:10.6027/​TN2011–​544. Card, D., J. Kluve, and A. Weber. Active Labour Market Policy Evaluations: A Meta-​Analysis. Economic Journal 120 (November 2010): F452–​F477. Grubb, D. Eligibility Criteria for Unemployment Benefits. OECD Economic Studies 31 (2000/​II). Paris: OECD Publishing, 2000. Hijzen, A., and S. Martin. The Role of Short-​Time Work Schemes During the Global Financial Crisis and Early Recovery: A Cross-​Country Analysis. IZA Discussion Paper No. 7291. Bonn, Germany: Institute for the Study of Labor (IZA), 2013. Hijzen, A., and D. Venn. The Role of Short-​Time Work Schemes During the 2008–​ 09 Recession. OECD Social, Employment and Migration Working Papers No. 115. Paris: OECD Publishing, 2010. Immervoll, H., and C. Knotz. How Demanding Are Activation Requirements for Jobseekers? OECD Social, Employment and Migration Working Papers, No. 215. Paris: OECD Publishing, 2018. Immervoll, H., and S. Scarpetta. Activation and Employment Support Policies in OECD Countries: An Overview of Current Approaches. IZA Journal of Labor Policy 1, no. 9 (2012): 1–​9. Kluve, J. The Effectiveness of European Active Labour Market Programs. Labour Economics 17, no. 6 (2010): 904–​918. Langenbucher, K. How Demanding Are Eligibility Criteria for Unemployment Benefits? Quantitative Indicators for OECD and EU Countries. OECD Social, Employment and Migration Working Papers No. 166. Paris: OECD Publishing, 2015. Organisation for Economic Cooperation and Development (OECD). OECD Employment Outlook 2006. Paris: OECD Publishing, 2006. Organisation for Economic Cooperation and Development (OECD). OECD Employment Outlook 2015. Paris: OECD Publishing, 2015. Organisation for Economic Cooperation and Development (OECD). Profiling Tools for Early Identification of Jobseekers Who Need Extra Support; Policy Brief on Activation Policies. Paris: OECD Publishing, 2018. Organisation for Economic Cooperation and Development (OECD). Public Employment Services in the Frontline for Jobseekers, Workers And Employers. Paris: OECD Publishing, 2020. Organisation for Economic Cooperation and Development (OECD). OECD Employment Outlook 2021; OECD Policy Responses to Coronavirus (Covid 19). Paris: OECD Publishing, 2021a. Organisation for Economic Cooperation and Development (OECD). Scaling Up Policies That Connect People with Jobs in the Recovery from COVID-​19; OECD Policy Responses to Coronavirus (Covid 19). Paris: OECD Publishing, 2021b.

3 7

Activation and Employment Support Policies  37 Van den Berg G., B. van der Klaauw, and J. van Ours. Punitive Sanctions and the Transition Rate from Welfare to Work. Journal of Labor Economics 22, no. 1 (2004): 211–​241. Van den Berg G., B. van der Klaauw. Counselling and Monitoring of Unemployed Workers: Theory and Evidence from a Controlled Social Experiment. International Economic Review 47, no. 3 (2006): 895–​936. Venn, D. Eligibility Criteria for Unemployment Benefits: Quantitative Indicators for OECD and EU Countries. OECD Social, Employment and Migration Working Papers No. 131. Paris: OECD Publishing, 2012.

38

3 EARLY ACTIVATION IN EUROPEAN UNION UNEMPLOYMENT INSURANCE PROGRAMS Márton Csillag and Anna Adamecz-​Völgyi

T

his chapter1 focuses on early labor market interventions for job seekers, drawing from the experience of European Union Member States. The prime aim of this chapter is to discuss the empirical evidence on the effectiveness of early activation, where by “early activation” we mean labor market services given to registered unemployed job seekers early in their unemployment period, not later than in the first six months of their unemployment. There has been a long policy discussion about the optimal timing of labor market interventions, including both activation services and active labor market policies (ALMPs) in the past two decades. There is a tradeoff between providing labor market interventions “early” versus “late.” Automatically provided early interventions may be costly if they are provided to all newly unemployed job seekers. Furthermore, they may generate deadweight loss because they would be given also to those who would have found a job anyway (Weber and Hofer 2004). On the other hand, if interventions are provided late, they might not be able to prevent some job seekers from slipping into long-​term unemployment that would radically reduce further chances of finding employment. In a theoretical model of labor market intervention provision, there should be an “optimal” allocation in terms of both the types of interventions

1. This chapter is a significantly shortened, updated, and rewritten version of the study “Early Activation and Employment Promotion” written for Directorate-​General for Employment, Social Affairs, and Inclusion of the European Commission. Márton Csillag and Anna Adamecz-​Völgyi, Early Activation in European Union Unemployment Insurance Programs In: Work and the Social Safety Net. Edited by: Douglas J. Besharov and Douglas M. Call, Oxford University Press. © Oxford University Press 2023. DOI: 10.1093/​oso/​9780190241599.003.0003

3 9

EU Unemployment Insurance Programs  39 matched to the individual needs of job seekers and in terms of their timing as well. Conditional on a well-​established job seeker profiling system, it might be possible to (1) assess the likelihood of becoming long-​term unemployed for each job seeker entering unemployment, and, based on this estimated probability, (2) schedule labor market activation services and measures in a way that maximizes the probability of finding a job and minimizes the costs of interventions. Before turning to the evidence on this topic, we first take a brief look at how early activation emerged in the toolbox of Public Employment Services (PES) in the European Union. The most recent development in the activation strategies of PES is the emergence of services for employees at risk of unemployment, which was initially introduced to prevent a large number of (low-​ skilled) workers from losing their jobs due to technological change. Arguably, these approaches have become even more prominent during the COVID-​19 crisis, as (1) a large number of persons who lost their jobs in the most affected sectors needed to be steered toward jobs with high labor demand, and (2) the crisis showed that persons with low digital skills had a much more difficult time in changing jobs. Given that the evidence on the effectiveness of preventive measures and services is extremely limited, we will not discuss them in this chapter.2 We also need to point out that any activation strategy hinges on finding the instrument that is most effective for a given individual (or groups of individuals). Indeed, with the advent of artificial intelligence (AI)-​based methods and new statistical procedures for Big Data, much progress has been made in statistical profiling of job seekers and targeting of measures (see Desiere et al. 2019). This topic, however, is also outside the scope of this chapter.

HOW EARLY ACTIVATION BECAME FOCAL IN THE TOOLBOX EUROPEAN PES Long-​term unemployment has plagued Europe for several decades,3 and, while there is no clear agreement precisely why “being on the dole” for a pronounced period is detrimental for job seekers, its destructive effect on future reintegration into the labor market is widely documented. Distinguishing whether finding a

2. We need to note that job search and upskilling programs for workers affected by collective dismissal are widespread in Europe, but, given the specificity of the topic, we will also avoid it. 3. Long-​term unemployment is likely partly due to the possibility of receiving Unemployment Insurance benefits, as in most EU Member States the length of the entitlement period and a relatively high replacement rate mean that job seekers receive more than 50% percent of their previous earnings, even after one year of unemployment (see Organisation for Economic Cooperation and Development [OECD] 2021).

40

40  Work and the Social Safety Net

Box 3.1  “Youth Guarantee” in EU Member States In order to prevent unemployment for young persons entering the labor market, one of the most significant policy changes in the European Union was the proposal to establish a “Youth Guarantee” aiming to put forward personalized services and early activation to all job seekers under age twenty-​five. The main points were that (1) all youth between sixteen and twenty-​five not in education or employment (NEETs) could benefit from the services if they registered for the program (effectively register as job seekers at the PES); (2) youth would initially enter a preparatory phase, where they would benefit from counseling; (3) the service provider (the PES) would propose a relevant and substantive offer, such as returning to education for drop-​outs (for those younger than the compulsory school leaving age, typically eighteen) or engaging in an active labor market policy (ALMP) measure. A further goal of the Youth Guarantee was to reach out to more vulnerable young persons and provide them with services (typically, mentoring) that would enable them to finish the measures successfully. The actual implementation of this far-​reaching policy started in 2015 in most EU Member States, and some initial evaluations of the program (or elements thereof) have appeared in the recent years. In this chapter we avoid the topic of measures specifically targeted at youth because this is a very far-​reaching issue and measures are often distinct from those offered to prime-​age job seekers.

job after a protracted period of unemployment is due to an obsolescence of skills caused by being out of work or if those who ended up as long-​term unemployed had lower re-​employment probability would be extremely useful from the point of view of designing early activation programs, but we have rather scarce evidence on this point. In extremely simplified terms, before the Great Recession, most continental European PES followed a similar activation strategy. In the first six to nine months of unemployment, job seekers only received some relatively low-​intensity services (job search clubs, etc.) as well as (automatic) referrals to job offers, and, if they did not find a job during this period, they were placed in active measures (training schemes or subsidized employment). However, the Great Recession forced a number of PES to reconsider this strategy in order to make their activation strategy more cost-​ effective. Due to the high inflow of unemployment insurance (UI) benefit claimants, PES had to consider how to reintegrate considerably more people into the labor market with a limited budget for services and measures. The shift in focus, broadly speaking, meant several interrelated changes. First, PES started developing a more individualized, tailor-​made approach based on the needs of job seekers. Second, there was a shift from remedy to prevention for those most at risk of becoming long-​term unemployed; in other words, a move to provide more and earlier support for this vulnerable group (see Box 3.1). This often means intensified job search counseling followed by active measures. Third, there also was a shift toward not supplying costly active measures

4 1

EU Unemployment Insurance Programs  41 and services closer to the labor market4 in order to decrease the potential for deadweight. Indeed, for these less vulnerable job seekers, online and blended service delivery became more prominent.

Representative Country Examples We highlight three country examples that are instructive, especially given their clear service strategies and the differences across them in their approach to early activation. The legislation relating to registration as job seeker after having received a notice of dismissal is rather exceptional in Germany. As soon as somebody knows that their current employment spell will come to an end, the individual is automatically required to register (in person) at the employment agency, in general three months before the employment is due to end. This means that prevention forms an important part of German PES goals, and the prevention of unemployment is one of the key performance indicators. In line with this, the local PES can deploy a large number of active services and measures for those at risk of becoming unemployed, with the most widely used being job search counseling and job referrals. In Denmark, following the Employment Reform of 2014, there is a well-​defined activation strategy based on a clear-​cut segmentation. For those closer to the labor market, during the first six months of the unemployment spell, the unemployed receive only (intensive) services in the form of monthly (face-​to-​face) meetings with their (personal) job counselor; for this group, active measures are typically only accessed during months 7–​15 of their unemployment spell. The integration strategy for persons further away from the labor market gives a larger role to active measures: they already can have access to a wider range of ALMPs (including longer retraining programs and work experience programs) in the first stages of their unemployment spell. Thus, activation starts at an early stage in Denmark but with different sets of tools based on distance from the labor market. In France, the PES launched a new service strategy in 2013 which clearly emphasizes job seekers’ needs and early (and complex) support for those further away from the labor market. Following an initial interview and profiling, job seekers are assigned one of four service streams. For those closest to the labor market, who are deemed to be autonomous job seekers, contact is kept online, and the client receives job offers and reports job search activity. For those needing some guidance in their job search, there are regular face-​to-​face or telephone contacts, and employment counselors offer job search advice. For those needing reinforced support, face-​to-​face contact between the counselor and the client is ensured once a month, and job seekers start receiving different additional guidance and skills development plans at an early stage in their unemployment 4. This was also facilitated by the development of online service provision, which began to be used intensively for those closer to the labor market and which is considerably cheaper than face-​to-​ face service provision.

42

42  Work and the Social Safety Net spell.5 Finally, for the most vulnerable, for whom the goal is social reintegration, a specific strategy is followed with the aim of holistic support and which is builds on regular contacts with the social services in addition to PES employment counselors.6

DETAILED REVIEW OF THE EMPIRICAL EVIDENCE Although there are several reviews and meta-​analyses of labor market services and measures in the literature, none of them targets interventions based on how “early” or “late” they are given within the first year of unemployment. In terms of the timing of entry to programs, general review papers only look at heterogeneous effects of labor market interventions on the long-​term unemployed who enter these programs after the first year of unemployment.7 Furthermore, although a handful of papers review the effects of labor market programs explicitly targeting the long-​term unemployed, none of these reviews looks at program entry within the first year of unemployment.8 This section provides a detailed review of academic papers focused on early labor market interventions.

The Effects of Caseworker Meetings and Job Search Counseling Individual and group caseworker meetings may have several (potentially overlapping) functions and goals. Meetings to evaluate the labor market potential/​ human capital strengths and weaknesses of job seekers and develop an individual plan are designed to help the matching procedure between job seekers and employers. A clear view of one’s own labor market potential may help to set realistic expectations in terms of positions and expected wages and thus may speed up the job search process. Meetings providing job search assistance aim at developing the job search skills and technologies of job seekers and at increasing their motivation. Meetings to offer explicit employment possibilities/​vacancies may decrease the costs of job search to basically zero. Meetings to monitor the job search efforts of job seekers and employ sanctions if those are not satisfactory are set to ensure that the job search behavior of job seekers reaches some acceptable level. While “monitoring meetings” sometimes occur without the offer of other services, independent of the actual declared goal of a meeting, more frequent 5. Note that those who have not been able to find a job (or have not been placed on an active measure) within nine months after their initial interview are transferred to this service track. 6. For instance, in 2017, around one-​third of registered job seekers in France were in the “closest to the labor market category,” 48% of PES clients were in the guidance category; 16% were in the reinforced support category, and 2% were in the complex support group. 7. To name a few of the most recent review papers: Card, Kluve, and Weber (2018); Kluve (2010); and Martin (2014). 8. Examples include Meager and Evans (1997).

4 3

EU Unemployment Insurance Programs  43 interactions between job seekers and caseworkers always carry the potential for monitoring job search compliance and the threat of potential sanctions. The literature shows that caseworker meetings in general do have a positive effect on the probability of finding a job; however, the size and significance of the effects differ. Maibom et al. (2017)9 look at the evidence emerging from thirty-​ seven papers that estimated the effects of caseworker meetings on labor market outcomes. According to their review, all papers found either positive and significant effects (thirty), or, positive but insignificant effects (seven) of caseworker meetings on labor market outcomes. Services in these studies include monitoring meetings, counseling meetings, job offers, and sanctions given in case of unsatisfactory job search efforts. Schiprowski (2020) looks at the effects of the quantity and quality of early individual caseworker meetings on the length of unemployment in Switzerland by exploiting random variation coming from caseworkers’ absenteeism. When a caseworker is missing from work, job seekers originally allocated to them for that day would either be rescheduled to a later date or reallocated to other caseworkers who are present. In this latter case, the substituting caseworkers would have more clients that day and thus spend less time on each job seeker. Thus, job seekers allocated to non-​absent caseworkers have more and higher-​ quality meetings earlier, while other job seekers have their meetings either later or of a lower quality.10 Exploiting this variation, she finds that job seekers stay in unemployment ten days longer if their caseworker was out of the office for at least ten days in the first three months of their unemployment spell, and the probability of exit goes down by 2.8 percentage points. Caseworker absence in the second three months of unemployment has no significant effect on unemployment duration. She also finds that this effect is larger if the office-​specific absent rate is higher (i.e., if missed meetings can be replaced in a smaller proportion due to capacity constraints). In addition to this, Schiprowski (2020) finds that the absence of more productive caseworkers11 has a higher negative effect on the exit rate from unemployment.12 McVicar (2008) explores how the Jobs and Benefits reform of 1999 in Northern Ireland, which brought variation in the length of counseling and monitoring meetings with a caseworker, affected the exit rate from unemployment to employment. The new reform merged job search assistance and

9. Note that Maibom et al. (2012) is the more detailed IZA Discussion Paper Series version of Maibom et al. (2017). 10. Caseworker absenteeism causes on average one lost meeting in the first half-​year of unemployment, and about half of these lost meetings are replaced by another caseworker. 11. Caseworker productivity is defined based on the six-​month rate of exit from unemployment when there is no caseworker absence. 12. She estimates that the absence of medium-​productivity caseworkers decreases the exit rate by 2.6 percentage points, the absence of low-​productivity caseworkers has zero effect, while the absence of high-​productivity caseworkers causes a 6.4-​percentage point reduction in the exit rate.

4

44  Work and the Social Safety Net monitoring meetings to one location and assigned more time to meetings that took place every two weeks. Moving to a single location required refurbishment of some of the offices, which in turn suspended the monitoring meetings in these offices for, on average, eight months. Hence, the author studies the effect of (1) no monitoring of job search and (2) enhanced job search assistance (and monitoring) versus the regular monitoring meetings. He found that the effect of no monitoring decreased the exit to employment by roughly 23%; however, the effect of enhanced job search assistance and monitoring was negligible. Van den Berg and Van der Klaauw (2006) examine the effects of counseling and monitoring on the probability of exiting unemployment of job seekers entitled to unemployment benefits. This small randomized experiment in the Netherlands provided face-​to-​face caseworker meetings to the treated group, while the control group received only written communication from the employment agency regarding their job search efforts. The researchers show that the intervention increased the individual transition rate of exiting unemployment to employment by 6% (i.e., reduces unemployment by about one week), however, this effect is insignificant. Thus, the authors do not find evidence that the intervention worked. Looking at the heterogeneity of the effects with respect to the intensity of the intervention, they find that the more intensive the intervention, the more likely it is to have some positive effect. Also important is the aim of the caseworker meeting: counseling-​type meetings might be more effective than monitoring-​only meetings or just increased job-​ search monitoring and sanctions alone.13 Based on an experiment conducted in Hungary, Micklewright-​Nagy (2010) shows mixed evidence for the effectiveness of job search monitoring meetings. In this setup, meeting frequency between counselors and newly registered UI benefit claimants was increased from once every three months to once every three weeks, and explicit monitoring of job search efforts was introduced (without a tightening of sanctions). The authors show that the treatment had a positive effect on the re-​employment rate for only a subset of job seekers (middle-​aged women). Behagel et al. (2014) compare the impact of an intensive, individual job search assistance program in a random experiment in France in 2007. While essentially the same program was delivered by both public and private providers, we only concentrate on their findings on the publicly provided arm of the experiment, which is measured against the standard service of the PES. In this six-​month program, instead of the usual procedure of monthly caseworker meetings,14 the experimental treatment offered intensified counseling, including bi-​weekly meetings with “personal advisors” and direct help in job search. The advisors did 13. This was pointed out by Ashenfelter et al. (2005) for the United States, who concluded that increased monitoring without teaching job seekers how to search for jobs more effectively has no positive effects. 14. Caseloads were effectively reduced from 120 job seekers/​caseworker to 40 job seekers/​advisor.

4 5

EU Unemployment Insurance Programs  45 not just search for jobs for the clients, but they also applied for vacancies in their names. The main goal of the treatment was to provide job search assistance, and it did not include increased job search monitoring; however, frequent caseworker meetings may also mean the stricter enforcement of search requirements.15 The target group consisted of job seekers at a high risk of long-​term unemployment. The results show that increased job search counseling increased the probability of finding a job by 11 percentage points after three months and by 10.2 percentage points after six months; most of these jobs were full-​time (lasting at least six months). In a 365-​day window, the overall length of unemployment (and, at the same time, unemployment benefit receipt) decreased by eighteen days. The results of a randomized experiment conducted in Denmark also shed further light on the effectiveness of meetings and job search programs. In the treatment, newly unemployed individuals received an early intensive job search course. Thereafter, they met job counselors bi-​weekly and were assigned to an activation program after four months of unemployment.16 The result was a large increase in job finding rates (by 20–​30%). Disentangling the effect of different program elements, it seems that both job search assistance and individual meetings had a positive effect on the job finding rate, while the activation programs had a negative effect in the short run (see Graversen and van Ours 2008; Vikström et al. 2013). Furthermore, analysis of longer-​term outcomes leads to the conclusion that participation in the program did not decrease re-​employment wages and (slightly) increased job stability (Graversen and van Ours 2011; Blasco and Rosholm 2011). Blasco and Rosholm (2011) further analyze how the positive effect of counseling meetings come about. Specifically, they ask whether it is the rapid reintegration of job seekers and this “early” employment experience that results in participants avoiding the potential “stigma” of an extended unemployment spell or if the counseling leads participants to finding jobs that represent better matches due to a more effective search strategy. Their evidence points to a larger role for the improvement in job seekers’ search “technology,” and they conclude that it is not necessarily the early timing of these meetings that mattered but rather the quality of counseling provided. In a later experiment conducted in Denmark, newly unemployed benefit recipients were randomized into one of three early interventions: (a) during the first three months of unemployment, job seekers had to attend a group job search meeting every week; (b) during the first three months of unemployment, job 15. In this study, the authors do not try to differentiate between the effects of (explicit) counseling and (implicit) monitoring. 16. Note that this is a significant increase in treatment intensity. Job search courses lasting two weeks were assigned after five weeks of unemployment, while in the baseline case they were used irregularly. The baseline frequency of meetings with job counselors was once every three months. The activation programs in principle lasted three months, and these could be a training program or wage subsidy program. In the baseline case, these programs became available at earliest at the sixth month of unemployment (but typically after the twelfth month of unemployment).

46

46  Work and the Social Safety Net seekers had to meet their job counselor six times; and (c) after three months of unemployment, job seekers were assigned a mandatory activation program lasting three months.17 Maibom et al. (2017) evaluated this experiment by looking at the total time spent in employment over a four and half-​year period after the start of the experiment. The authors find that group meetings tend to increase employment, but results are not statistically significant.18 Early activation in general does not increase total time spent in employment, which is due to a lock-​in effect during the first year after treatment. Individual meetings with job counselors had the largest beneficial effect, with those assigned to intensive schedules having spent more than 5% longer in employment. Interestingly, the effect over the initial two years is even larger, being close to 10%. This is primarily due to women who tend to leave unemployment much more quickly and find equally stable jobs relative to the control group while men tend to only slightly increase their job-​finding rates but are successful at landing better, longer-​lasting jobs than the control group. The positive short-​run effect of individual meetings on job finding is also confirmed by the study of van den Berg et al. (2012). Using Danish data, they analyze how the exit rate to employment reacts to the timing of meetings. They find that job-​finding rates increased dramatically during the week a meeting was held, and the effect tended to taper off over the next week. However, when the next meeting was held, the job-​finding rate increased again, and, as a result, the effect of a string of meetings tends to be a gradual increase in the exit rate from unemployment to employment.19 A novel experiment from Sweden (see Cheung et al. 2019) also sheds light on how early intensive caseworker meetings might work. In the experiment, job seekers were randomized into more intensive job search assistance regimes during the first quarter of their unemployment spell, which could mean more caseworker meetings (either in-​ person or on the phone/​ online) or group meetings. Caseworker meetings (but not group meeting) have a positive effect on re-​employment rates (increase of about 10%), and it also seems that the more vulnerable job seekers gain more from this intensified support. Relying on detailed data, the authors show that the positive outcomes are likely due to an increased number of job referrals from counselors to which job seekers apply.20

17. This means empirically that, over a period of one year, in treatment (a) unemployed participated on average in seven additional group meetings compared to the control group; in treatment (b) they participated in five additional individual meetings, and in treatment (c) unemployed took part in an activation program for additional three months compared to the control group. 18. In general, the effect accrues only one year after the start of the experiment, which is a sign that treated persons tend to find more stable jobs but do not exit unemployment more quickly. 19. However, the timing of the first meeting does not matter for the exit to employment. 20. A number of other channels could be ruled out including tighter job search effort monitoring, more job search training, and a larger pool or increased range of vacancies which job seekers apply to.

4 7

EU Unemployment Insurance Programs  47 In Germany, several pilot projects have been conducted in the past to examine the impact of intensified individual support (more frequent contacts and more in-​depth counseling) through reducing caseloads.21 In a large-​scale pilot project known as “1:70,” conducted between 2007 and 2010, average caseloads for counselors serving primarily short-​term unemployed were reduced to 1:40 in fourteen pilot employment offices (as opposed to 1:100 in nonparticipating offices).22 Evaluation results in Hainmueller et al. (2016) suggest that the lower caseloads resulted in better outcomes. In pilot offices relative to comparable non-​ pilot offices, the UI rate decreased by 14% and the re-​employment rate of UI benefit recipients increased by 9%. A final pilot (“Berlin Job-​Offensive,” conducted between 2011 and 2012) concentrated on uninsured unemployed classified as being “near to the market”23 and reduced contact intervals from four months to once a month. An evaluation study (Fertig 2014) examined the impact of the pilot project by comparing outcomes of clients at pilot job centers with matched non-​pilot job centers and found a statistically significant 10% increase in the re-​employment probability in unsubsidized jobs on the primary labor market.

Ex-​Ante Effects of Caseworker Meetings Meetings (and labor market measures in general) might have both ex ante and ex post effects (Maibom et al. 2017). Ex ante effects may be realized when job seekers, knowing that they have to meet their caseworkers soon and thus that their job search intensity will be monitored, are more likely to find a job even before the prospective meeting than do similar job seekers with no forthcoming meeting appointments. The theoretical mechanism behind such effect might be considered as either that the costs of being unemployed go up or the utility derived from being unemployed goes down. The behavioral response of being more likely to find a job might be caused by increased job search efforts, more motivation to find a job, or by lowered reservation wages (Hägglund 2011). Ex post effects are realized if job seekers are more likely to find a job after a meeting than are job seekers who had no meeting. Such effects might be realized due to increased human capital or self-​awareness due to the services given, increased job search efforts and motivation after the meetings, more effective job search

21. There was an earlier small-​scale pilot (known as “FAIR”) conducted in 2005, partly concentrating on insured unemployed, whereby unemployed randomized into the pilot were served by counselors whose caseloads were 1:90 to 1:150 (as opposed to 1:180 to 1:500 for regular clients). Schiel et al. (2008) show that job seekers randomized to be served by the FAIR team relative to unemployed who received regular services in the same local employment offices saw a 22% increase in exits to sustainable jobs during the year of the pilot. 22. The actual caseloads were lower (1:40) than originally planned (1:70) since unemployment decreased significantly in the months before the start of the pilot. 23. We need to note that these job seekers were recipients of UB2 (means-​tested unemployment benefits), and, while they were mostly formally short-​term registered unemployed, many of them were in factual non-​employment for more than a year.

48

48  Work and the Social Safety Net technology, more information on the matching procedure on the labor market, more realistic expectations on positions or wages, or accepting directly offered vacant positions.24 Hägglund (2011) looks at the ex ante impacts of job search meeting groups and individual job search monitoring in three randomized experiments conducted in Sweden in 2004 and finds mixed results. In one of the three experiments, he estimates that random referral to both in-​group job search assistance and individual job search monitoring increased the probability of exit from unemployment by 46% (or reduced the length of unemployment by two weeks) before the actual interventions could have taken place. Allocation to individual job search monitoring meetings alone, on the other hand, had no effect ex ante on exiting unemployment. In two other experiments that targeted specifically young and highly educated job seekers, he finds no ex ante effects. From these results, it seems that group meetings are more likely to have ex ante effects than are individual meetings or job search monitoring and, these effects are more likely to be realized for “general job seekers” rather than on specific groups of job seekers. On the other hand, ex post individual meetings seem to be more effective. Van den Berg et al. (2021) also look at the anticipatory effect of signing an integration agreement (IA), as in their experiment one of the treatment arms contained an early announcement of the date at which the IA will be signed. While the authors find evidence of the threat effect, insofar as job seekers who know in advance about the IA begin to adjust their search intensity at the announcement, while the other job seekers abruptly react to the signing of the IA, empirically this does not matter for the speed at which the two groups are re-​employed.

The Effects of Early Activation Walls Early activation walls are supposed to make use of the “threat effect” of allocating job seekers early to time-​consuming activation measures.25 The evidence from the literature on early activation walls (i.e., the incentive power of the “threat effect”) is mixed: researchers usually find some positive effects on the probability of finding a job, however, whether the gains exceed the costs remains an open question. In particular, earlier investigations of their impacts usually look for exit rate effects and not their effects on the probability of finding stable, long-​term jobs. Also, most evaluations of not just activation walls but also some other labor market measures as well do not look at the quality of the jobs found. Theoretically, 24. Indeed, part of the reason that meetings (specifically job search monitoring meetings) might have an ex post effect on job finding is that, in case of noncompliance with job search requirements, sanctions ultimately leading to significant benefit cuts are initiated after such a meeting. We do not discuss the literature on benefit sanctions here. 25. The notion was originally introduced by Black et al. (2003) who found a positive impact of the threat effect on both employment and wages in the United States in a randomized experiment conducted in 1994–​1996.

4 9

EU Unemployment Insurance Programs  49 there might be a tradeoff between how quickly job seekers find a job and its quality: due to the threat of ALMP placement, job seekers may reduce their reservation wages and accept lower-​quality jobs. In a model of social benefits, if quickly found jobs are of a lower quality in terms of wages (or other parameters), social welfare might not necessarily increase due to such an intervention in spite of positive employment effects. On the other hand, early activation walls might not just lower the reservation wages of job seekers but they might also motivate job search and thus prevent losses in human capital due to long unemployment, which in turn might lead to higher (or, at least, not lower) wages. According to the empirical evidence, there are examples of all such scenarios. The notion of activation walls was originally introduced by Black et al. (2003) who found a positive impact of the threat effect on both employment and wages in the United States in a randomized experiment conducted in 1994–​1996. Hägglund (2009) similarly finds that the threat of intensive ALMP care increased the probability of both finding a job and higher wages in Sweden. The results of the two already cited experiments in Denmark also shed some light on the threat effect of activation. Both experiments included a mandatory assignment to a three-​month program after three to four months of unemployment. It is clear that participation in a program had a negative short-​run effect on re-​employment and only a slight positive effect in the medium term (see Graversen and van Ours 2008; Vikström et al. 2013; Maibom et al. 2017). However, part of the reason that mandatory activation sped up transitions to work is due to a “threat” or compulsion effect (Rosholm 2008). Indeed, the anticipation of having to participate in such a program increases the transition to work prior to actual participation (Vikström et al. 2013; Maibom et al. 2017). This is particularly pronounced if the unemployed person lives far from the job center (Graversen and van Ours 2011), when the potential time-​cost of participation is particularly high. However, the threat effect of activation does not seem to lead to quicker re-​employment when cyclical conditions are unfavorable (Maibom et al. 2017). Interestingly, quicker job-​finding due to the threat effect of activation did not seem to lead to worse-​quality (more unstable or low-​wage) jobs in Denmark.

The General Equilibrium Effects of Caseworker Meetings and Job Search Counseling All the papers discussed above examine the partial effects of labor market interventions because they look at their impacts on the participants of the programs only. These studies implicitly or explicitly assume that the programs have no spillover effects; that is, they do not affect those who did not participate. However, this assumption may not hold if programs are run on a large scale. Job search assistance given to a large number of job seekers, for example, might help program participants to find a job, but it might also affect nonparticipants

50

50  Work and the Social Safety Net negatively by decreasing the number of leftover vacancies.26 Large-​scale programs might affect the whole labor market, average wages, the professional structure of workers, or firm behavior (i.e., vacancy supply) as well. These effects are referred to as “general equilibrium effects” in the literature. Although they are theoretically important in the case of labor market interventions, the empirical evidence on the general equilibrium effects of labor market measures is scarce as measuring spillover effects necessitates large-​ scale, two-​level randomization procedures.27 The paper by Gautier et al. (2018) looks at the effects of a multiregional randomized labor market intervention on nonparticipant job seekers in Denmark. The authors compare labor market outcomes of nonparticipating unemployed job seekers in the treated regions and those of job seekers in nontreated regions in a difference-​in-​differences evaluation framework. They find that the program decreased the probability of finding a job in the treated regions for nonparticipating job seekers. Thus, the program that increased exit rates to employment for its participants (similarly to several other studies) had, at the same time, negative spillover effects on nonparticipants. As a next step, they built a theoretical model to estimate the general equilibrium effects of the program, taking into account both its positive effects on participants and its negative effects on nonparticipants and concluded that, on a very large scale (with a take-​up of greater than 20%), it would have decreased overall social welfare. More direct evidence on displacement effect comes from a recent experiment conducted in Sweden (Cheung et al. 2019), where randomization took place both within and across local labor markets. Using this feature, researchers estimated that about half of the beneficial effect of early intensified meetings came at the expense of other job seekers in the same local labor market. Through the additional vacancy referrals given to the “treated” job seekers, competition for jobs increased, which hurt “nontreated” job seekers especially in weak local labor markets. Using a general equilibrium model of the labor market, the authors show that, on the one hand, full-​scale rollout of intensive meetings would decrease the unemployment rate, but it would have a neutral effect on the overall public budget.28 26. Alternatively, large-​scale wage subsidy programs, for example, might increase the employment probability of beneficiaries, but it might also decrease the employment probability of those who are not eligible. Furthermore, substitution effects might also occur if firms replace regular workers with subsidized job seekers. 27. The seminal paper by Crépon et al. (2013) looks at a job search counseling program for young, educated long-​term unemployed. Given that differing portions of eligible youth were assigned to the program in different micro-​regions, the authors were able to estimate general equilibrium (displacement) effects. In fact, part of the employment gains come from the lower employment chances of youth who did not participate in the program. 28. Clearly, estimating the long-​run welfare effects of the program hinges on a number of crucial assumptions. The most important of these is the proportion of “treated” individuals and the reaction of firms (vacancy creation).

5 1

EU Unemployment Insurance Programs  51

Information, Nudging, and, Job Search Criteria Novel data collections and research show that an important reason that job seekers might slip into long-​term unemployment is biased beliefs about their labor market prospects and the lack of updating of these beliefs (Mueller et al. 2021). Furthermore, we have some evidence that part of the reason that job search programs work is that they contribute to more realistic beliefs and ultimately lead to lowering of job seekers’ “pickiness.” A natural option might be for the PES to provide information and “nudging” on job search strategies as early as possible, and channel targeted information on the state of the labor market to those job seekers who might hold biased beliefs. A more direct form of influencing job seekers’ “pickiness” is the enforcement of wider job search criteria. Many European PES require that after a certain amount of time spent in unemployment, the job seeker accepts a wider range of job offers. This begs the question whether starting with wider search criteria from the beginning of the unemployment spell might lead to positive outcomes and whether it might be particularly effective for those at risk of slipping into long-​term unemployment. The evidence on this issue of providing additional information for job search is rather scare and mixed. In an experiment providing brochures containing information about the labor market situation and motivation for job search early in job seekers’ unemployment spell in Germany (Altmann et al. 2018), there was very minor positive effect overall. However, it was precisely those at risk of long-​term unemployment for whom the largest positive effect could be detected because their employment and earnings rose by about 4% following the provision of the information brochure. Van Landeghem et al. (2017) evaluate a simple and inexpensive intervention in a field experiment in Flanders, in 2014. The treatment consists of sending an invitation to a mandatory group information session regarding the PES and job search (followed by a short face-​to-​face interview with a job counselor), while the regular procedure schedules the same intervention after five months spent in unemployment. The authors argue that such an intervention might work through giving some useful job search information or by reinforcing unemployed persons’ job search effort through higher perceived social norms or higher perceived monitoring. They examine the probability of finding a job in the first 120, 150, and 180 days of unemployment. The paper finds that sending out the invitation did not increase the probability of exit from unemployment in general; however, among low-​educated clients, it increased the probability of exit by 50% and increased the number of days spent in employment by 4.7 working days in the first 120 days. The effects measured in days increase if they look at longer periods: to 6.4 working days in 150 days and to 7.7 working days in 180 days. However, this effect primarily comes as an immediate jump in re-​employment, which tends to fade out

52

52  Work and the Social Safety Net over time.29 They argue that this effect on low-​educated job seekers is indeed important as they are the most vulnerable to long-​term unemployment and the intervention is cost-​effective. The timing of a more direct nudging device is examined based on an experiment by van den Berg et al. (2021). In Germany, it is mandatory for job seekers (and the PES counselor) to sign IAs, which are documents about the rights and duties of job seekers. However, as the authors illustrate, it is more akin to a “reminder” to the job seeker about their job search obligations and is viewed by employment counselors as a way to control the effort of job seekers. In the experiment, the timing of IA signing was randomly assigned to take place one, three, or six months after entry into employment, while PES counselors were instructed to not alter anything else in their integration strategy. The results show that the timing of the IA indeed matters, but only for those with relatively low employability; it has no effect for those closer to the labor market. In particular, delaying IA signing to the sixth month decreases the re-​employment rate of low employability job seekers one year after initial registration by 8 percentage points (which is roughly a 15% decrease), which is strong evidence for using early “nudging” for those with unfavorable labor market prospects. The results of providing more directed job search advice in an online format so that job seekers may consider alternative jobs within similar occupations were positive in a randomized experiment conducted in the United Kingdom (Belot et al. (2019). Indeed, those who searched for relatively narrow occupations and who had been unemployed for a considerable time responded by significantly altering and enlarging the pool of vacancies considered. This ultimately led to more job applications and a more than 40% larger number of job interviews; however, it was not possible to estimate the potential effect on re-​employment. The evaluation of a recent experiment in the Netherlands provides much less support for expanding the range of jobs that job seekers are required to consider. In the experiment, during a caseworker meeting after six months of unemployment, job search criteria leading to the obligation to apply to vacancies that had lower educational requirements was imposed. This was, however, estimated to lead to slower re-​employment and to more unstable jobs.

The Empirical Evidence on the Effects of the Timing of Labor Market Interventions In this section, we review the evidence on the effects of different timing schedules of labor market services and measures. While in the earlier sections we discussed the impacts of specific interventions vis-​à-​vis a control group that received the

29. As the time goes by, the effect of the treatment on being employed decreases (9.4 working days on the 120th, 5.8 working days on the 150th, and 5.2 working days on the 180th day after entering unemployment), and it loses its significance starting from 150 days.

5 3

EU Unemployment Insurance Programs  53 “usual” PES service, in the literature reviewed in this section, we concentrate on the heterogeneity of these effects with respect to the timing of the interventions. Overall, the evidence on timing is mixed. First, the empirical evidence on the effects of intervention timing is very narrow. To our best knowledge, only a handful of papers look at the causal impacts of timing. Second, although the literature does not agree that earlier entry to classic ALMPs is more effective, some papers do find higher effects of job search programs and counseling caseworker meetings in case of early entry. In most of these studies, general composition effects (i.e., those entering programs early may be inherently different from those who enter late) are controlled for; however, as time goes by after entry to unemployment, there might be a sequential selection process at every point in time that is hard to take into account. The first attempt to evaluate the effects of intervention timing is that of Weber and Hofer (2004). They look at the “job coaching” program in Austria, introduced in 1999 as a complementary tool to training programs. The six-​week program concentrated on the training of job search skills as well as leaving time for job seekers to actively search for jobs. Theoretically, it was planned to be available for job seekers in the first four months of unemployment, but practically only 63% of participants entered the program in this period; 22% entered between four months and one year, and the remainder of the participants were unemployed for more than one year. The authors use the timing-​of-​event method to identify the causal effect of entry time on the duration of unemployment.30 They find no variation in the effects of the program within the first year of unemployment. They estimate that job coaching decreased the duration of residual unemployment (the number of days spent in unemployment after entering the program) by around 30% (or about thirty days) if participants entered the program on the 60th, 120th, 180th, 240th, or 300th day of unemployment. When reaching one year spent in unemployment, however, the effect of the program starts to decrease, and results turn negative after 480 days. A further paper finding little evidence on the effects of early versus late entry to interventions comes from Carling and Richardson (2004). They look at eight Swedish labor market programs using detailed administrative data from 1995 to 1997. The eight programs include both services (job search training) and classic ALMPs (see Table 3.1). This evaluation is special in the sense that they do not compare treated groups to a control group that had not received any measures; they compare job seekers who participated in at least one of these eight

30. The timing-​of-​event method is a multivariate procedure (Weber and Hofer 2004). The effect of personal characteristics and other control variables on the timing of entry and the effect of program entry (along with other control variables) on the outcome variable are estimated together, allowing for correlated time-​invariant unobserved individual effects. The identification strategy thus relies on the assumption that while unobserved individual effects are constant in time, the program effect starts only after program entry.

54

54  Work and the Social Safety Net Table 3.1  The evidence on the effectiveness and cost efficiency of different early interventions Intervention

Evidence base

Effectiveness

Cost-​to-​gains ratio

Intensified early support, individual job counseling for non-​disadvantaged

Strong

+​+​

+​+​

Intensified early support, individual job counseling for disadvantaged

Strong

+​+​

+​

Early group job counseling, non-​disadvantaged

Medium

+​

0

Early group job counseling, disadvantaged

Medium

+​

+​

Information and nudging, non-​disadvantaged

Weak

0

0

Information and nudging, disadvantaged

Weak

+​

+​

Intensified early activation (participation in active measures)

Strong

+​

0

Early participation in training programs

Medium

+​/0​

?

programs to each other. For those who participated in more than one program, they evaluate the effects of the first program. They estimate the effects of timing by interacting the program indicators with the length of the unemployment spell at the time of program entry in their duration model and only find significant coefficients on the interaction term in the case of two measures: classroom computer training and a type of wage subsidy. In spite of this, the rank of the effects of each measure on the probability of re-​employment does not change whether or not these interaction terms are included; thus, they conclude that “the timing of placement in programs does not affect the relative efficiency of the programs.” Sianesi (2004) also looks at the effects of 1990s Swedish labor market programs conditional on the timing of entry to the programs. She uses an identification strategy that differs from that of Carling and Richardson because her control group includes job seekers who did not take up any programs until the time when the treated individuals entered the programs (although they might afterward). Her identification strategy relies on the assumption that all characteristics affecting program participation and labor market outcomes are observed (selection on observables), which is a strong assumption. She finds mixed results, including that earlier allocation to labor market programs increases the probability of finding a job but, at the same time, increases the probability of staying in the UI benefit system as well. By contrast, Vikström (2017), who also evaluates the timing and sequencing of active measures in Sweden, finds very little evidence

5 5

EU Unemployment Insurance Programs  55 supporting the notion that earlier program entry leads to larger re-​employment effects. He looks at three types of programs: training, wage subsidies, and work trials, and he controls for dynamic selection into the programs. Fitzenberger et al. (2008) use a methodology similar to Sianesi (2004) to look at three different (West) German training programs: on-​the-​job training (in practice firms), short-​term skills training, and vocational retraining. They examine whether participating in a program at different points of an unemployment spell (0–​6 months, 6–​12 months; 12–​24 months) influences the effect of the program on the medium-​and long-​term employment outcomes of participants relative to nonparticipants. Similar to other evaluations of training programs, all three programs are effective in the long-​run, but they all have a substantive lock-​ in period.31 Most notably, the authors find that the employment outcomes of those who participated in training in the first six months of their unemployment spell are no better, but likely worse, than those who started later. This is due to the fact that the period of initially lower employment rates than the comparison group is much longer for these individuals. In terms of sequencing, early skills training followed by job search assistance seems to do better than the other way around. Lechner and Wiehler (2007) estimate the effects of timing interventions and the effects of participation in a sequence of interventions. Their method is able to take into consideration that participation in a program may depend on past participation in an earlier program and also on some intermediate outcomes in between. We concentrate on their evaluation of the timing for our purposes. They use very detailed Austrian administrative data from 1985 to 2005 and a dynamic evaluation framework that allows for selection in each point in time, conditional on all available contemporaneous and past information in each period, including past participation in labor market programs.32 They look at three types of labor market interventions: job search training, qualification measures developing basic skills, and financial support for courses by external providers. Furthermore, they differentiate between entering each program in the first, second, or third four-​month period (i.e., trimester) of the first year in unemployment. Their main findings include that there is no difference in the effects of any program if entering in the first or second trimester, but program entry in the third instead of the first trimester increases the duration of unemployment in the case of all programs. One potential explanation of this phenomenon could be the lock-​in effect: those who entered a program 31. Clearly, the length of the initial lock-​in period differs across the programs and is related to the length of the program itself. Based on the cumulative employment effect of the programs up to six years after inflow into unemployment, vocational retraining only has a small positive effect (roughly seven additional weeks in employment), while on-​the-​job training has a sizeable positive effect (roughly twenty-​seven weeks more spent in employment), and general skills training has the largest beneficial effect (around thirty-​five weeks of additional employment). 32. More precisely, identification is based on the weak dynamic conditional independence assumption (WDCIA) and in practice it uses dynamic propensity scores (Lechner and Wiehler 2007).

56

56  Work and the Social Safety Net in the third trimester are expected to be in the program for more trimesters to come; thus, they will remain unemployed longer. However, in the case of job search training and qualification measures this difference in outcomes stays significant even after four years so it cannot be explained by lock-​in effect alone. Comparing program entry in the second versus third trimester shows similar effects: earlier program entry decreases the probability of subsequent unemployment by 6–​10% even after 3–​4 years.

CONCLUSION: AN OVERVIEW OF THE RESULTS ON EARLY ACTIVATION The evidence presented above refers to several options for reinforcing timely and tailor-​made support. These can be summarized along the following dimensions: • Intensified services for registered unemployed in the first stage of their unemployment spell (within the first six months) • Ensuring an offer of access to active measures already in the first stage of an unemployment spell • Defining preferred sequences of measures (and services) during the first year of an unemployed person’s registration The most important findings and conclusions for these options are summarized in Table 3.1, which includes a general assessment of the effectiveness and the cost-​benefit ratio. To present an overview on the cost-​effectiveness estimates that have been determined in different formats in the original evaluation studies, a stylized comparison has been used. Interventions that can be deemed highly cost-​effective (i.e., those in which gains to the public budget are likely to be more than double the costs of the intervention) are denoted with “+​+​,” while interventions where gains are likely to outweigh costs are represented with “+​ .” The opposite holds for “−” and “− −.” Last, “?” denotes missing evidence. This stylized comparison is necessary since the different evaluation studies used different approaches to represent the cost-​effectiveness of the interventions. Furthermore, the table indicates that the assessment of different interventions is based on the available evidence, which is quite scarce for some of the options. In particular, in interventions for which only one or two studies are available, or where these studies pertain to only one country, the evidence base is marked as “weak.” Interventions with a relatively rich evidence base (more than two studies from at least two countries, preferably with randomized control trials) are denoted with “strong.” Finally, we present whether different measures’ effectiveness and cost-​benefit ratio differ across disadvantaged (low-​skill) and non-​ disadvantaged (medium and high skilled) groups.

5 7

EU Unemployment Insurance Programs  57 Our reading of the literature is thus: for early access to active measures, the evidence on effectiveness remains mixed about to what extent enrolling in a program during the first six months of an unemployment spell is more beneficial than joining a program later. Early (mandatory) assignment to active measures seems to speed up job finding rates to some extent because job seekers do not value these programs and may increase their search effort rather than enroll in these programs. However, this compulsion via early activation does not seem to be conducive to earlier re-​employment in those regions and at times when the labor market is weak. The optimal sequencing of services and measures is only in its nascent stage, hence no firm conclusions can be reached. The only types of early intervention that have been robustly shown to be effective for increasing unemployed persons’ employment rates is group job search training and early, intensive (face-​to-​face) meetings between the unemployed person and her (personal) job counselor. However, novel evidence on equilibrium effects of these interventions shows that, due to potential displacement effects, intensive meetings are likely more effective in tight labor markets. Furthermore, as can be gleaned from the literature, it is not straightforward to examine the cost efficiency of early activation measures. On the one hand, their impacts might be estimated with large errors already present that makes the calculation of potential gains vague. On the other hand, while the costs of the measures can be known more easily, their benefits are much harder to assess even in the short run. This is rendered even more complicated since potential (negative) spillover effects might need to be estimated, which implies building an equilibrium model of the labor market. Thus, although early activation might be effective, it might not necessarily be also cost-​efficient.

REFERENCES Altmann, S., Falk, S., Jager, S., and Zimmermann, F. Learning about Job Search: A Field Experiment with Job Seekers in Germany. Journal of Public Economics 164 (2018): 33–​49. Ashenfelter, O., Ashmore, D., and Deschenes, O. Do Unemployment Insurance Recipients Actively Seek Work? Evidence from Randomized Trials in Four U.S. States. Journal of Econometrics 125 (2005): 53–​75. Behagel, L., Crepon, B. and Gurgand, M. Private and Public Provision of Counselling to Job Seekers: Evidence from a Large Controlled Experiment. American Economic Journal: Applied Economics 6, no. 4 (2014): 142–​174. Belot, M., Kircher, P., and Muller, P. Providing Advice to Job Seekers at Low Cost: An Experimental Study on Online Advice. Review of Economic Studies 86, no. 4 (2019): 1411–​1447.

58

58  Work and the Social Safety Net Black, D., Smith, J., Berger, M., and Noel, B. Is the Threat of Reemployment Services More Effective Than the Services Themselves? Evidence from Random Assignment in the UI System. American Economic Review 93, no. 4 (2003): 1313–​1327. Blasco, S., and M. Rosholm. The Impact of Active Labour Market Policy on Post-​ Unemployment Outcomes: Evidence from a Social Experiment in Denmark. IZA Discussion Paper No. 5631. Bonn: IZA, April 2011. Card, D., Kluve, J., and Weber, A. What Works? A Meta Analysis of Recent Active Labour Market Program Evaluations. Journal of the European Economic Association 16 (2018): 894–​931. Carling, K., and K. Richardson. The Relative Efficiency of Labor Market Programs: Swedish Experience from the 1990s. Labour Economics 11 (2004): 335–​354. Cheung, M., Egebark, J., Forslund, A. et al. Does Job Search Assistance Reduce Unemployment? Experimental Evidence on Displacement Effects and Mechanisms. Working Paper Series 2019: 25. Uppsala: Institute for Evaluation of Labour Market and Education Policy (IFAU), 2019. Crépon, B., Dejemeppe, M., and Gurgand, M. Counseling the Unemployed: Does It Lower Unemployment Duration and Recurrence? IZA Discussion Paper No. 1796. Bonn: IZA, October 2005. Desiere, S., Langenbucher, K., and Struyven, L. Statistical Profiling in Public Employment Services: An International Comparison. OECD Social, Employment and Migration Working Papers 224. Paris: OECD Publishing, 2019. Fertig, M. Quantitative Wirkungsanalysen zur Berliner Joboffensive. Nurnberg: IAB-​Forschungsbericht, 2014. Fitzenberger, B., Osikominu, A., and Völter, R. Get Training or Wait? Long-​ Run Employment Effects of Training Programs for the Unemployed in West Germany. Annales D’Économie Et De Statistique 91–​92 (2008): 321–​355. doi:10.2307/​27917250. Gautier, P., Muller, P., van der Klaauw, B. et al. 2018. Estimating Equilibrium Effects of Job Search Assistance. Journal of Labor Economics 36 (2007): 1073–​1125. Graversen, B. K., and J. C. van Ours. How to Help Unemployed Find Jobs Quickly: Experimental Evidence from a Mandatory Activation Program. Journal of Public Economics 92 (2008): 2020–​2035. Graversen, B. K., and J. C. van Ours. An Activation Program as a Stick to Job Finding. Labour 25 (2011): 167–​181. Hägglund, P. Experimental Evidence from Intensified Placement Efforts Among Unemployed in Sweden. IFAU Working Paper No. 2009: 16. Uppsala: Institute for Evaluation of Labour Market and Education Policy (IFAU), 2009. Hägglund, P. Are There Pre-​Programme Effects of Swedish Active Labour Market Policies? Evidence from Three Randomized Experiments. Economic Letters 112, no. 1 (2011): 91–​93. Hainmueller, J., B. Hofmann, G. Krug, and K. Wolf. Do Lower Caseloads Improve the Performance of Public Employment Services? New Evidence

5 9

EU Unemployment Insurance Programs  59 from German Employment Offices. Scandinavian Journal of Economics 118, no. 4 (2016): 941–​974. Kluve, J. The Effectiveness of European Active Labor Market Programs. Labour Economics 17, no. 6 (2010): 904–​918. Lechner, M., and S. Wiehler. Does the Order and Timing of Active Labor Market Programs Matter? IZA Discussion Paper No. 3092. Bonn: IZA, October 2007. Maibom, J. Experimental Evidence on the Effects of Early Meetings and Activation. IZA Discussion Paper No. 6970. Bonn: IZA, 2012. Maibom, J., Rosholm, M., and Svarer, M. Experimental Evidence on the Effects of the Early Meetings and Activation. Scandinavian Journal of Economics 119 (2017): 3. doi: 10.1111/​sjoe.12180. Martin, J. P. Activation and Active Labour Market Policies in OECD Countries: Stylized Facts and Evidence on Their Effectiveness. IZA Policy Paper No. 84. Bonn: IZA, 2014. McVicar, D. Job Search Monitoring Intensity, Unemployment Exit and Job Entry: Quasi-​Experimental Evidence from the UK. Labour Economics 15 (2008): 1451–​1468. Meager, N., and Ceria Evans. The Evaluation of Active Labour Market Measures for the Long-​Term Unemployement, Employment and Training. Papers No. 16. Geneva: ILO, 1997. Mickewright, J., and Gy. Nagy. The Effect of Monitoring Unemployment Insurance Recipients on Unemployment Duration: Evidence from a Field Experiment. Labour Economics 17, no. 1 (2010): 180–​187. Mueller, A., Spinnewijn, J., and Topa, G. Job Seekers’ Perceptions and Employment Prospects: Heterogeneity, Duration Dependence and Bias. American Economic Review 111, no. 1 (2021): 324–​363. Organisation for Economic Cooperation and Development (OECD). Benefits in Unemployment, Share of Previous Income, 2021. https://​data.oecd.org/​ benw​age/​benef​i ts-​in-​unemp​loym​ent-​share-​of-​previ​ous-​inc​ome.htmin​dica​ tor-​chart. Rosholm, M. Experimental Evidence on the Nature of the Danish Employment Miracle. IZA Discussion Paper No. 3620. Bonn: IZA, July 2008. Schiel, S., H. Schröder, and R. Gilberg. Das arbeitsmarktpolitische Programm FAIR: Endbericht der Evaluation. In Mehr Vermittlungen durch mehr Vermittler? Ergebnisse des Modellversuchs ‘Förderung der Arbeitsaufnahme’(FAIR), ed. T. Kruppe, IAB-​Bibliothek 312. Bielefeld: Bertelsmann, 2008. Schiprowski, A. The Role of Caseworkers in Unemployment Insurance: Evidence from Unplanned Absences. Journal of Labor Economics 38, no. 4 (2020): 1189–​1225. Sianesi, B. An Evaluation of the Swedish System of Active Labour Market Programs in the 1990s. Review of Economics and Statistics 86, no. 1 (2004): 133–​155.

60

60  Work and the Social Safety Net Van den Berg, G., and B. van der Klaauw. Counseling and Monitoring of Unemployed Workers: Theory and Evidence from a Controlled Social Experiment. International Economic Review 47, no. 3 (2006): 895–​936. Van den Berg, G., Kjaersgaad, L., and Rosholm, M. To Meet or not to Meet (Your Caseworker), That is the Question. IZA Discussion Paper No. 6476. Bonn: IZA, April 2012.van den Van den Berg, G., Hofmann, B., Stephan, G., and Uhlendorff, A. Mandatory Integration Agreements for Unemployed Job Seekers: A Randomized Controlled Field Experiment in Germany. IZA Discussion Paper No. 14026. Bonn: IZA, January 2021. Van Landeghem, B., Cörvers, F., and de Griep, A. Is There a Rationale to Contact the Unemployed Right from the Start? Evidence from a Natural Field Experiment. Labour Economics 45 (2017): 158–​168. Vikström, J. IPW Estimation and Related Estimators for Evaluation of Active Labor Market Policies in a Dynamic Setting. Labour Economics 49 (2017): 42–​54. Vikström, J., Roshol, M., and Svarer, M. The Relative Efficiency of Active Labour Market Policies: Evidence from a Social Experiment and Non-​Parametric Methods. Labour Economics 24 (2013): 58–​67. Weber, A., and H. Hofer. Employment Effects of Early Interventions on Job Search Programs. IZA Discussion Paper No. 1076. Bonn: IZA, March 2004.

 61

4 UNEMPLOYMENT INSURANCE AFTER THE GREAT RECESSION Jacob Alex Klerman

U

nemployment—​actively seeking work but unable to find it—​is a pervasive and important facet of modern economies and their labor markets. In partial response, modern economies provide unemployment insurance (UI): cash payments to those who are unemployed and who meet other program eligibility criteria. Unemployment and UI have been present on both sides of the Atlantic during the past century and the first two decades of this century. Details of unemployment and UI vary by place and time. With respect to place, the varying programs provide policy environments and rich evidence for understanding policy effects. On both sides of the Atlantic, UI programs continue to evolve, particularly in response to short-​term fiscal crises, longer-​term budget pressures, and the ongoing economic development of what were formerly less-​developed countries. The chapter proceeds in five sections. The first section provides a brief overview of the American UI system and its evolution in the twenty-​first century. Drawing heavily on the European papers presented in this volume, the second and third sections present brief high-​level reviews of the voluminous literatures on the economic theory of the effects of UI programs and the empirical evidence on those effects. The fourth section describes the US macroeconomic environment in the years that followed the Great Recession and changes in UI policy. The final section attempts to identify some lessons for the United States from the European experience. In sum, the available evidence suggests that UI programs lengthen unemployment spells and that low-​intensity activation programs—​ when rigorously enforced—​shorten unemployment durations. Jacob Alex Klerman, Unemployment Insurance After the Great Recession In: Work and the Social Safety Net. Edited by: Douglas J. Besharov and Douglas M. Call, Oxford University Press. © Oxford University Press 2023. DOI: 10.1093/​oso/​9780190241599.003.0004

62

62  Work and the Social Safety Net Two major economic shocks—​the Great Recession of the early 2010s and the recession that resulted from the COVID-​19 pandemic in the early 2020s—​both led to temporary and large expansion of UI benefits. In both cases, the expansions appear to have increased unemployment, but in neither case do the expansions appear to have been a major contributor to the high rate of unemployment (and were an even smaller contributor to the low rates of employment).

AMERICAN UI PROGRAMS Like many aspects of American social policy, American UI programs are a federal–​state partnership with their origins in the 1935 New Deal legislation. The federal statute sets the broad outlines of policy, and states set their own policies within the framework provided by the federal statute. In particular, states vary in their replacement rate (the UI benefit relative to earnings when employed) and maximum duration of benefits, as well as in the details of what it takes to qualify for UI benefits. At least compared to European UI programs, the most salient feature of American programs in “normal times” is their spareness. On the eve of the Great Recession, the Organisation for Economic Cooperation and Development (OECD) (2009) placed the American programs last in “generosity.” In “normal times” (see below), the maximum duration of benefits is about twenty-​six weeks and often shorter. The “normal times” replacement rate drifted down during this period. It was 47% in 2009 before the Great Recession, falling to 45% in 2019 before the pandemic recession (US Department of Labor 2021a). Even these figures substantially overstate the generosity of the system. A requirement of recent employment makes about a quarter of job losers totally ineligible. A requirement of sustained and substantial recent employment limits eligibility for many other job losers to considerably less than the maximum twenty-​six weeks. In net, less than a third of the unemployed are collecting UI—​28% in 2019 (US Department of Labor 2021b). Some were never eligible (e.g., recent entrants and reentrants, those who quit their jobs), and some have expended their UI eligibility but have not found jobs. (It should be noted that the Biden administration has proposed to “modernize” the UI system, in part by making it available to more of the unemployed, for longer, with higher benefits; Stone 2021; US Executive Office of the President 2021.) Furthermore, the stated replacement rate refers only to cash earnings (OECD 2009). In the United States, health insurance is usually provided as an employee benefit. Thus, many employees also lose their health insurance when they lose their jobs. The US Department of Labor estimates that health insurance makes up about 8% of total compensation, and all benefits (including health insurance) make up about 31% of compensation (Bureau of Labor Statistics [BLS] 2019). These benefits are not included in the official wage base for UI and in the

6 3

UI After the Great Recession  63 computed replacement rate. Some of the cost of lost health insurance is covered by the federal Consolidated Omnibus Budget Reconciliation Act of 1985 (COBRA) provision that allows the uninsured to continue paying the group insurance rate for up to eighteen months, but the payment is out-​of-​pocket for the employee. Partially offsetting this loss of employer-​provided health insurance is that, under the Affordable Care Act, most of the unemployed are eligible for large subsidies for the purchase of health insurance. Another salient feature of American UI programs is their minimal activation component. By statute, there is a requirement to be available for and actively searching for work; in practice, this requirement has traditionally been only very weakly enforced (e.g., Barron and Mellow 1979; St. Louis et al. 1986). Furthermore, with the switch in enrollment and reenrollment from in-​person to telephone and Internet, along with cuts in funding for the employment service that would have done any enforcement, it is widely believed that the degree of enforcement fell in the late 1990s and early 2000s (O’Leary and Wandner 2005; O’Leary and Eberts 2008; Grubb 2011a, 2011b). This fall in intensity of monitoring was partially offset by the increased use, since 1996, of worker profiling as part of the Worker Profiling and Reemployment Service (i.e., the use of statistical models to identify those likely to have long UI spells and to target them for more intensive services). Use of those statistical models continued and was expanded first as part of the Reemployment and Eligibility Assessment (REA) from 2005 to 2015 (Wandner 2010) and then by the follow-​on Reemployment Services and Eligibility Assessment (RESEA) Program. However, neither program appears to strongly enforce the work test (Klerman et al. 2019). Other activities for the unemployed (e.g., training) are minimal, as is public funding for training more broadly in the United States. I discuss changes that occurred during the Great Recession, in its immediate aftermath, and in response to the pandemic recession below. More broadly, the requirement to actively search for work and accept an offer of employment is weak and appears to be weakly enforced. By federal statute, “The requirement that an individual be available for work does not require an active work search on the part of the individual. States may, however, require an individual to be actively seeking work to be considered available for work, or States may impose a separate requirement that the individual must actively seek work” (20 CFR 604.5). Furthermore, there is only a minimal requirement to accept a job offer. Specifically, only “suitable” job offers need be accepted, where the definition of suitable is determined by state UI Law, taking into consideration “the education and training of the individual, the commuting distance from the individual’s home to the job, the previous work history of the individual (including salary and fringe benefits), and how long the individual has been unemployed” (20 CFR 604.5). In practice, states appear to find it difficult to penalize workers for insufficient work effort or for turning down suitable job offers. Programs to monitor these

64

64  Work and the Social Safety Net requirements are far from universal. Furthermore, there is no obvious or easy-​ to-​identify insufficient work effort or refusing a suitable job offer. Finally, even when a strong suspicion exists (e.g., a remark of the UI claimant), it is difficult to generate the formal documentation required to suspend or terminate benefits. The main UI statute includes a permanent Extended Benefit (EB) period for those who exhaust their period of benefits “during periods of high unemployment.” This Extended Benefit program adds thirteen more weeks to UI benefits (split federal/​state funding), and states have an option to add an additional seven weeks (pure state funding), for a total of twenty (US Department of Labor 2010b). In response to past recessions (including the Great Recession), Congress authorized Emergency Unemployment Compensation (EUC), but this provision required explicit renewal. During the Great Recession, the EUC was enacted on June 30, 2008, with several extensions and modifications, until it expired at the end of 2013 (most recently P.L. 112–​96, The Middle Class Tax Relief and Job Creation Act of 2012, passed February 22, 2012; see Hock et al. 2016; Mastri et al. 2016 for more on recent EUC experience). This EUC was entirely federally funded (and included federal funding for some traditionally split state/​federal benefits). In practice, policy response to the Great Recession had two phases. As noted by Gilbert and Besharov (2011), in the initial phase, “The Stimulus Bill” (American Recovery and Reinvestment Act of 2009 [ARRA]) made the following explicitly temporary changes to make UI more generous: (1) increased the maximum duration of UI receipt to ninety-​nine weeks (nearly two years) “in states with high unemployment,” (2) created a tax credit equivalent to 65% of the cost of COBRA health insurance, and (3) increased the UI benefit slightly and exempted much of it from federal income taxes. Since the initial stimulus bill, the trend was toward making UI less generous and more burdensome. The higher UI payments were in place only for calendar year 2009, the exemption from federal income taxes lasted until December 2010, the COBRA credit ended in late 2011, and the EUC was allowed to expire at the end of 2013. As of 2019, the maximum UI duration in most states was twenty-​six weeks (see Congressional Research Service 2019). In two states, the maximum duration was greater than twenty-​six weeks (twenty-​eight weeks in Massachusetts, thirty weeks in Montana); in eight states, the maximum UI duration was less than—​in some states, much less than—​twenty-​six weeks, in part due to changes adopted during the Great Recession and not undone (twenty weeks in Michigan, Missouri, and South Carolina; sixteen weeks in Arizona; and twelve to twenty-​three weeks in Alabama, Florida, Georgia, Idaho, Kansas, and North Carolina—​depending on the state unemployment rate). Again in response to the pandemic recession, the generosity of the UI system was sharply but temporarily expanded (CBPP 2020). Specifically, in March 2020, the Coronavirus Aid, Relief, and Economic Security (CARES) Act created three new programs:

6 5

UI After the Great Recession  65 • Pandemic Unemployment Assistance (PUA) extended UI to those who would otherwise be ineligible (e.g., insufficient work history, gig workers, the self-​employed). • Pandemic Unemployment Compensation (PUC) increased weekly UI benefits by $600 per week (lowered to $300 per week at the end of 2020), relative to the benefit under regular state UI law (sunsetted September 26, 2021, and a few months earlier in about half of the states). • Pandemic Emergency Unemployment Compensation (PEUC) applied conventional EUC to the pandemic recession. As Grubb (2011a) emphasizes, this cyclical increase in the generosity of American UI programs in response to the Great Recession was in contrast to secular (not cyclical) trends in Europe in the 1990s and early 2000s (see also Gilbert and Besharov 2011; Martin, Chapter 11, this volume; Tatsiramos and van Ours 2014 present case studies of changes in Austria in 1989 and Slovenia in 1998.) During that period in Europe, there was a sharp shift toward less generous programs, increased monitoring, and more low-​intensity labor market programs. The changes in Germany under the Hartz reforms have been widely noted. Several European countries made changes that cut the maximum duration of UI (Netherlands, early and mid-​2000s), cut the amount of the benefit either directly (Italy, in 2012, for longer durations; Germany, in early-​to mid-​2000s under the Hartz reforms), or cut the benefit indirectly by shortening the maximum UI duration and thereafter rolling individuals into a less generous welfare/​ social assistance system (Denmark, between 1994 and 2010; Germany, with the mid-​2000s Hartz reforms; Italy, in 2012; and the United Kingdom, in 2007). In addition, several European countries moved to require more activities from UI recipients (usually some combination of job counseling, programs to improve job search skills, and proof of active job search), with some loss of benefits for failure to comply with the requirements (Denmark, in 1994; Germany, in early-​ to mid-​2000s under the Hartz reforms; Netherlands, in the early-​to mid-​2000s; United Kingdom, starting in the mid-​2000s under the New Deal for Long Term Unemployed and the 2010 Work Programmes).

THE THEORY OF UI The simple argument for UI is that job search takes time: without UI, what will the unemployed live on while searching for their next job? This simple argument—​ like the canonical search models—​implicitly ignores the possibility of saving, borrowing, or spousal earnings. More generally, the broader the eligibility for UI and the higher the UI benefits, the higher consumption will be during periods of unemployment. (There is a moderate-​sized literature on these issues: on the effect of savings and UI on consumption smoothing of the uninsured, see Gruber

6

66  Work and the Social Safety Net 1997, 1998, 2001 and Browning and Crossley 2001; on the effect of UI on savings, see Engen and Gruber 2001; on the effect of UI on spousal employment [what used to be called the “added worker effect”], see Cullen and Gruber 2000.) A slightly more subtle argument would posit that, in the absence of UI, the unemployed would have to take (nearly) the first job offered. With UI, they can search more carefully for a job yielding better job matches (i.e., higher wages/​ earnings and longer job tenure). More generally, the broader the eligibility for UI and the higher the UI benefits, the higher the wages will be when a job is accepted and the longer those jobs will last. Except for concerns about cost, these two arguments—​consumption and better job matches—​both suggest more generous UI programs. However, a conventional moral hazard argument pushes in the other direction. UI—​and more generous UI—​makes unemployment more attractive relative to any particular job. (It is important to note that Chetty’s 2008 work suggests that liquidity constraints are as important as moral hazard. If the liquidity constraints are more important, the less salient is the moral hazard argument, and optimal UI benefits would be higher.) We would therefore expect UI to cause more job exits among the employed (followed by receipt of UI), fewer job-​to-​job transitions without receipt of UI, less intensive job search, longer unemployment durations, and higher UI program costs. In fact, for some purposes, it is useful to think of UI as leisure. Average search intensity is quite low—​approximately an hour per day (Krueger and Mueller 2010). Thus, for those with a relatively high value of leisure, UI may be more attractive than the previous job and most plausible next jobs, leading to deferral of job search and the starting date for offered jobs until late in the period of UI eligibility. This was likely particularly salient during the pandemic recession, when the $600 bump-​up in UI payments made income while unemployed clearly higher than income in the previous job (Ganong et al. 2020). In the formal literature, these insights are derived from (and motivated the development of) job search theory. The canonical model is sufficient to begin our discussion (McCall 1970; Mortensen 1986). People only search for jobs when unemployed, and they become unemployed exogenously (i.e., entering unemployment is not affected by the details of UI). Jobs arrive stochastically with a frequency that is affected by search effort. Search effort is costly. In this model, UI raises income while unemployed and therefore the attractiveness of being without a job. With higher income while unemployed, people can, and do, search longer, leading to better (higher wages, perhaps longer lasting) job matches. In addition, with higher wages people have less of an incentive to search, so they search less intensively. Thus, all else equal, eligibility for UI and the size of the UI benefit lower search intensity and lengthen unemployment spells. Standard generalizations to this theory imply that when UI has a finite period, the unemployed search more intensively as they near the end of the period of UI. The

6 7

UI After the Great Recession  67 higher the probability of finding a job in a given period (with enough effort), the longer many of the unemployed will defer search and the later they will find jobs. This is the canonical model. The crucial insight is to embed it in a government budget constraint (Baily 1978). Consider a group of identical workers and a tax on earnings sufficient to cover the UI benefits (and any cost of administering the UI program). Now the polity faces a tradeoff. Higher UI benefits provide more income support while unemployed; however, higher UI benefits also induce moral hazard—​the unemployed do not search as intensively as they would if UI benefits were not available. In the absence of a way to distinguish those who “could” find a job from those who cannot find a job, the UI system must balance income support while unemployed against moral hazard (i.e., people will remain unemployed longer/​take more leisure at the expense of the program). There are no good answers. Note that the previous model assumes that the unemployed will return to work at the end of their unemployment spell. In practice, some become discouraged and leave the labor force neither employed nor unemployed (formally defined as not employed but actively searching for work). For some this is temporary; for some it is early retirement or the beginning of a transition to receipt of disability payments. In as much as (or perhaps better once) the plan is (becomes) to leave UI for retirement or disability insurance, then using all weeks of UI eligibility will often be an optimal strategy. Furthermore, the unemployed/​ out-​of-​the-​labor-​force distinction is fuzzy. UI receipt appears to lead to people reporting being unemployed (meaning that people reported having actively been searching for a job on a labor market survey) (Rothstein 2011; Farber and Valetta 2015; Farber, Rothstein, and Valetta 2015). This pathway increases the measured unemployment rate but without any first-​order impact on employment (there may be a second-​order impact in that the availability of UI benefits increases transitions to retirement or disability benefits). The previous discussion implicitly assumes that it is not possible to distinguish those who could find a job if they searched intensively (or have already found a job) from those who would not have found a job even if they had searched intensively from the beginning of their unemployment and therefore need the consumption safety net that UI provides. As in the welfare literature (Besley and Coate 1992), one way out of this tradeoff is to impose nonmonetary costs of collecting UI. If those costs are productive (e.g., improve job search skills), even better; however, even costs that are a net social loss (i.e., the programs cost money to run but do not improve job search skills or work skills) may be useful in sorting those who truly need the benefit from those who do not. Since much of the value of UI receipt is increased leisure, one natural approach to making UI less attractive while retaining the cash safety net is to take away the leisure. Ways of taking away the leisure might include mandatory job search, job skills training, or even unpaid “make work” (Fredriksson and Holmlund 2006a, 2006b). Furthermore, mandatory job search will often be productive (i.e., it will

68

68  Work and the Social Safety Net sometimes lead directly to more effective job search and therefore shorter UI spells). Finally, and crucially, there is an emergent literature on the optimal pattern of UI benefits over the business cycle (Kiley 2003; Sánchez 2008; Andersen and Svarer 2009; Landais et al. 2010; Schmieder et al. 2012; Kroft and Notowidigdo 2011; but see Mitman and Rabinovic 2011, who suggest the opposite). This literature confirms earlier conjectures that UI benefits should become more generous—​higher benefits and longer maximum duration—​during recessions (e.g., Krueger and Meyer 2002; in contrast to Ljungqvist and Sargent 1998, 2008). Heuristically, the argument is that jobs are harder to find during recessions, so the moral hazard is less severe. Evidence (reviewed below) that exits from UI to employment are not very sensitive to UI payment duration are consistent with this perspective (in contrast, exits to out-​of-​the-​labor-​force appear to be moderately sensitive to UI payment duration). (See Martin, Chapter 11, this volume, who cites cross-​national evidence that, nevertheless, more generous programs appear to lead to larger growth in unemployment rates during recessions, suggesting that the within-​country studies may be missing something.)

EMPIRICAL EVIDENCE ON THE EFFECTS OF UI POLICY The empirical literature on the effects of UI policy is vast. Several recent and good reviews exist (e.g., Fredriksson and Holmlund 2006a, 2006b). Additional recent research reviews and augments that literature (Grubb 2011a; Tatsiramos and van Ours 2014; Røed 2012). The discussion here provides only the highest-​ level summary and makes three main points. First, the evidence that the size and length of the benefit affect unemployment is overwhelming. High-​quality empirical studies consistently show that higher UI benefits and longer potential UI durations lead to longer UI spells. There is strong empirical evidence supporting these theoretical presumptions, though the magnitude of the estimated impacts varies widely. The key empirical issue is finding exogenous variation. Cross-​state and cross-​time variation in American UI programs supports a difference-​in-​difference strategy. Holmlund (1998) provides surveys suggesting that the elasticity of duration with the benefit is 0.3 to 0.9. Grubb (2011a) surveys parts of the American literature. Tatsiramos and van Ours (2014) survey the recent European literature that exploits variation over time in program details for population subgroups to generate difference-​ in-​difference estimates. For the elasticity of duration with respect to the amount of the benefit, recent elasticity estimates (i.e., percentage change in duration for percentage change in benefit amount) from studies in Austria, Finland, Norway, and Sweden are at the top of the American range, often close to 1.0 (Carling, Holmlund, and Vejsiu 2001; Lalive, van Ours, and Zweimüller 2006; Røed and Zhang 2003; and Uusitalo and Verho 2010).

6 9

UI After the Great Recession  69 For the relation of UI duration to maximum duration, research from three European countries (Austria, Germany, and Slovenia) show that estimates vary, with a median of about 0.10—​another week of eligibility increases average duration by roughly a tenth of a week (Lalive 2008; Lalive and Zweimüller 2004; Lalive et al. 2006; van Ours and Vodopivec 2006; Schmieder, Von Wachter, and Bender 2012). Plausibly, there is some evidence that maximum duration affects exit rates more at longer durations, while benefit levels affect exit rates more at shorter durations. For the United States, there is evidence of a decline over time in the sensitivity of UI duration to maximum duration. This decline has been attributed to a decline in temporary layoffs, which was often deliberately scheduled for the length of the available UI benefit (Katz 2010). In addition, many studies find an increase in exit rates from insured unemployment toward the end of the eligibility window. Evidence of a spike can be found in Japan (Kajitani 2008) and throughout Europe, including Belgium (Cockx and Ries 2004), France (Dormont et al. 2001), Norway (Røed and Zhang 2005), Poland (Adamchick 1999), Portugal (Portugal and Addison 2003), Slovenia (van Ours and Vodopivec 2004), and Spain (Alba-​R amirez et al. 2007). The exception is the negative finding for Austria (Card et al. 2007b); Grubb (2011b), however, critiques these results as they relate to the broader Austrian safety net. For the United States, Farber and Valetta (2015) find evidence of a spike for their earlier period (2000–​2005), but not for the Great Recession. Evidence that offering bonuses to leave UI early lead to increased exit rates is also consistent with a latent ability to find a job for a fraction of the UI population (Meyer 1995; but see Card et al. 2007b who argue that many of those exits are not to employment). A thinner literature also suggests increased in-​flows into UI (Winter-​Ebmer 2003; Lalive and Zweimüller 2004). There is some evidence that higher benefits and longer potential durations lead to better job matches (higher earnings and longer job durations), but the evidence is stronger in Europe than in the United States. The Europeans, with access to richer administrative data, have studied this issue in more detail. For Portugal, Centeno and Novo (2007) find the expected negative effect. Finding worse jobs toward the end of benefit eligibility is also consistent with this theoretical expectation (for Germany, see Caliendo et al. 2009; for the United Kingdom, see Petrongolo 2009). Findings of no effect include Lalive (2007) and Card et al. (2007b), for Austria—​but see Grubb’s (2011b) critique of this paper—​and van Ours and Vodopivec (2008) and Fitzenberger and Wike (2010), for Germany. Data limitations have focused the US literature on UI durations but not on the quality of post-​unemployment jobs. It should be noted, however, that American evidence suggests that earnings have very high variance, such that quite large samples (larger than are available in many of these studies) would be needed to detect any—​likely modest—​impacts on earnings, wages, and job tenure.

70

70  Work and the Social Safety Net Finally, for monitoring of job search and low-​intensity mandated programs (reviewed below), there is more consistent evidence of negative impacts on subsequent jobs (Arni et al. 2009; van den Berg and Vikstron 2009). Given the evidence of low search intensity—​approximately an hour a day (Krueger and Mueller 2010)—​it is plausible that on-​the-​job search while employed would be nearly as effective as among the unemployed. Second, the magnitude of these disincentives to taking a job is a matter of some debate. From one perspective, while the estimates are consistently statistically significant and in the expected direction, their absolute magnitude seems small. Here the study by Card and Levine (2000) is representative. They estimate that permanently extending UI from twenty-​six to thirty-​nine weeks would increase the average duration of unemployment by only about 1.3 weeks, similar to the European estimates discussed above. Similarly, Moffitt (1985) and Meyer (1990) estimate that a 10% increase in the UI benefit would lead to an increase of 4–​8% in UI durations. From a complementary perspective, we might ask what UI policy parameters informed workers would choose, knowing that they would pay the average value of the benefits in the form of lower wages. Basic UI operates through a trust fund such that payouts are raised by taxes on employees. Nominally, some of the taxes are paid by employers, but standard tax incidence analysis implies that most of the taxes nominally paid by employers are in fact paid by employees in the form of lower wages. However, extended UI is paid from general federal tax revenues. As such, to a great extent, it is a transfer from wealthier Americans (who pay most of the taxes) to middle-​and lower-​class workers who experience most of the unemployment and receive most of the UI benefits. From that perspective, workers trade the cost of the moral hazard of the insurance against the inability to find a job. A moderate-​sized literature has tried to compute the optimal level of the UI benefit. Those computations are quite sensitive to the details of the model; details at a level well beyond what is discussed in this simple overview. Some studies suggest that UI is too generous (e.g., Baily 1978; Fredriksson and Holmlund 2001, 2006a, 2006b; Boone et al. 2007); others that it is not generous enough (e.g., Chetty 2006, 2008). Finally, there is very strong evidence—​from both the United States and Europe—​that low-​intensity activation programs (e.g., a requirement to provide evidence of active job search or a requirement to participate in some minimal training in job search, such as resume preparation or using computer job boards) and requirements to check in with a caseworker, along with strong penalties for failure to comply (usually termination of the UI benefit) affect UI durations and UI caseloads. Tatsiramos and van Ours (2014) provide a thorough review of the European literature. Røed (2012) presents a complementary review with case study detail about several Scandinavian changes which reinforce the evidence that low-​intensity monitoring and required program participation can have

7 1

UI After the Great Recession  71 large impacts on durations. Grubb (2011b) also reinforces that position, which presents case study details for Ireland and Spain. Specific analyses showing the expected impact and often large magnitudes of low-​intensity activation programs (e.g., a few hours per week of required activities or reported and verified job search) in European countries include Denmark (Geerdsen 2006; Graversen and Larsen 2013; Jensen et al. 2003; Svarer 2011), Germany (Boockmann et al. 2009), the Netherlands (Abbring et al. 2005; Gorter and Kalb 1996; van den Berg et al. 200; Lammers, Bloemen, and Hogchuertel 2013), Northern Ireland (McVicar 2010); Norway (Røed et al. 2008; Røed and Westlie 2011), Sweden (Røed et al. 2008), Switzerland (Lalive et al. 2005), and the United Kingdome (Dolton and O’Neill 1996; Petrongolo 2009). For Australia, Borland and Tseng (2007) provide evidence of large impacts of a strictly enforced requirement to report active job search. This is also the finding of Card, Kluve, and Weber’s (2018) meta-​analysis. (See Scarpetta, Chapter 2, this volume, for a similar argument.) For the United States, Black, Smith, Berger, and Noel (2003) exploit the details of who is required to participate in the low-​intensity activation program. They find that the program decreases UI benefit receipt by 2.2 weeks and that the impact does not appear to be from the services themselves but rather from notification of the requirement to receive the services. Similarly, and even more convincingly, for Denmark, Graverson and van Ours (2008, 2011) show that a program requiring the unemployed to attend a two-​week job search course, followed by regular one-​on-​one meetings with job counselors in which they received advice about job search and their job search efforts were monitored (and, for those who did not find a job, the possibility of being assigned to a training program) decreases average UI duration from 14 weeks to 11.5 weeks and raises earnings in the subsequent quarters (i.e., in net, more work). Like Black et al. (2003), they find that much of the impact on UI receipt (the only outcome they measure at a weekly frequency) occurred between being notified and the two-​week job search course. They also find that the impact rises with distance between home and the course location (and therefore, presumably, the time required to participate). Finally, they find no impact on post-​employment job quality (e.g., wages). They conclude: “the activation program mainly worked because it was compulsory and the unemployed did not like it. The activation program worked as a stick to job finding” (Graverson and van Ours 2011, 167). A recent set of experimental evaluations of the US REA program also supports this inference that activation programs lower UI duration, speed reemployment, and increase earnings.1 The REA program is low-​intensity—​usually a meeting of 1. Key papers include and Benus et al. (2008); Poe-​Yamagata et al. (2011); Michaelides and Muester (2018, 2020); Klerman et al. (2019); Manoli, Michaelides, and Patel (2018); Manoli and Patel (2019); and Michaelides et al. (2021).

72

72  Work and the Social Safety Net about an hour with a caseworker and sometimes a follow-​on reemployment activity of at most a few hours. Klerman et al. (2019) review this literature. Average impacts on earnings are about a week, employment at two quarters after initial UI claim increases about 2.5 percentage points, and earnings increase by about $500 over the year after the initial claim. Klerman et al. (2019) present the results of a four-​state multiarmed experimental study. The totality of the evidence suggests that much of the impact of the program comes from claimants not showing for the meeting and having their benefits suspended, some of the impact comes from assistance provided at and through the meeting, and little of the impact comes from enforcement of job search requirements or detection of noncompliance with those requirements. Unlike other studies, there is little evidence of claimants leaving before the meeting. In fact, the authors argue that the optimal strategy is to keep on claiming through the meeting. There is little penalty to not doing so, and suspension of benefits is neither sure nor substantial (i.e., many who do not attend the meeting continue to receive benefits at least for some time past the initial missed meeting).

UI POLICY AND UNEMPLOYMENT IN THE US GREAT RECESSION Interest in these issues is prompted by the macroeconomic situation in the United States, Europe, and the rest the world during the Great Recession and more recently during the pandemic recession. In 2008–​2009, the United States experienced a sharp recession, certainly the sharpest recession in a generation and perhaps since the Great Depression. Thereafter, the unemployment rate dropped to quite low levels, but the labor force participation rate remained quite low. Then, with the onset of the pandemic in early 2020, the unemployment rate jumped to even higher levels than during the Great Recession and labor force participation rate dropped sharply. In 2021, apparently in part with the arrival of vaccines, some of these changes have reversed. Nevertheless, through mid-​2021 unemployment rates remain quite elevated and labor force participation rates quite low relative to their pre-​pandemic levels. This increased unemployment during the Great Recession temporarily shifted the long-​term relative ranking of US and European unemployment rates. Historically, US unemployment rates have been well below comparable European rates. Using the OECD’s Harmonized Unemployment Rate, which uses common definitions (available through OECD 2012), the 2007 annual UI rate for the United States was 4.6% compared to 7.3% for the EU countries. However, the Great Recession started more severely in the United States than in Europe, such that, in 2009 and 2010, the US harmonized rate increased quickly

7 3

UI After the Great Recession  73 and briefly surpassed the EU. However, trends after 2011—​a steady and long US expansion and slower expansion in Europe—​returned US rates to well below EU rates (3.6% vs. 6.6% for December 2019). Furthermore, this discussion is for the conventional unemployment rate (i.e., only people who are jobless, looking for jobs, and available for work are considered unemployed). Broader definitions imply much higher rates. (See Bregger and Haugen 1995 for a careful discussion of these concepts. See Bureau of Labor Statistics 2011 for the raw data.) Thus, the American U-​6 rate (total unemployed, plus discouraged workers, plus all marginally attached workers, plus total employed part-​time for economic reasons, as a percent of the civilian labor force plus all marginally attached workers) for 2011 was 15.9% (Manoli, Michaelides, and Patel 2018), considerably above the conventional American Unemployment Rate (U-​3) of 8.9%. However, by December 2019, both measures had dropped sharply, U-​6 to 6.8% (by more than half) and U-​3 to 3.6% (by more than half). Similar patterns are present for the COVID recession. The conventional unemployment rate spiked from 3.5% in February 2020 to 14.8% in April 2020. Since then it has drifted down, reaching 5.2% in August 2021 (the most recent data as of this writing). As noted earlier, the United States responded to Great Recession and the associated sharp increase in unemployment by substantially increasing potential UI durations and making several smaller policy changes which had the effect of raising the effective replacement rate. These changes of making UI more generous are in contrast to the (already noted) changes in Europe in the 2000s which made UI less generous (Denmark, Germany, Italy, Netherlands, United Kingdom) (Grubb 2011b). Also as noted, similar changes were made in response to the pandemic recession: broader eligibility, longer eligibly, and much higher benefits. The extent to which the Great Recession UI extensions in the United States were themselves the cause of high and prolonged unemployment is the subject of considerable academic debate. Everyone agrees that the extensions have lengthened UI durations and the aggregate unemployment rate. The point of contention is: How large is the effect? Those arguing that the effect has been substantial—​ several percentage points—​ include Barro (2010) and Grubb (2011a, 2011b). Simulation studies suggesting an effect of about 1 percentage point include Fujita (2011) and Mazumder (2011). Recent and more careful cross-​ state microeconomic studies estimate a smaller impact (Valetta and Kuang 2010, 2012; Rothstein 2011; Farber and Valetta 2015; Farber et al. 2015). Representative of this literature, Farber and Valetta (2015) estimate that an additional month of UI eligibility increases average UI duration by 0.06 months—​ both in 2000–​2005 and 2007–​2012. Furthermore, they conclude that most of the effect on unemployment is on discouraging exits to out-​of-​the-​labor-​force, not on discouraging exits (i.e., returns) to employment. In net, they estimate

74

74  Work and the Social Safety Net that extended UI explained nearly a quarter of the increase in long-​term (more than six months) unemployment, but those spells were neither common nor a large share of the unemployed during a recession. As a result, they conclude that extended UI explained only about 0.4 percentage points of the increase in the unemployment rate. This appears to be, in part, because the behavioral response to UI generosity was smaller as the macroeconomy worsened; additionally, these papers used more data from the Great Recession. However, as Grubb (2011a) emphasizes (and as Farber and Valetta 2015 concede), it seems likely that these papers underestimate the total effect of UI policy on the level of the US unemployment rate. First, measurement error in the data and imperfect understanding of the timing of changes by the unemployed probably imply that these papers underestimate the intermediate-​term impact of longer UI benefits. Second, the difference-​in-​difference approach used by these recent papers only estimates the impact of lengthening durations on those currently unemployed. It thus misses the impact of more generous benefits (as noted above, slightly higher payments, tax deductibility, subsidy for health insurance) and increased entry into UI (as noted above, there is European evidence for such an impact; see Winter-​ Ebmer 2003; Lalive and Zweimüller 2004). Third, Grubb (2011a) (in part citing Holmes 2011) presents some evidence that the UI responses to the Great Recession weakened incentives to monitor active job search. In particular, funding for the staff to do monitoring has not risen fast enough to keep up with the sharp increase in the UI caseload, so the intensity of monitoring has almost certainly declined (see Scarpetta, Chapter 2, this volume, for a similar argument). Each of these factors suggests that while the recent literature carefully applies modern econometric methods, it likely underestimates the impact of UI policy on the US unemployment rate. In the other direction, Farber and Valetta (2015) present evidence that their difference-​in-​difference strategy does not fully control for the economy. They argue that this evidence implies a smaller true impact. Three other studies with better identification strategies appear to suggest larger impacts of the temporary increases in maximum UI duration and therefore a larger role for UI policy in causing the increase and then decrease in the unemployment rate. The first study is Johnston and Mas (2015). In 2011, Missouri cut its maximum duration of UI from twenty-​six weeks to ten weeks. Comparing job finding just before and after the implementation of the (almost totally unexpected and unpredictable) change, they estimate that a one-​month cut in eligibility cut UI duration by fifteen days, with slightly more than half of the change being an increase in employment. The changes were not clustered at the new, shorter exhaustion date; rather, rates of exit from UI increase at nearly all durations. These results for Missouri are very different from those in Farber, Rothstein, and Valetta (2015) and other papers using their simple cross-​time/​cross-​state identification strategy. While only applying to one state, this Missouri evidence is of higher quality than the other studies. The unexpected and unpredictable

7 5

UI After the Great Recession  75 change eliminates concerns about anticipatory behavior (for which they find some evidence), and the administrative data provide large samples free of the concerns about response error in the cross-​state studies. It should be possible to implement similar studies in the other states that cut maximum duration. The second study (Hagedorn et al. 2015) explores the impact of the (again unexpected) non-​renewal of EB and EUC, in December 2013, on employment. Previously the number of weeks of eligibility added varied across states from zero to forty-​seven; with the non-​renewal, in all states the eligibility fell to zero. They control for local economic conditions by comparing counties on either side of state borders. Again, they use aggregate administrative data (rather than survey data, which has smaller samples and issues with nonresponse bias and response errors). Finally, they use a method that accounts for general equilibrium effects (e.g., increased job creation). Using these data and methods, they estimate that the “failure” to renew EB and EUC explained between 50% and 80% of aggregate employment growth in 2014. The third study (Hagedorn et al. 2013) estimates the impact of the effects of UI benefit duration changes on employment (not merely the 2013 end of EB and EUC). That paper also uses aggregate administrative data and estimates impacts of similar magnitude. They conclude that “our estimates do not imply that the large increase in unemployment at the outset of the Great Recession was due to extensions of unemployment benefits. However, we do find that the extensions of benefits contributed significantly to the slow decline of unemployment thereafter. . . . Unemployment in 2011 would have been 2 percentage points lower had benefits not been extended” (with almost all of the difference being employment). These three sets of estimates are much larger than would have been expected given the results of Farber, Rothstein, and Valetta (2015) and other papers using their simple cross-​time/​cross-​state identification strategy. Which set of results is most accurate has major implications for optimal UI policy—​in general and in recessions. Similar debates are under way with respect to the pandemic policy reactions. Some observers have attributed the only slow decline in the unemployment rate to the high UI benefits. More careful analyses suggest that the changes have pushed up the unemployment rate but not by much (Coombs et al., 2022; Marinescu et al. 2021). Some states ended the higher benefits earlier. This should provide strong tests, but, as of this writing, it is too early to tell.

THREE POSSIBLE INTERPRETATIONS OF RECENT EXPERIENCE There are at least three possible interpretations of the Great Recession and the associated and prolonged sharp increase in unemployment.

76

76  Work and the Social Safety Net

Keynesian Some have interpreted the Great Recession and the associated unemployment as evidence of a lack of aggregate demand. There is wide agreement that the Great Recession was induced by a speculative bubble in residential housing and, to a lesser extent, in commercial construction and financial assets. This interpretation suggests a policy response involving fiscal stimulus. The American Recovery and Reinvestment Act of 2009 provided a large fiscal stimulus. The recession ended, but recovery was slow albeit steady. Unemployment dropped to very low levels, but earnings growth was anemic and labor force participation remained low through late 2019 and the start of the pandemic recession. Proponents of this position respond that the stimulus was not large enough and suggested a second, perhaps larger, round of stimulus funds.

Policy-​Induced Others have argued that adjustments to the UI program converted a moderate real business cycle event into a deep recession with prolonged unemployment (Barro 2010; Grubb 2011a, 2011b). These observers point to the major extensions to maximum UI duration and, to a lesser extent, to the other expansions of the benefit (including the exclusion of the benefit from federal income taxes and the subsidy to health insurance purchase). This interpretation suggests that the appropriate policy response would be to undo the expansions of the UI program. Those programs have been undone, and the unemployment rate has dropped sharply. Moreover, the econometric evidence reviewed here suggests that while this line of argument may have some explanatory power, changes in UI benefits were neither the main explanation of the sharp rise in the unemployment rate nor of the subsequent sharp fall of the unemployment rates to quite low levels.

Neo-​Classical Economics A final possible explanation is that the United States (and also Europe) has suffered a long-​run decline in competitiveness. Echoing Bruce Springsteen (“My Hometown,” Born in the USA, 1984), in the 2012 presidential debate, President Obama stated “There are some jobs that are not going to come back. Because they are low wage, low skill jobs” (presidential debate, October 16, 2012). To some extent, this long-​run decline was hidden by the speculative bubble of the mid-​2000s (and, in Europe, the positive effects of conversion to the Euro). That same speculative bubble would suggest that for many workers there is no status quo ante to which to return. The immediate pre-​recession level of output and employment was the result of an unsustainable housing boom. Through that housing boom and accelerating during the recession, American manufacturing employment has disappeared at a rapid pace (Charles et al. 2016). This interpretation suggested that a return to higher employment would require dropping, or at least stagnant, real earnings and restoring competitiveness (i.e., a drop in

7 7

UI After the Great Recession  77 wages, leading to a drop in domestic costs of production, leading to a drop in the price of domestic production, leading to a substitution of domestic production for some current imports and, finally, to increased exports). Recent stagnant or declining real wages are consistent with this interpretation. Facing lower wage offers, many have chosen to exit the labor market, leading to the continuing lower labor force participation rate (Solon 1979; Howell and Azizoglu 2011). The limited available evidence suggests that UI did not have a large depressive impact on return to work. It may, however, have eased the transition to retirement or disability. Those links seem worthy of further study. Recovery from the pandemic recession may suggest that different forces are now at work. Some have speculated that the pandemic recession has caused some workers to rethink their willingness to work for low wages. Some have also noted that the aging US population and decreased immigration are causing a labor shortage that will push up wages and the labor force participation rate. As I write this in mid-​2021, it is simply too soon to distinguish short-​term reactions as the pandemic recession recedes from longer-​term trends.

CONCLUSION This chapter has reviewed and contrasted the American and European experiences and evaluation literature with respect to unemployment and unemployment insurance policy. Three related points emerge regarding (1) the role of UI in the rise and fall of the unemployment rate during the great recession, (2) potential changes to the ongoing UI programs, and (3) the role of UI in the continued low level of the employment rate.

The Role of UI in the Rise and Fall of the Unemployment Rate During the Great Recession There is considerable debate about the role of UI in causing the spike in the unemployment rate and lowering employment during the Great Recession and its aftermath and then with the pandemic recession and its aftermath. Cross-​state studies consistently find only small impacts on unemployment and even smaller impacts on employment (i.e., in the absence of extended UI benefits, those individuals would not have found work; rather, they would have left the labor force). More recent state-​specific studies find evidence of large impacts on the unemployment rate and on employment. To the degree that unemployment is induced by a loss of competitiveness (i.e., a neo-​classical interpretation), it is plausible that the more generous UI benefits prolonged the recession and slowed the recovery by deferring necessary changes (as in Charles et al. 2016). The American UI system implicitly assumes stable or

78

78  Work and the Social Safety Net growing wages. For frictional unemployment, it is plausible that the new job will be similar to the old job. For cyclical unemployment, it is also plausible that—​ at least once the macro-​economy recovers—​the new job will be similar to the old job. When unemployment is due to structural changes and a loss of competitiveness, however, the new job will often pay considerably less than the old job. Careful studies from large-​scale industrial dislocations suggest an earnings loss of 30% or more (Jacobson et al. 1993; Couch and Placzek 2010). Furthermore, long unemployment durations—​in part facilitated by the long maximum UI durations—​further depress likely future earnings. Past work skills depreciate, new work skills are not developed, and employers infer a lack of interest in work. This was Ljungqvist and Sargent’s (1998) explanation of the rise in European unemployment in the 1980s. To the degree that long-​term labor market dislocations have induced a sharp downward shift in likely earnings, there might be reason to rethink the UI system so that it is encouraging, rather than retarding, painful but necessary labor market adjustments and that it is not easing the transition out of the labor market (i.e., to retirement or to disability insurance). In any recession, broadening the definition of suitable employment is likely to lead to more employment until the economy comes out of recession. The more the recession reflects not cyclical but secular changes, the more appropriate is broadening the definition of suitable wage. The earlier wage may not be available even once the economy improves. Better to move to addressing the secular changes rather than using the UI system to delay facing them. (See Scarpetta, Chapter 2, this volume, for a discussion of similar issues.) In as much as addressing those losses of long-​term earnings is a policy goal, policies outside the UI system are likely to be appropriate (e.g., wage insurance) (Klerzer and Litan 2001; Kletzer 2004; LaLonde 2007; Obama 2016).

Potential Changes to Ongoing UI Programs As discussed earlier, analyses of nonmonetary costs of program participation, recent empirical work in the United States and Europe on mandatory low-​intensity job search assistance programs, and reforms in the United States and Europe suggest several promising and complementary strategies to lower UI durations while still providing funds for consumption during unemployment: • Mandatory maintenance of a job search log with some level of employer job contacts and job offers received. Failure to keep the log and show the required level of job search would result in termination of the UI benefit. Here, some form of auditing of claims of active job search would probably be appropriate. As noted, the REA and RESEA programs include “eligibility

7 9

UI After the Great Recession  79 assessment” in their name. Nevertheless, the programs appear to include little additional enforcement of job search requirements. • Mandatory participation in low-​ intensity job search assistance, perhaps a several-​day job search assistance seminar followed by meetings of increasing frequency with a job search counselor. Here again, failure to participate in the program would result in termination of the benefit. Again, the change from REA to RESEA in the United States moved in the opposite direction (i.e., toward activities for a narrower group of the unemployed). • Broadened definition of a “suitable job” and a requirement to accept suitable job offers. Failure to accept such a job offer results in termination of the UI benefit. (See Scarpetta, Chapter 2, this volume, for a discussion of similar issues.) In the United States, these changes appear to require neither federal nor state statutory changes. Statutory authority exists. Limited available evidence suggests that states, in policy, and UI staff, in practice, are not imposing allowable requirements (e.g., a log, regular meetings with UI staff to verify search behavior, low-​intensity participation in job search assistance activities)—​not prior to the pandemic recession and certainly not during that recession. Furthermore, penalties for noncompliance are not swift, not sure, and not substantial. Imposing such requirements and penalizing those who do not comply would likely require more staff and a fundamental change in the perceived role of such requirements—​ from pure assistance to a balanced mix of assistance and pushing those who can find a job (or already have one which they are not reporting) off of UI. Since most UI durations are short, it is neither cost-​effective nor appropriate to impose these strategies on everyone. Instead, it might be more appropriate to impose these requirements for “longer” durations. Perhaps, upon receipt of UI, announce and schedule, for approximately two months later, a multiday job search assistance seminar followed immediately by a meeting with a counselor and a review of the job search log to date. Thereafter, perhaps one meeting at the end of month three, two meetings in month four, and weekly meetings in months five and six. Furthermore, states might widen the definition of what constitutes a suitable job with each passing month. “Profiling”—​the statistical modeling of those likely to have longer UI durations—​might help to further target some UI recipients for earlier and more intensive services. These strategies are consistent with the changes in Europe in the previous decade. Furthermore, these strategies would increase in intensity and apply earlier in the UI spells strategies included in the now-​sunsetted US Middle Class Tax Relief and Job Creation Act (February 2012) for EUC. In the United States and Europe, there is clear evidence that these strategies lead to shorter UI spells and lower total program cost (benefits plus services) and have little to no impact on job quality.

80

80  Work and the Social Safety Net These strategies would, however, require increasing state UI program staffing and making that staffing more counter-​c yclical. Currently, staffing changes relatively little with the unemployment rate such that the intensity of services and oversight drops sharply during recessions. Implementing a proposal of this form would benefit from an automatic requirement to increase staffing with the number of UI cases and a dedicated stream of funding to do so—​either dedicated federal funding or a requirement to draw down the UI trust fund to increase staffing. (See Scarpetta, Chapter 2, this volume, for a similar argument.) These broad directions of policy are consistent with the existing literature. Nevertheless, we know little about the optimal balance of program components and the optimal timing of their implementation. Especially if we are moving to a more active policy with more intensive casework interventions, the effects of these programs and their cost-​effectiveness are clearly evaluable using random assignment methods. Given the current state of knowledge, such formal impact evaluation and associated cost-​benefit analysis seem worthwhile. Such experimentation could be built directly into enabling legislation (see Klerman 2011).

The Role of UI in the Ongoing Low Employment Rate Post Great Recession, the US unemployment rate had fallen below pre-​recession levels, but the employment rate remained low (i.e., a lot of people are not working). Some, but not all, of the explanation appears to be demographics; an aging workforce is likely to have lower employment levels (Charles et al. 2016; Aaronson et al. 2015). However, demographics do not appear to explain all of the decline. The explanation in Europe—​where the maximum UI durations are much longer—​may be different; however, in the United States, given the short maximum UI durations of rarely over twenty-​six weeks and often well below twenty-​six weeks, the UI program is not a plausible explanation of the post-​ recession low employment rates. Thus, in the United States, the long-​term unemployed are not collecting UI. No change to the UI program will directly affect their choices. (A UI program that was more successful in placing those receiving UI on jobs would help, but we do not know how to run such a program, and impacts of any such program are likely to be small.) It seems plausible that other aspects of the US social safety net—​ Disability Insurance, Supplemental Security income, and perhaps the Supplemental Nutrition Assistance Program—​affect the employment-​to-​population ratio, but, in the United States, those programs are totally separate from UI (see the discussion in Charles et al. 2016).

8 1

UI After the Great Recession  81

REFERENCES Aaronson, D., Luojia Hu, Arian Seifoddini, and Daniel G. Sullivan. Changing Labor Force Composition and the Natural Rate of Unemployment. Chicago Fed Letter 2015: 338. Abbring, Japp H., Gerard J. van den Berg, and Jan C. van Ours. The Effect of Unemployment Insurance Sanctions on the Transition Rate from Unemployment to Employment. The Economic Journal 115, no. 505 (2005): 602–​630. Adamchick, Vera. The Effect of Unemployment Benefits on the Probability of Re-​Employment in Poland. Oxford Bulletin of Economics and Statistics 61, no. 1 (1999): 95–​108. Alba-​Ramirez, Alfanso, Jose M. Arranz-​Munoz, and Fernando Munoz-​Bulló. Exits from Unemployment: Recall or New Job. Labour Economics 14, no. 5 (2007): 788–​810. Andersen, Torben M., and Michael Svarer. 2009. “State Dependent Unemployment Benefits.” Journal of Risk and Insurance 78, no. 2 (2009): 325–​344. Arni, Patrick, Rafael Lalive, and Jan C. van Ours. How Effective are Unemployment Benefit Sanctions? Looking Beyond Unemployment Exit. IZA Discussion Paper No. 4509. Bonn, Germany: Institute for the Study of Labor (IZA), 2009. Baily, Martin N. Some Aspects of Optimal Unemployment Insurance. Journal of Public Economics 10, no. 3 (1978): 379–​402. Barro, Robert. The Folly of Subsidizing Unemployment. Wall Street Journal, August 30, 2010. http://​onl​ine.wsj.com/​arti​cle/​ SB1000142405274870395970 4575454431457720188.html. Barron, John M., and Wesley Mellow. Search Effort in the Labor Market. Journal of Human Resources 14, no. 3 (1979): 389–​404. Besley, Timothy, and Stephen Coate. Workfare versus Welfare: Incentive Arguments for Work Requirements in Poverty Alleviation Programs. American Economic Review 82, no. 1 (1992): 249–​261. Benus, J., E. Poe-​Yamagata, Y. Wang, and E. Blass. Reemployment and Eligibility Assessment (REA) study FY2005 initiative. Columbia, MD: IMPAQ International, 2008. Black, Dan A., Jeffery A. Smith, Mark C. Berger, and Brett J. Noel. Is the Threat of Reemployment Services More Effective Than the Services Themselves? Evidence from Random Assignment in the UI System. American Economic Review 93, no. 4 (2003): 1313–​1327. Bureau of Labor Statistics (BLS), U.S. Department of Labor. Labor Force Statistics (CPS): Table A.15: Alternative Measures of Labor Underutilization, 2011. http://​www.bls.gov/​weba​pps/​leg​acy/​cpsata​b15.htm. Bureau of Labor Statistics (BLS), U.S. Department of Labor. Employer Cost for Employee Compensation, 2019. data.bls.gov/​cgi-​bin/​surveymost?cm.

82

82  Work and the Social Safety Net Boockmann, Bernhard, Stephen L. Thomsen, and Thomas Walter. Intensifying the Use of Benefit Sanctions: An Effective Tool to Shorten Welfare Receipt and Speed Up Transitions to Employment? IZA Discussion Paper No. 4580. Bonn, Germany: Institute for the Study of Labor (IZA), 2009. Boone, Jan, Peter Fredriksson, Bertil Holmlund, and Jan C. van Ours. Optimal Unemployment Insurance with Monitoring and Sanctions. Economic Journal 117, no. 518 (2007): 399–​421. Borland, Jeff, and Yi-​Ping Tseng. Does a Minimum Job Search Requirement Reduce Time on Unemployment Payments? Evidence from the Jobseeker Diary in Australia. Industrial and Labor Relations Review 60, no. 3 (2007): 357–​378. Bregger, John E., and Stephen E. Haugen. BLS Introduces New Range of Alternative Unemployment Measures. Monthly Labor Review 118 (1995): 19–​26. Browning Martin, and Thomas F. Crossley. Unemployment Insurance Benefit Levels and Consumption Changes. Journal of Public Economics 80, no. 1 (2001): 1–​23. Caliendo, Marco, Konstantinos Tatsiramos, and Arne Uhlendor. Benefits Duration, Unemployment Duration and Job Match Quality: A Regression Discontinuity Approach. IZA Discussion Paper No. 4670. Bonn, Germany: Institute for the Study of Labor (IZA), 2009. Card, David, and Phillip B. Levine. Extended Benefits and the Duration of UI Spells: Evidence from the New Jersey Extended Benefit Program. Journal of Public Economics 78, no. 1 (2000): 107–​138. Card, David, Raj Chetty, and Andrea Weber. Cash-​on-​Hand and Competing Models of Intertemporal Behavior: New Evidence from the Labor Market. Quarterly Journal of Economics 122, no. 4 (2007a): 1511–​1560. Card, David, Raj Chetty, and Andrea Weber. The Spike at Benefit Exhaustion: Leaving the Unemployment System or Starting a New Job? American Economic Review 97, no. 2 (2007b): 113–​118. Card, David, Andrew Johnston, Pauline Leung, Alexandre Mas, and Zhuan Pei. The Effect of Unemployment Benefits on the Duration of Unemployment Insurance Receipt: New Evidence from a Regression Kink Design in Missouri, 2003–​2013. American Economic Review 105, no. 5 (2015): 126–​130. Card, D., J. Kluve, and A. Weber. What works? A meta analysis of recent active labor market program evaluations. Journal of the European Economic Association 16, no. 3 (2018): 894–​931. Carling, Kenneth, Bertil Holmlund, and Altin Vejsiu. Do Benefit Cuts Boost Job Findings? Swedish Evidence from the 1990s. Economic Journal 111, no. 474 (2001): 766–​790. Centeno, Mário, and Álvaro Novo. The Bright Side of the Moon: Unemployment Insurance Generosity and Post-​Unemployment Wages. Banco de Portugal Discussion Paper. Lisbon: Banco de Portugal, 2007. https://​edito​rial​expr​ess. com/​cgi-​bin/​con​fere​nce/​downl​oad.cgi?db_​n​ame=​pej2​007&paper​_​id=​64.

8 3

UI After the Great Recession  83 Center for Budget and Policy Priorities (CBPP). Pandemic Unemployment Insurance Provisions: What They Mean for Access to SNAP, Medicaid, and TANF, 2020. https://​www.cbpp.org/​resea​rch/​econ​omy/​pande​mic-​unemp​ loym​ent-​insura​nce-​pro​visi​ons-​what-​they-​mean-​for-​acc​ess-​to-​snap. Charles, Kerwin Kofi, Erik Hurst, and Matthew J. Notowidigdo. The Masking of the Decline in Manufacturing Employment by the Housing Bubble. Journal of Economic Perspectives 30, no. 2 (2016): 179–​200. Chetty, Raj. A General Formula for the Optimal Level of Social Insurance. Journal of Public Economics 90, no. 10 (2006): 1879–​1901. Chetty, Raj. Moral Hazard versus Liquidity and Optimal Unemployment Insurance. Journal of Political Economy 116, no. 2 (2008): 173–​234. Cockx, Bart, and Jean Ries. The Exhaustion of Unemployment Benefits in Belgium: Does it Enhance the Probability of Employment? IZA Discussion Paper No. 1177. Bonn, Germany: Institute for the Study of Labor (IZA), 2004. Congressional Research Service. Unemployment Insurance: Programs and Benefits. Washington, DC: Congressional Research Service, 2019. Coombs, K., Dube, A., Jahnke, C., Kluender, R., Naidu, S., and Stepner, M. Early Withdrawal of Pandemic Unemployment Insurance: Effects on Employment and Earnings. In AEA Papers and Proceedings (Vol. 112, pp. 85–​90). American Economic Association. Couch, Kenneth A., and Dana W. Placzek. Earnings Losses of Displaced Workers Revisited. American Economic Review 100, no. 1 (2010): 572–​589. Cullen, Julie, and Jonathan Gruber. Does Unemployment Insurance Crowd Out Spousal Labor Supply? Journal of Labor Economics 18, no. 3 (2000): 546–​572. Dolton, Peter, and Donal O’Neill. Unemployment Duration and the Restart Effect: Some Experimental Evidence. The Economic Journal 106, no. 435 (1996): 387–​400. Dormont, Brigitte, Denis Fougère, and Ana Prieto. L’effet de l’allocation Unique Dégressive sur la Reprise d’emploi. Economie et Statistique 343, no. 1 (2001): 3–​28. Engen, Eric, and Jonathan Gruber. Unemployment Insurance and Precautionary Savings. Journal of Monetary Economics 47, no. 3 (2001): 545–​579. Farber, Henry S., Jesse Rothstein, and Robert G. Valletta. The Effect of Extended Unemployment Insurance Benefits: Evidence from the 2012–​2013 Phase-​ Out. The American Economic Review 105, no. 5 (2015): 171–​176. Farber, Henry S., and Robert G. Valletta. Do Extended Unemployment Benefits Lengthen Unemployment Spells? Evidence from Recent Cycles in the US Labor Market. Journal of Human Resources 50, no. 4 (2015): 873–​909. Fitzenberger, Bernd, and Ralf A. Wilke. New Insights into Unemployment Duration and Post Unemployment Earnings in Germany. Oxford Bulletin of Economics and Statistics 72, no. 6 (2010): 794–​826. Fredriksson, Peter, and Bertil Holmlund. Optimal Unemployment Insurance in Search Equilibrium. Journal of Labor Economics 19, no. 2 (2001): 370–​399.

84

84  Work and the Social Safety Net Fredriksson, Peter, and Bertil Holmlund. Improving Incentives in Unemployment Insurance: A Review of Recent Research. Journal of Economic Surveys 20, no. 3 (2006a): 357–​386. Fredriksson, Peter, and Bertil Holmlund. Optimal Unemployment Insurance Design: Time Limits, Monitoring, or Workfare? International Tax and Public Finance 13, no. 5 (2006b): 565–​585. Fujita, Shigeru. Effects of Extended Unemployment Insurance Benefits: Evidence from the Monthly CPS. Federal Reserve Bank of Philadelphia Working Paper No. 10-​35/​R. Philadelphia, PA: Federal Reserve Bank of Philadelphia, 2011. Ganong, Peter, Pascal Noel, and Joseph Vavra. US unemployment insurance replacement rates during the pandemic. Journal of Public Economics 191 (2020): 104273. Geerdsen, Lars P. Is There a Threat Effect of Labour Market Programmes? A Study of ALMP in the Danish UI System. Economic Journal 116, no. 513 (2006): 738–​750. Gilbert, Neil, and Douglas Besharov. Welfare States Amid Economic Turmoil: Adjusting Work-​Oriented Policy. Policy & Politics 39, no. 3 (2011): 295–​308. Gorter, Cees, and Guyonne R. J. Kalb. Estimating the Effect of Counselling and Monitoring the Unemployed Using a Job Search Model. Journal of Human Resources 31, no. 3 (1996): 590–​610. Graversen, Brian K., and Jan C. van Ours. How to Help Unemployed Find Jobs Quickly: Experimental Evidence from a Mandatory Activation Program. Journal of Public Economics 92, nos. 10–​11 (2008): 2020–​2035. Graversen, Brian K., and Jan C. van Ours. An Activation Program as a Stick to Job Finding. Labour 25, no. 2 (2011): 167–​181. Graversen, Brian K., and Larsen, Brian. Is There a Threat Effect of Mandatory Activation Programmes for the Long-​ Term Unemployed? Empirical Economics 44, no. 2 (2013): 1031–​1051. Grubb, David. Assessing the Impact of Recent Unemployment Insurance Extensions in the United States. Paper presented at a workshop on Unemployment Insurance in the Recession, Institute for the Study of Labor (IZA), Bonn, Germany, June 27–​28, 2011a. http://​www.iza.org/​confe​renc​e_​fi​ les/​UnIRe2​011/​grub​b_​d6​938.pdf. Grubb, David. Unemployment Benefits and Activation as Influences on Labour Market Outcomes. Presentation at the Labour Activation in a Time of High Unemployment conference, Paris, November 14–​ 15, 2011b. http://​umdc​ ipe.org/​ conferences/​LaborActivationParis/​Papers/​David%20Grubb-​2011-​ UBactivation_​11_​1108.pdf. Gruber, Jonathan. The Consumption Smoothing Benefits of Unemployment Insurance. American Economic Review 87, no. 1 (1997): 192–​205. Gruber, Jonathan. Unemployment Insurance, Consumption Smoothing, and Private Insurance: Evidence from the PSID and CEX. Research in Employment Policy 1, no. 3 (1998): 3–​32.

8 5

UI After the Great Recession  85 Gruber, Jonathan. The Wealth of the Unemployed. Industrial and Labor Relations Review 55 (2001): 79–​94. Hagedorn, Marcus, Fatih Karahan, Iourii Manovskii, and Kurt Mitman. Unemployment Benefits and Unemployment in the Great Recession: The Role of Macro Effects. NBER Working Paper No. 19499. Cambridge, MA: National Bureau of Economic Research (NBER), 2013. Hagedorn, Marcus, Iourii Manovskii, and Kurt Mitman. The Impact of Unemployment Benefit Extensions on Employment: The 2014 Employment Miracle? NBER Working Paper No. 20884. Cambridge, MA: National Bureau of Economic Research (NBER), 2015. Hock, Heinrich, Walter Nicholson, Karen Needels, Joanne Lee, and Priyanka Anand. Additional Unemployment Compensation Benefits During the Great Recession: Recipients and Their Post-​Claim Outcomes. Washington: Mathematica Policy Research, 2016. Holmes, D. Improving Efforts to Help Unemployed Americans Find Jobs: Hearing Before the Subcommittee on Human Resources Committee on Ways and Means. 112th Cong. Statement of Douglas J. Holmes, President, UWC, 2011. http://​ www.waysa​ndme​ans.house.gov/​News/​ DocumentSingle.aspx?DocumentID=​ 228810. Holmlund, Bertil. Unemployment Insurance in Theory and Practice. Scandinavian Journal of Economics 100, no. 1 (1998): 113–​141. Howell, David R., and Bert M. Azizoglu. Unemployment Benefits and Work Incentives: The U.S. Labor Market in the Great Recession. Oxford Review of Economic Policy 27, no. 2 (2011): 221–​240. Jacobson, Louis S., Robert J. LaLonde, and Daniel G. Sullivan. Earnings Losses of Displaced Workers. The American Economic Review 83, no. 4 (1993): 685–​709. Jensen, Peter, Michael Rosholm, and Michael Svarer. The Response of Youth Unemployment to Benefits, Incentives, and Sanctions. European Journal of Political Economy 19, no. 2 (2003): 301–​316. Johnston, Andrew C., and Alexandre Mas. Potential Unemployment Insurance Duration and Labor Supply: The Individual and Market-​Level Response to a Benefit Cut. NBER Working Paper No. 22411. Cambridge, MA: National Bureau of Economic Research (NBER), 2015. Kajitani, Shinya. Does Skill-​ Development Make Elderly Japanese More Marketable? Japan Labor Review 5, no. 2 (2008): 89–​111. Katz, Lawrence. Long-​Term Unemployment in the Great Recession: Testimony Before the Joint Economic Committee, 111th Cong., April 29, 2010. Kiley, Michael T. How Should Unemployment Benefits Respond to the Business Cycle? B.E. Journal of Economic Analysis and Policy 3, no. 1 (2003): 1–​22. Klerman, Jacob. Performance Measurement Systems and Evaluation: Towards a Mutually Reinforcing Relationship. Paper presented at Improving the Quality of Public Services conference, Moscow, Russia, June 28–​29, 2011. http://​umdc​ ipe.org/​conf​eren​ces/​Mos​cow/​pap​ers/​CQI​_​v24​jak.docx.

86

86  Work and the Social Safety Net Klerman, J. A., C. Saunders, E. Dastrup, Z. Epstein, D. Walton, and T. Adam, with B. S. Barnow. Evaluation of Impacts of the Reemployment and Eligibility Assessment (REA) Program: Final Report. Prepared for the U.S. Department of Labor. Cambridge, MA: Abt Associates, 2019. Kletzer, Lori G. Trade‐Related Job Loss and Wage Insurance: A Synthetic Review. Review of International Economics 12, no. 5 (2004): 724–​748. Kletzer, Lori G., and Robert E. Litan. A Prescription to Relieve Worker Anxiety. Brookings Policy Brief Series No. 73. Washington: Brookings Institution, 2001. Kroft, Kory, and Matthew J. Notowidigdo. Should Unemployment Insurance Vary with the Unemployment Rate? Review of Economic Studies 83, no. 3 (2016): 1092–​1124. Krueger, Alan B., and Bruce Meyer. Labor Supply Effects of Social Insurance. In Handbook of Public Economics, vol. 4, ed. Alan J. Auerbach and Martin Feldstein. Amsterdam: North Holland, 2002, 2327–​2392. Krueger, Alan B., and Andreas Mueller. Job Search and Unemployment Insurance: New Evidence from Time Use Data. Journal of Public Economics 94, no. 3 (2010): 298–​307. Lalive, Rafael. Unemployment Benefits, Unemployment Durations, and Post-​ Employment Jobs: A Regression Discontinuity Approach. The American Economic Review 97, no. 2 (2007): 108–​112. Lalive, Rafael. How Do Extended Benefits Affect Unemployment Duration? A Regression Discontinuity Approach. Journal of Econometrics 142, no. 2 (2008): 785–​806. Lalive, Rafael, and Josef Zweimüller. Benefit Entitlement and the Labor Market: Evidence from a Large-​ Scale Policy Change. In Labor Market Institutions and Public Regulation, ed. Jonas Agell, Michael Keen, and Alfons J. Weichenrieder, 63–​100. Cambridge, MA: MIT Press, 2004, 63–​100. Lalive, Rafael, Jan C. van Ours, and Josef Zweimüller. The Effect of Benefit Sanctions on the Duration of Unemployment. Journal of the European Economic Association 3, no. 6 (2005): 1386–​1417. Lalive, Rafael, Jan C. van Ours, and Josef Zweimüller. How Changes in Financial Incentives Affect the Duration of Unemployment. Review of Economic Studies 73, no. 4 (2006): 1009–​1038. LaLonde, Robert John. The Case for Wage Insurance. Council Special Report No. 30. New York: Council on Foreign Relations, 2007. Lammers, Marloes, Hans Bloemen, and Stefan Hochguertel. Job Search Requirements for Older Unemployed: Transitions to Employment, Early Retirement and Disability Benefits. European Economic Review 58 (2013): 31–​57. Landais, Camille, Pascal Michaillat, and Emmanuel Saez. Optimal Unemployment Insurance Over the Business Cycle. NBER Working Paper No. 16526. Cambridge, MA: National Bureau of Economic Research (NBER), 2010.

8 7

UI After the Great Recession  87 Ljungqvist, Lars, and Thomas Sargent. The European Unemployment Dilemma. Journal of Political Economy 106, no. 3 (1998): 514–​550. Ljungqvist, Lars, and Thomas Sargent. Two Questions About European Unemployment. Econometrica 76, no. 1 (2008): 1–​29. Mastri, Annalisa, Wayne Vroman, Karen Needels, and Walter Nicholson. States’ Decisions to Adopt Unemployment Compensation Provisions of the American Recovery and Reinvestment Act. Princeton, NJ: Mathematica Policy Research, 2016. Manoli, D. S., Michaelides, M., and Patel, A. Long-​term effects of job-​search assistance: Experimental evidence using administrative tax data (No. w24422). National Bureau of Economic Research, 2018. Manoli, D., and A. Patel, May. Long-​Term Treatment Effects of Job Search Assistance and Training: A Summary of Recent Evidence. In AEA Papers and Proceedings (Vol. 109). 2019, 340–​343. Marinescu, I., D. Skandalis, and D. Zhao. The Impact of the Federal Pandemic Unemployment Compensation on Job Search and Vacancy Creation (No. w28567). National Bureau of Economic Research, 2021. Mazumder, Bhash. How Did Unemployment Insurance Extensions Affect the Unemployment Rate in 2008–​10? Chicago Fed Letter 285 (2011): 1–​4. McCall, John J. Economics of Information and Job Search. Quarterly Journal of Economics 84, no. 1 (1970): 113–​126. McVicar, Duncan. Does Job Search Monitoring Intensity Affect Unemployment? Evidence from Northern Ireland. Economica, 77, no. 306 (2010): 296–​313. Meyer, Bruce D. Unemployment Insurance and Unemployment Spells. Econometrica 58, no. 4 (1990): 757–​782. Meyer, Bruce D. Lessons from the U.S. Unemployment Insurance Experiments. Journal of Economic Literature 33, no. 1 (1995): 91–​131. Michaelides, M., and P. Mueser. The Labor Market Effects of U.S. Reemployment Programs During the Great Recession (Working Paper No. 08-​2015). Nicosia, Cyprus: University of Cyprus, Department of Economics, 2016. Michaelides, M., and P. Mueser. Are reemployment services effective? Experimental evidence from the Great Recession. Journal of Policy Analysis and Management 37, no. 3 (2018): 546–​570. Michaelides, M., and P. Mueser. The Labor Market Effects of US Reemployment Policy: Lessons from an Analysis of Four Programs during the Great Recession. Journal of Labor Economics 38, no. 4 (2020): 1099–​1140. Michaelides, M., P. R. Mueser, and J. A. Smith. Do Reemployment Programs for the Unemployed Work for Youth? Evidence from the Great Recession in the United States. Economic Inquiry 59, no. 1 (2021): 162–​185. Mitman, Kurt, and Stanislav Rabinovic. Pro-​Cyclical Unemployment Benefits? Optimal Policy in an Equilibrium Business Cycle Model. University of Pennsylvania Department of Economics PIER Working Paper No.

8

88  Work and the Social Safety Net 11-​023. Philadelphia: Penn Institute for Economic Research, Department of Economics (PIER), 2011. Moffitt, Robert. Unemployment Insurance and the Distribution of Unemployment Spells. Journal of Econometrics 28, no. 1 (1985): 85–​101. Mortensen, Dale. Job Search and Labor Market Analysis. In Handbook of Labor Economics, vol. 2, ed. Orley Ashenfelter and Richard Layard. Amsterdam: North-​Holland, 1986, 849–​920. Obama, Barack. Remarks of President Barack Obama—​State of the Union Address as Delivered, 2016. https://​www.whi​teho​use.gov/​the-​press-​off​i ce/​ 2016/​01/​12/​rema​rks-​presid​ent-​bar​ack-​obama-​%E2%80%93-​prepa​red-​deliv​ ery-​state-​union-​addr​ess. O’Leary, Christopher, and Randall W. Eberts. The Wagner-​Peyser Act and U.S. Employment Service: Seventy-​Five Years of Matching Job Seekers and Employers. W.E. Upjohn Institute Report No. 9. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research, 2008. O’Leary, Christopher, and Stephen A. Wandner. Do Job Search Rules and Reemployment Services Reduce Insured Unemployment? Upjohn Institute Working Paper No. 05-​ 112. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research, 2005. OECD (Organisation for Economic Co-​operation and Development). OECD Employment Outlook: Tackling the Jobs Crisis. Paris: OECD Publishing, 2009. OECD (Organisation for Economic Co-​operation and Development). OECD. Stat Extracts Accessed March 21, 2012. http://​stats.oecd.org/​index.aspx. Petrongolo, Barbara. The Long-​Term Effects of Job Search Requirements: Evidence from the UK JSA Reform. Journal of Public Economics 93, no. 11 (2009): 1234–​1253. Poe-​Yamagata, E., J. Benus, N. Bill, H. Carrington, M. Michaelides, and T. Shen. Impact of the Reemployment and Eligibility Assessment (REA) initiative. IMPAQ International, June 2011. https://​www.impaq​int.com/​sites/​defa​ult/​ files/​files/​ETAOP_​2012_​08_​I​mpac​t_​of​_​the​_​REA​_​Ini​tiat​ive.pdf Portugal, Pedro, and John T. Addison. Six Ways to Leave Unemployment. IZA Discussion Paper No. 954. Bonn: Institute for the Study of Labor (IZA), 2003. Røed, Knut, Peter Jensen, and Arna Thoursie. Unemployment Duration and Unemployment Insurance: A Comparative Analysis Based on Scandinavian Micro Data. Oxford Economic Papers 60, no. 2 (2008): 254–​274. Røed, Knut, and Lars Westlie. Unemployment Insurance in Welfare States: The Impacts of Soft Duration Constraints. Journal of the European Economic Association 10, no. 3 (2011): 518–​554. Røed, Knut, and Tao Zhang. Does Unemployment Compensation Affect Unemployment Duration? The Economic Journal 113, no. 484 (2003): 190–​206. Røed, Knut, and Tao Zhang. Unemployment Duration and Economic Incentives: A Quasi Random-​Assignment Approach. European Economic Review 49, no. 7 (2005): 1799–​1825.

8 9

UI After the Great Recession  89 Rothstein, Jesse. Unemployment Insurance and Job Search in the Great Recession. Brookings Papers on Economic Activity 43, no. 2 (2011): 143–​213. Sánchez, Juan M. Optimal State-​Contingent Unemployment Insurance. Economics Letters 98, no. 3 (2008): 348–​357. Schmieder, J. F., Von Wachter, T., and Bender, S. The long-​term effects of UI extensions on employment. American Economic Review 102, no. 3 (2012): 514–​519. Schmieder, Johannes F., Till von Wachter, and Stefan Bender. The Effect of Unemployment Benefits and Nonemployment Durations on Wages. American Economic Review 106, no. 3 (2016): 739–​777. Solon, Gary. Labor Supply Effects of Extended Unemployment Benefits. Journal of Human Resources 14, no. 2 (1979): 247–​255. St. Louis, Robert D., Paul L. Burgess, and Jerry L. Kingston. Reported vs. Actual Job Search by Unemployment Insurance Claimants. Journal of Human Resources 21, no. 1 (1986): 92–​117. Stone, Chad. Congress Should Heed President Biden’s Call for Fundamental UI Reform. Washington, DC: Center for Budget and Policy Priorities, 2021. Svarer, Michael. The Effect of Sanctions on Exit from Unemployment: Evidence from Denmark. Economica 78, no. 312 (2011): 751–​778. Tatsiramos, K., and Van Ours, J. C. Labor market effects of unemployment insurance design. Journal of Economic Surveys 28, no. 2 (2014): 284–​311. US Department of Labor. Unemployment Insurance Chartbook, 2010a. http://​ www.dol​eta.gov/​unemp​loy/​chartb​ook.cfm. US Department of Labor. Unemployment Insurance Extended Benefits, 2010b. http://​workfo​rces​ecur​ity.dol​eta.gov/​unemp​loy/​exten​ben.asp. US Department of Labor. UI Replacement Rates, 2021a. https://​oui.dol​eta.gov/​ unemp​loy/​rep​l_​ra​tio/​rep​l_​ra​tio_​rpt.asp. US Department of Labor. Unemployment Insurance Chartbook, 2021b. https://​ oui.dol​eta.gov/​unemp​loy/​chartb​ook.asp. US Executive Office of the President. Fact Sheet: The American Families Plan. Washington, DC: US Executive Office of the President, 2021. Uusitalo, Roope, and Jouko Verho. The Effect of Unemployment Benefits in Re-​ Employment Rates: Evidence from the Finnish Unemployment Insurance Reform. Labour Economics 17, no. 4 (2010): 643–​654. Valetta, Rob, and Katherine Kuang. Extended Unemployment and UI Benefits. FRBSF Economic Letters 12 (2010):1–​4. Valetta, Rob, and Katherine Kuang. Why Is Unemployment Duration So Long? FRBSF Economic Letters 3 (2012): 1–​5. Van den Berg, Gerard J., Bas van der Klaauw, and Jan C. van Ours. Punitive Sanctions and the Transition Rate from Welfare to Work. Journal of Labor Economics 22 no. 1 (2004): 211–​241. Van den Berg, Gerard J., and Johan Vikstrom. Monitoring Job Offer Decisions, Punishments, Exit to Work, and Job Quality. Swedish Ministry of Employment

90

90  Work and the Social Safety Net IFAU Working Paper No. 2009:18. Uppsala, Sweden: Institute for Labour Market Policy Evaluation (IFAU), 2009. http://​www.ifau.se/​upl​oad/​pdf/​se/​ 2009/​wp09-​18.pdf. Van Ours, Jan C., and Milan Vodopivec. How Changes in Benefits Entitlement Affect Job-​Finding: Lessons from the Slovenian Experiment. IZA Discussion Paper No. 1181. Bonn, Germany: Institute for the Study of Labor (IZA), 2004. Van Ours, Jan C., and Milan Vodopivec. How Shortening the Potential Duration of Unemployment Benefits Entitlement Affects the Duration of Unemployment: Evidence from a Natural Experiment. Journal of Labor Economics 24, no. 2 (2006): 351–​378. Van Ours, Jan C., and Milan Vodopivec. Does Reducing Unemployment Insurance Generosity Reduce Job Match Quality? Journal of Public Economics 92, no. 3 (2008): 684–​695. Wandner, Stephen A. Solving the Reemployment Puzzle: From Research to Policy. Kalamazoo, MI: W. E. Upjohn Institute for Employment Research, 2010. Winter-​Ebmer, Rudolf. Benefit Duration and Unemployment Entry: Quasi-​ Experimental Evidence for Austria. European Economic Review 47, no. 2 (2003): 259–​273.

 91

5 ACTIVATION IN EIGHT EUROPEAN SOCIAL ASSISTANCE PROGRAMS Ivar Lødemel and Amílcar Moreira

O

ver the course of two decades, activation became a key feature of European minimum income systems. By the beginning of the Great Recession in 2008, activation policies had not only spread to all EU nations, but had also undergone a gradual change that increasingly impacted the nature of European safety nets. In most of the nations covered in this chapter, this change was from an emphasis on developing the skills of the participants, often described as a human capital approach, to an activation strategy where the emphasis was on incentives, motivation, and stronger obligations to work, often described as a work-​first approach. Given the experience of European countries during the Great Recession and its aftermath, as well as the recession that resulted from the coronavirus pandemic, the issue of the activation of minimum income benefit recipients is likely to be much discussed in future years. Activation—​the policy of designing benefit rules and employment/​training services with the view of moving unemployed income benefit recipients into work—​is not to be confused with workfare. In Europe, different programs for the workless are widespread and feature both similarities and important differences. Defined as an ideal type, workfare refers to “programmes or schemes that require people to work in return for social assistance benefits” (Lødemel and Trickey 2001, 7). Like workfare, activation is compulsory: if work or training requirements are not met, benefits can be reduced or even withdrawn. Unlike workfare, where individuals are required to become self-​sufficient through unsubsidized paid employment or to participate in work-​ for-​benefit schemes (see Lødemel and Trickey 2001, 6–​7), activation considers a broader variety of options for returning income benefit recipients back to work, Ivar Lødemel and Amílcar Moreira, Activation in Eight European Social Assistance Programs In: Work and the Social Safety Net. Edited by: Douglas J. Besharov and Douglas M. Call, Oxford University Press. © Oxford University Press 2023. DOI: 10.1093/​oso/​9780190241599.003.0005

92

92  Work and the Social Safety Net including training, education, and the provision of job-​search assistance services. This is also not to be confused with active labor market policies (ALMPs) that, despite being part of the policy mix used to activate minimum income recipients, do not usually place eligibility conditions on the right to income protection and have a broader aim of dealing with more structural imbalances in the labor market, such as structural unemployment and regional development. (Hvinden 1999, 28–​29; Kildal 2001, 5). While activation is most commonly associated with minimum income, some authors extend the coverage to also include unemployment assistance and disability programs for both the insured and uninsured workless, such as unemployment insurance (UI) and disability benefits (as described in other chapters in this volume; see also Eichhorst et al. 2008a, 7–​8; Lødemel and Trickey 2001, 9–​11). We should also differentiate between activation policy and activation programs. In some countries, such as the United Kingdom or the Netherlands, policymakers opt for creating activation programs that offer a specific set of activation options and (sometimes) apply a specific set of rules and sanctions. However, activation, as a policy, covers interventions in other areas of public policy, such as the structuring of the tax system to create financial incentives to take up work or the organization of the delivery of employment and training services to jobless income recipients. This chapter, however, only describes activation for minimum income programs. By “minimum income schemes,” we mean schemes that provide a financial safety net for individuals/​households whose income is below the national social minimum. (Such schemes are also often referred to as social assistance or, in the United States, as cash welfare or simply “welfare.”) Because of their subsidiary nature, minimum income schemes are only available to persons/​households who—​ with some exceptions, such as family benefits—​are no longer eligible for other forms of income protection. Unlike social insurance, which provides insurance-​ based protection against (work-​related) social risks such as unemployment, old age or illness, minimum income schemes are a noncontributory form of protection (Lødemel and Schulte 1992, 8–​9). Unlike categorical social assistance, minimum income schemes are (quasi)universal; that is, they are not targeted at particular groups or social risks (Eardley et al. 1996, 28). The focus of this chapter is reforms in the governance of activation taking place in the two decades leading up to the 2008 financial crisis and the responses in the immediate aftermath of the Great Recession. In the first two sections, we report on the results of a project we conducted on activation reforms in minimum income programs of eight European countries (the Czech Republic, Denmark, France, Germany, the Netherlands, Norway, Portugal, and the United Kingdom).1 1. The project was funded by Oslo and Akershus University College (now OsloMet University) and by REASSESS (Nordic Centre of Excellence: Reassessing the Nordic Welfare Model).

9 3

European Social Assistance Programs  93 In doing so, we identify two waves of activation reforms. The first dates back to the late 1990s and is marked by the introduction of activation requirements as a condition to an entitlement to minimum income protection and by the introduction of activation programs for minimum income recipients. The second wave took place (more or less) in the first decade of the twenty-​first century and involves a number of measures aimed at both strengthening the focus on work in the activation of minimum income recipients and at improving the delivery of services. Making use of the heuristic potential of two-​step cluster analysis, we look at the map of the governance of activation in Europe that, to a certain degree, results from the second wave of activation reforms. The third part of the chapter highlights some of the changes that took place after the onset of the financial crisis and in the “austerity” period that followed. In the final discussion, we take a longer time perspective in order to reflect on the changing strategies pursued in national activation policies and programs.

THE FIRST TWO WAVES OF ACTIVATION REFORMS IN EUROPE Looking at developments in the activation of minimum income recipients in eight European nations (the Czech Republic, Denmark, France, Germany, the Netherlands, Norway, Portugal, and the United Kingdom), we can identify two reasonably clear waves of reform that occurred prior to the Great Recession. The first wave, which coincides with the last decade of the twentieth century, is marked by the introduction of activation requirements as an eligibility condition to minimum income benefits. This was the case with the 1998 Act on Active Social Policy in Denmark, the introduction of the Guaranteed Minimum Income scheme in Portugal in 1996, and the introduction of activation programs, as was the case of the 1998 introduction of the New Deal for Young People and the New Deal for Long-​Term Unemployed in the United Kingdom (see Eardley et al. 1996; Lødemel and Trickey 2001; Hanesh and Baltzer 2001; Moreira 2008). This first wave of reforms sets the institutional background for a second generation of activation reforms. Not only because these reforms try to deal with the implementation problems that emerged during the first set of activation measures (see Van Berkel 2009), but also because it sets the institutional rationale that shapes the set of institutional reforms that follow. Hence, it is not surprising that this second wave of reforms, rather than weakening, was actually about the strengthening of the importance of work as an obligation for minimum income recipients and about the improvement of the delivery of services—​in line with New Public Management thinking. In the paragraphs below, we describe in more detail these second-​wave developments.

94

94  Work and the Social Safety Net

STRENGTHENING THE IMPORTANCE OF WORK As noted, underlying this second set of activation reforms is a clear desire to strengthen the importance of work as an obligation for minimum income recipients. This is visible in attempts to structure the framework of minimum income protection by reference to individuals’ ability and availability to work. A good example of this was the reform introduced in Germany, in 2005, where social assistance recipients deemed able to work at least three hours per day were moved to the new unemployment assistance scheme—​Arbeitslosengeld II (ALG-​ II) (Eichhorst, Grienberger-​ Zingerle, and Konle-​ Seidl 2008, 30–​ 34; Mosley 2007, 1–​3). This is also visible in reforms that redefine the content of the activation requirement that delimits the right to minimum income protection. This was the case in France with the introduction of the Revenu Solidarité Active (RSA), in 2009, which replaced the Revenue Minimum d’Insertion (RMI). In the RMI, to be eligible for minimum income benefits, recipients were required to engage in a set of activities that would help them in their social insertion. In this sense, the obligation of looking for work was on par with other activities/​obligations, such as improving one’s education or health situation. The new scheme established a clear hierarchy between professional and social forms of insertion, where the latter form was reserved for a minority of beneficiaries who were considered not job ready and must be seen as a first step toward professional insertion (Clegg and Palier 2014). Another example of this increased focus on work can be seen in attempts at tightening the type of jobs recipients can refuse without losing the right to minimum income protection. This was the case in France, where the RSA required that job offers be compatible with the qualifications and professional competencies of job seekers. However, after three months of receiving benefits, there were an increasing set of restrictions on the type of jobs recipients can refuse. Thus, within the first three months on RSA, recipients can refuse jobs that pay less than their previous jobs. In the following three months, however, recipients can only refuse offers that pay less than 95% of their previous ones. After six months, they can refuse an offer that pays less than 85% of their previous jobs. After twelve months, recipients cannot refuse jobs that pay more than the benefit they receive (Moreira and Lødemel 2014, 321). (This is similar to activation requirements in many countries’ unemployment assistance programs.) A third way of strengthening the importance of work consisted of the introduction or strengthening of financial incentives. Reflecting on the second-​wave reforms introduced during the first decade of the twenty-​first century, we can identify two different approaches to the use of financial incentives.

9 5

European Social Assistance Programs  95 1.  Financial incentives that are used to increase the financial gain of making the transition to work. This is done through in-​work benefits, such as income disregards, integration allowances, or tax credits. A good example of this was the introduction of the Working Tax Credit in the United Kingdom, in 2003, which replaced the Working Families’ Tax Credit (Griggs et al. 2014). Another good example here was the 2002 introduction of Mini-​and Midi-​Jobs in Germany (discussed below), which replaced the previous mechanism of social insurance contribution reductions for low-​wage jobs (Fertig and Kluve 2004, 24). 2.  Benefit cuts that are used as a way of reducing the attractiveness of being on benefits. The best example of this was the introduction of the Danish 2002 More in Work Act, which introduced a ceiling on benefits after six months on social assistance. Moreover, in a move to de facto target non-​Western ethnic minority families, the Act introduced a series of additional benefit reductions. Married couples receiving social assistance for more than six months had their benefits reduced. In addition, if one of the spouses (normally the woman) was considered not available to work, social assistance was withdrawn and replaced by a homemaker supplement (Goul-​ Andersen and Pedersen 2007, 16). Another visible sign of this increased focus on work, albeit less evident, can be seen in the decreasing importance given to training and education in the activation of minimum income recipients. A good example of this was the 2009 introduction of the Flexible New Deal (FND) in the United Kingdom, which replaced the New Deal for Young People (NDYP), ND25+​, and Employment Zones. Unlike the NDYP or ND25+​, under which (among other offers) participants were entitled to a training and education offer, with the FND, the offer of training or education was made subject to the decision of local Job Centre Plus offices (Griggs et al. 2014). The exception here was Germany, where the introduction of ALGII and the consequent strengthening of the role of the Bundesagentur für Arbeit (BfA, or the public employment service) in the activation of minimum income recipients signified a strengthening of the importance given to education and training. Thus, whereas previously help toward work did not (typically) include a training/​education offer, minimum income recipients were able to take part in short-​term (typically four-​to eight-​week) courses that included qualification and counseling and job search support components. In addition, they were entitled to a training voucher that they could use to participate in vocational training courses (Moreira and Lødemel, 2014; Weishaupt, Jørgensen, and Nunn, Chapter 9, this volume). Curiously, this increased focus on work was not, as some might expect, accompanied by a universal increase in the use of sanctions for noncompliance. Three

96

96  Work and the Social Safety Net countries, however, either increased sanction length or made sanctions easier to apply. In the Czech Republic, the period of benefit suspension for minimum income recipients who refuse a (suitable) job offer was increased (in 2004) from three to six months (Sirovatka 2014). In France, temporary partial suspensions of minimum income benefits were introduced for individuals who refuse to sign who or discharge the responsibilities inscribed in their insertion contract (Dujol and Grass 2009, 306). In the Netherlands, the option—​consecrated in the 2004 Work and Social Assistance Act (WWB)—​was to give local authorities full power of discretion in the application of sanctions to beneficiaries (Moreira and Lødemel, 2014).

RESTRUCTURING THE DELIVERY OF SERVICES As mentioned earlier, besides a strengthening of the focus on work, this second wave of activation reforms focused on improving the delivery of activation services. As discussed below, these reforms were fundamentally shaped by what scholars have deemed New Public Management (NPM) and how this has developed over the years. Originally, NPM revolved around • A focus on issues of efficiency and outcome accountability, rather than on issues of universalism or equity; • The separation between policymaking, funding, and monitoring and the delivery of service; • A preference for the devolution of responsibilities and management control to lower levels of the policy system (i.e., decentralization); • A preference for specialized, “single-​purpose” organizations in the delivery of services; • A preference for the privatization of certain public services, or outsourcing the delivery of services to private providers; • A preference for the use of market-​based mechanisms—​such as financial incentives, contractual relations, and quasi-​markets in the public sector—​ for securing higher levels of effectiveness in the delivery of services; and • An increased focus on measuring/​monitoring performance (Pollitt 2002, 474; Boston 2010, 20–​21). As a product of the (perceived) failure of NPM in delivering on the efficiency gains it promised and the (vertical and horizontal) coordination problems it generated (Christensen and Lægreid 2010, 22–​31), a new wave of public-​sector reforms emerged during the 1990s, which some labeled “post-​NPM” (Christensen and Lægreid 2011, 395). Rather than replacing the full range of NPM ideas, this new set of reforms seemed to be mostly targeted at addressing the vertical and horizontal coordination problems that emerged with the first generation of NPM

9 7

European Social Assistance Programs  97 policies. Thus, in order to improve vertical coordination, there was a move toward the introduction of strategic units, reviews, public service agreements, and the enhancement of central control mechanisms—​such as stronger audit systems, tightening financial management, and strengthening accountability. In order to improve horizontal coordination, there was a move toward the integration of services—​going from the creation of one-​stop shops to the merger of government agencies—​and the creation of issue-​based structures such as task forces, intergovernmental councils, government czars, etc. (Christensen and Lægreid 2011, 394–​395). Consequently, reforms in the delivery of activation services to minimum income recipients reflected the development of NPM thinking. Therefore, we can identify a number of reforms that were in line with the principles of old-​ style NPM. A good example was the outsourcing of the delivery of employment services to private providers. Perhaps the most extreme move in this domain was the 2002 Structure Implementation Work and Income Act (SUWI) in the Netherlands. Following on the footsteps of the 1990 Act on Public Employment Services (PES), which ended the monopoly of the PES in the provision of employment services, the SUWI Act went further and mandated that all employment services be subcontracted with for-​and not-​for-​profit organizations (Sol et al., 2008, 189–​190). Given its disappointing results, however, the mandatory outsourcing of employment services was later reversed (van Berkel and de Graaf 2011, 139). The influence of old-​style NPM thinking can be seen in the strengthening of customer focus in the delivery of activation services. Examples include the introduction of personal action plans as a mechanism for defining the set of services offered to Qualification Programme participants in Norway (Gubrium et al. 2014) or to social assistance recipients in the Czech Republic (Sirovatka 2014). More resource-​intensive forms of personalized delivery, such as use of case managers or personal advisers, were less common. Even where case managers have been designated, as in Norway, they often lacked the resources—​or ability—​to deliver what was needed. One possible exception here is France, where RSA recipients were eligible to receive assistance from a personal advisor through the local PES (Monthly Personalized Support, or SMP) (Béraud and Eydoux 2009, 6; Moreira and Lødemel 2014). In addition, we can identify a set of reforms shaped by post-​NPM thinking. In some cases, namely in schemes where the administration of minimum income benefits was devolved to local authorities, reforms were aimed at improving vertical coordination. The most common practice involved the use of financial mechanisms that were meant to strengthen the ability of the central government to steer the delivery of services on the ground. For instance, in Denmark, the central government reimburses 65% of the costs to local authorities with social assistance recipients in activation, while it only reimburses 35% of the expenses related to those not on activation (Kvist and Harsløf 2014).

98

98  Work and the Social Safety Net In other cases, reforms were aimed at improving the coordination of service delivery on the ground—​in other words, aimed at improving horizontal coordination. In the Netherlands, this was done though the introduction, in 2002, of one-​stop shop “Locations for Work and Income” (LWI) where the various providers of employment services to the unemployed—​Centres for Work and Income (CWI), Employee Insurance Agency (UVW), and local authority social services—​were co-​located (Tergeist and Grubb 2006, 17). (CWIs provided services to unemployed persons who were expected to find work easily. UVWs provided services for unemployed persons with strong employability deficits and who were entitled to social insurance benefits. Local authority social services provided services to hard-​to-​employ unemployed social assistance recipients; see Tergeist and Grubb 2006, 17.) In Germany, the option was to create new bodies jointly managed between local national authorities and the PES (Eichhorst et al. 2008, 46–​48). While functioning under a unified leadership, there was a significant division of responsibilities between the BfA and local authorities within the Arbeitsgemeinschaft (ARGE) consortia. The BfA was responsible for the payment of Unemployment Benefits-​II (UB-​II) and for the delivery of activation services to recipients. Local authorities were responsible for the reimbursement of housing and childcare costs, as well as for the provision of social counseling services (Eichhorst et al. 2008, 46–​48).

THE SECOND WAVE OF REFORM The range of reforms that we have highlighted in the previous section had a significant impact on the policy landscape of activation in Europe (for a comprehensive mapping of different approaches to activation, see Lødemel and Trickey 2001). To capture the new geography of activation in Europe as it emerged in the first decade of the twenty-​first century, we looked at the governance of the activation of minimum income recipients in the eight countries mentioned previously. Following Carmel and Papadopoulos (2003, 94), we see governance as “the attempt to ‘steer’ the behavior of individuals, groups or institutions toward particular social and politico-​economic goals via a set of institutions and processes that aim to maintain or change the status quo.” These eight countries represent the different models of welfare protection in Europe (see Ferrera, Hemericjck, and Rhodes 2000; Fenger 2007) and different forms of minimum income provision (see Immervoll 2009). Departing from Fenger’s (2007) typology of welfare regimes, we included three countries (France, Germany, and the Netherlands) from the “conservative-​ corporativist” type. As can be seen in Table 5.1, while in France and the Netherlands the right to a minimum income is secured through general social assistance schemes (RSA and WWB, respectively), in Germany (since 2005) individuals able to work more than three hours per week are covered by ALG-​II—​an

9 9

European Social Assistance Programs  99 unemployment assistance scheme. To a certain degree, the option adopted in Germany mirrors that of the United Kingdom, which can be seen as representative of a more liberal model of welfare provision, where the (Income-​Related) Jobseekers Allowance (JSA), the UK’s unemployment assistance scheme until 2013, acted as the main financial safety net for work-​able individuals. In addition to these, we included two representatives of the social-​democratic model of welfare provision—​Denmark and Norway. We included the Czech Republic as a representative of the post-​communist European type. Finally, we included the Rendimento Social de Inserção (RSI) in Portugal, a social assistance scheme, as a representative for the southern-​European rim.2 To capture different approaches to activation, we look at three sets of indicators (see Appendix B for a full description of the indicators). We start by looking at the type of minimum income benefit scheme in the country under analysis and the distribution of responsibilities in the administration of minimum income benefits between the central and local governments. Then, in line with Carmel and Papadopoulos’s (2003) distinction between formal and operational governance, we look at the formal dimension of the governance of activation—​the rules and regulations that set the framework for the activation of minimum income recipients (Carmel and Papadopoulos 2003, 94). Here, we begin by looking at the main aim of activation as defined by the nature of the activation requirement that is imposed on minimum income recipients (Guibentif and Bouget 1997; Lødemel and Trickey 2001). We look specifically at the importance given to work in the set of obligations that condition the eligibility to minimum income protection and the type of work that recipients are expected to accept as a condition for the receipt of minimum income benefit. Acknowledging that the definition of what constitutes a suitable job covers a variety of issues, we focus on two criteria that can be seen as helping to secure more sustainable professional integration trajectories: whether job offers allow recipients to continue their previous occupations and whether job offers secure a level of pay that is at least equal to the salary of workers’ previous jobs. We then look at the harshness of the sanctions that are imposed on minimum income recipients for failing to take an activation or job offer. This is measured by looking at whether sanctions involve the termination of eligibility to minimum income benefits (repressive sanctions) and, when this is not the case, by looking at the maximum cut in minimum income benefit in the monthly payment subsequent to the first refusal of a (suitable) activation or job offer (restitutive sanctions).

2. Fenger (2007, 24–​25) identifies three different welfare regimes in Eastern Europe: the former USSR type, the developing welfare state type, and the post-​communist European type. It should be noted that Fenger (2007, 22), unlike us, does not include Portugal as one of the southern-​rim nations.

10

100  Work and the Social Safety Net Table 5.1  Cross-​tabulating welfare-​state and minimum income typologiesa,b Fenger 2007 +​Ferrera, Hemerijck and Rhodes 2000

Immervoll 2009

Social Democratic

Kontanthjælp, Denmark (2010) Qualification Programme, Norway (2009)

Liberal

Social assistance

Conservative-​ Corporativist

WWB, Netherlands (2009) RSA, France (2009)

Post-​Communist

Assistance in Material Need, Czech Republic (2009)

Southern European

RSI, Portugal (2008)

Unemployment assistance

JSA/​Flexible New Deal, UK (2010) ALG-​II, Germany (2008)

Notes: a When selecting the timeframe for our comparison, we decided to look at the activation of minimum income recipients in 2008 (i.e., before the impact of the 2008 financial crisis and subsequent foreign debt crisis). In those cases, such as the UK, France, the Netherlands, Norway or the Czech Republic, where policy developments were not directly influenced by the crisis, we extend our timeframe to cover the impact of these reforms. b Our selection of cases also acknowledges the fact that in some instances minimum income recipients are (sometimes mandatorily) enrolled in activation programs that offer a specific set of activation options and apply a distinctly different set of rules (namely on sanctions). Thus, in Norway, instead of looking at what goes on with social assistance, we look at the activation of participants in the Qualification Programme (QP), which was aimed at improving the services to long-​term, hard-​to-​employ social assistance recipients. In the same way, in the UK, we look at the activation of JSA recipients in the context of the Flexible New Deal, which was introduced in 2010.

To capture the different approaches in the provision of employment and training services to minimum income recipients, we look at the importance given to training and education as a means to assist recipients in entering or returning to the labor market. The task of mapping the range of options offered to minimum income recipients is a particularly difficult task. This is due to several changes that we have observed during this second wave of reform (Moreira and Lødemel 2014). First, policymakers have strengthened the power of discretion of local authorities in the provision of this type of service, thus making it more difficult to capture a general approach. Second, there was a move away from one-​size-​fits-​all approaches toward a tailoring of offers to the needs of recipients, which are more difficult to classify. Finally, (increasingly) the offer a recipient gets depended on the duration of the benefit episode or of the unemployment episode. The option adopted in this study, although far from comprehensive, provides key insight into policymakers views of the purpose of activation. These options are measured by whether there is a legal guarantee of the provision of training and education and, if so, when this option becomes available.

10 1

European Social Assistance Programs  101 Finally, we look at the use of financial incentives as a means to motivate minimum income recipients to enter/​return to the labor market. Rather than looking at the type of tools used by policymakers, the objective here is to capture whether they introduced the use of in-​work benefits with the intention of increasing the amount of disposable income that a recipient can earn from paid work (“positive incentive” approach) or whether they tried to reduce/​devalue minimum income benefits with the objective of making work more appealing (“negative incentive” approach). Such reductions can go from an outright cut in benefits levels, to the introduction of lower benefits after a given period on benefits or for recipients who are not available for work, to the elimination/​undermining of the mechanisms used to adjust the level of benefits to changes in the cost of living. This is a particularly complex area as the financial incentive to take up work is the product of the interaction of policies that are targeted specifically at minimum income recipients, such as income disregards or integration allowances, and of (broader) policies that regulate the functioning of the labor market (and the economy), such as minimum wage legislation or rates of taxation for earned income (see Immervoll and Pearson 2009). Bearing in mind that a comprehensive study of the various types of tools is impossible in this chapter, we limit our analysis to those tools that are specifically targeted at minimum income recipients, such as rules determining the level and adjustment of minimum income benefits and the use of in-​work benefits (income disregards, tax credits, integration allowances, or reductions in employee social insurance contributions). Having looked at the formal dimensions of the governance of activation, we then look at the operational dimensions—​the organizational arrangements and procedures for policy delivery (Carmel and Papadopoulos 2003, 94). As Streeck and Thelen (2005, 109) rightly mention, there is a gap between the pattern of the rule and the real pattern of life under it. While we acknowledged this, our focus is on the formal rules and structures that govern the activation of minimum income recipients, as they are the object (and product) of the processes of deliberation that shape policymaking in this area. We look at three indicators: first, the level of integration between the administration of minimum income benefits and the provision of activation services (see Bradshaw 2000; Bergmark and Minas 2010). Second, we look at the involvement of private providers in the delivery of activation services, notably job-​brokering services and the organization of training courses. We differentiate between “state-​centered” models of service provision, where private providers were not involved, and schemes where private providers were involved through market-​based mechanisms (such as tendering or voucher schemes) or through partnerships (Considine 2001, 23–​24; Newman 2007, 368–​371). Finally, we look at the degree to which policymakers structured the delivery of services to better adjust them to the needs of minimum income recipients, which is measured by whether the organizations responsible for the provision of activation services were required to offer a personal action plan and a case-​ manager/​personal adviser (“single contact points”) to minimum income recipients.

102

102  Work and the Social Safety Net Table 5.2  Models of activation in Europe (two-​step cluster analysis) Aims

MI scheme

Model of benefit administration

Involvement of Country/​ies private providers

Suitable work

UA

Centralizsed

Market-​Oriented

DE-​2008; UK-​2010

Work

SA

Decentralized -​Strong central steering

Market-​Oriented

DK-​2010; NTH-​2009

Social insertion

SA

Centralized

Network-​Oriented

PT-​2008

Suitable work

SA

-​

Market-​Oriented

FR-​2009; CZ-​2009

Strengthening employability

TARGETED

Decentralized -​Strong central steering

Market-​Oriented

NO-​2009

As can be seen in Table 5.2, the institutional map that results from this second wave of reforms is rather diverse, with five different models within the eight countries covered.3 When compared with previous attempts at mapping the activation of social assistance recipients (see Lødemel and Trickey 2001; Moreira 2008), we observe that, despite the introduction of the variables covering the operational dimension of activation, how the right to a minimum income is structured remains a critical factor in shaping the different models of activation. On the other hand, the role of sanctions in differentiating different models of activation becomes much less evident (see Figure 5.1). Looking at the cases in our sample, we can identify an initial set of clusters where the model of governance is fundamentally shaped by how the right to a minimum income was guaranteed. This is the case of the United Kingdom and Germany, where the right to a minimum income (for work-​able persons) was provided through an unemployment assistance scheme, not through general social assistance. This shapes both the objectives of activation and how this is implemented. Thus, the focus was on helping recipients back into the labor market, rather than dealing with other social vulnerabilities. However, as recipients were not seen as having the depth of employability problems as traditional social assistance recipients, they were given more freedom to choose the type of job they want. For example, in the United Kingdom, during the first thirteen weeks on benefits JSA, recipients could restrict their job search to jobs that secure usual pay in a usual occupation (Department for Work and Pensions 2015).

3. We tried different cluster solutions. However, none of them satisfied the criteria of quality offered by the Statistical Package for the Social Sciences (SPSS).

10 3

European Social Assistance Programs  103 Predictor Importance Aims Model of Benefit Administration Involvement of Private Providers Financial incentive approach Importance given to training and education Integration of benefit and activation services Sactions Individual focus in the provision of services 0,00

0,20

0,40

Least Important

0,60

0,80

1,00

Most Important

Figure 5.1  Importance of variables in determining clustering.

In addition, the model of governance in the United Kingdom and Germany can be characterized by three key traits. The first trait was the use of positive financial incentives to strengthen the appeal of returning to the labor market. However, the two countries vary significantly in how this strategy is implemented. In the United Kingdom, this was done through the Working Tax Credit scheme, by which JSA recipients who move into low-​paid jobs were entitled to a top-​up through the tax system. In Germany, on the other hand, this was done through exemptions/​reductions in personal social insurance contributions. Thus, under the Mini-​Jobs program, jobs with wages up to €400 per month were exempted from income tax and social security contributions for the employee. Under the Midi-​Jobs program, jobs with wages between €401 and €800 received a partial reduction of social security contributions for employees, which decrease proportionally until reaching the full rate of contribution (approximately 21%) (Fertig and Kluve 2004, 24). The second trait is the use of private actors, on a market-​based model, in the delivery of activation services. While this has been a relatively common approach in Germany (see Weishaupt, Jørgensen, and Nunn, Chapter 9, this volume), in the United Kingdom, this reflected a move—​one that can be traced back to the introduction of Employment Zones in 2000—​of increasing the role of private providers in the delivery of employment services and of using market competition to improve the quality of services offered to job seekers. This trend reached its pinnacle with the introduction of the Flexible New Deal, by which

104

104  Work and the Social Safety Net employment services were delivered through a mixture of public, private, and third-​sector providers (Griggs et al. 2014). The third trait concerns the fact that both countries have moved toward the integration of benefit and activation models. However, reflecting the different paths of institutional development, the formulas of service integration are quite distinct. In the United Kingdom, where benefit administration and the provision of employment services were allocated to central government agencies (the Benefits Agency and the Employment Service, respectively), the option was to merge them into one single state agency: Job-​Centre Plus (Griggs et al. 2014). In Germany, reflecting the previous split between a decentralized model of benefit administration (where local authorities play a central role) and a centralized model of employment service provision (allocated to the Federal Employment Agency, BfA), the option was to create new bodies jointly managed between local national authorities and the BfA—​the ARGE consortia (Eichhorst, Grienberger-​ Zingerle, and Konle-​Seidl 2008, 46–​48). Similar to the United Kingdom and Germany, the governance of activation in Norway reflects the specific role the Qualification Programme (QP) plays in that country’s social protection structure. Introduced in 2007, the QP was aimed at improving the job prospects of long-​term social assistance recipients, identified as having a significantly lowered ability to work. In this sense, QP was intended to provide a more intensive level of support than that provided to other social assistance recipients. Critical to the success of this new approach was the introduction of the Welfare and Employment Agency (NAV), in 2006. The NAV brought together state employment and national insurance services and municipal social assistance services in an effort to improve the range of activation offers available to QP participants. Prior to 2006, recipients of municipal social assistance did not have access to the wide range of activation programs offered for the insured workless through the national employment agency. However, the objective of improving the employability of QP participants was marred by the fact that the QP did not effectively guarantee the provision of a training/​education opportunity, instead leaving it to local NAV offices to determine individual eligibility—​as was also the case before the introduction of QP. Additionally, in line with what was common practice in the activation of social assistance recipients, the provision of both job-​brokering and training services was increasingly subcontracted to private providers. Whereas in the previous cases the model of governance is influenced by how the right to a minimum income is provided, the following two models are more influenced by a set of ideological standpoints on how to deal with those at the bottom of the income ladder—​who are traditionally the core target of social assistance schemes. Underlying the model of governance in Denmark and the Netherlands is the idea that the market is the best mechanism for the allocation of social resources. This can be seen in the nature of the activation requirement that regulates the eligibility to minimum income benefits, where recipients were

10 5

European Social Assistance Programs  105 required to return to the labor market as soon as possible, with no exceptions as to the types of jobs recipients can refuse. In line with this ideology, there was no effective guarantee that recipients were offered an opportunity to improve their qualifications through training or education. The influence of market-​oriented ideas, namely of old-​style NPM, is also visible in the fact that private providers (selected through market-​based mechanisms, such as public tenders) were given a significant role in the provision of training and job-​brokering services. Additionally, the influence of market-​oriented ideas is also visible in the use of financial mechanisms by the central government as a means to strengthen local authorities’ focus on activation and on getting minimum income recipients back to the labor market. As mentioned above, the Danish government only reimbursed 35% of the costs of social assistance recipients not in activation, while it reimbursed 65% of the expenses related to those on activation (Kvist and Harsløf 2014). In the Netherlands, this financial incentive involved dividing central government funding to local authorities in two yearly budgets: one for paying benefits and one for reintegration activities. To strengthen the focus on activation, municipalities that spend more on benefit payments than their budget allows will have to cover the cost of such expenses within their own municipal budget. Moreover, money from the reintegration budget that was not spent was to be paid back to the national government (Spies and van de Vrie 2014). In contrast to the influence of market-​based ideas in the Netherlands or Denmark, in Portugal, the activation of social assistance recipients reflected a more sociological understanding of societal functioning, where individuals were seen as part of a broader multidimensional network of social relations. In this sense, activation was not about returning individuals to the labor market, but about engaging them in a process that promotes their (re-​) insertion into this set of relations. Consequently, the obligation of looking for/​accepting a job was “on par” with other social obligations, such as dealing with health or addiction problems, improving one’s housing situation, or ensuring that one’s children attend school. This supportive approach justified the great attention given to education and training and the absence of negative financial incentives as a way of promoting the return to the labor market. This more sociological understanding of the functioning of society was also at the base of the model of delivery adopted. Thus, the scheme was implemented through a semi-​decentralized structure based on Local Support Commissions (Comissões Locais de Acompanhamento [CLAs]) intended to facilitate the coordination of the network of actors (government departments, local authorities, and civil society organizations [CSOs]) involved in the provision of services to RSI recipients (Capucha, 1998, 32–​36). Nevertheless, it is important to point out that, despite the intent of decentralizing the implementation of the scheme, the ministry of Social Security played a key role in the implementation of the measure. First, Social Security regional

106

106  Work and the Social Safety Net services were responsible for the determination of entitlement to minimum income. Second, the coordinator of the CLA was usually the representative from the Social Security regional office. Third, Social Security provided the necessary resources (funding, space, social workers, and administrative support) for the implementation of the scheme (see Law [Lei] 19-​A/​96, Art. 15, No. 1, and Art. 16, No. 4). In contrast with the more or less integrated models just described, the model of governance in France and the Czech Republic is more difficult to typify. On the one hand, the nature of the activation requirement—​based on the idea that minimum income recipients should try to find suitable work—​puts it closer to the model of activation adopted in Germany or the United Kingdom. Thus, in the Czech Republic, recipients were allowed to refuse jobs that were not in line with their occupation (see Sirovátka 2014). As described above, in France, besides being able to refuse a job that is not in line with their previous occupation, during the first three months on RSA, recipients could refuse a job that pays less than their previous job. In the following three months, recipients could only refuse an offer that pays less than 95% of their previous job. After six months, recipients could refuse an offer that pays less than 85% of their previous job. After twelve months on benefits, recipients could not refuse a job that pays more than the benefit they received. On the other hand, the decentralized model of benefit administration puts these countries closer to the Dutch or Danish model of activation. (In France, decision on eligibility for RSA and the power to decide the enforcement of sanctions, was allocated to the president of the Conseil Général [local authority]; however, the benefit was administered on behalf of the local authorities by the Caisse d’Allocations Familiales [Gineste and Brunhes 2012].) There are, however, significant differences with regards to the ability of central government in steering developments on the ground. This mostly relates to how the central government repaid local authorities for their costs incurred during the administration of minimum income benefits. In the Czech Republic, the central government compensated municipalities for the totality of costs associated with social assistance, no strings attached (Sirovatka 2014). In France, however, in what can be seen as an attempt to control the growth of expenditure on RSA benefit entitlements, the state delayed the payment of these costs for a period of two to three years. To summarize, we have so far been able to identify five clusters of governance that emerged around 2008, which we have described as the end of the second wave of activation reforms in Europe. In the first cluster—​Germany and the United Kingdom—​the governance of activation was largely shaped by the way minimum income is provided. More than elsewhere, the focus in these two nations was to facilitate transitions to work, often by the use of positive incentives, extensive use of private actors in ALMPs, and a move toward greater integration of the system of benefits and of activation. Norway, the second cluster, shares similarities with the first cluster, where a new program introduced at the end of

10 7

European Social Assistance Programs  107 the period placed greater emphasis on work and was implemented as part of a large-​scale integration of benefit and activation agencies, thereby broadening the range of ALMP options available. A third cluster, consisting of Denmark and the Netherlands, is characterized by an ideologically driven turn toward a reliance on market mechanisms to best motivate both the workless and the agencies in charge of implementing activation. In contrast, activation in Portugal, the fourth cluster, was found to be driven more by a sociological understanding in which, at the time, activation was focused here more on processes of (re-​)insertion to a set of social relations than mechanisms to ensure a quick integration into ordinary work. A fifth cluster, consisting of France and the Czech Republic, was harder to typify. The two nations shared an emphasis on decentralization of both benefits and ALMPs combined with a central financial responsibility for benefits. In the decade leading up to the Great Recession the activation policies and programs that had emerged in the previous decade underwent changes in both strategy and in how programs were governed. The emphasis on a conditional activation more in line with the European Social Model and different from the perceived tough workfare policies in the United States, had changed into a picture more similar to that prevailing in the United States. As a result, European nations faced the onset of the Great Recession and the following austerity period with a safety net that had more holes and offered less protection compared to the minimum income schemes found in the 1990s. The following two sections will address how the safety net responded to, first, the increasing needs created by the Great Recession and its aftermath of austerity, and second, the combined health and economic crisis resulting from the ongoing COVID-​19 pandemic.

DEVELOPMENTS IN THE AFTERMATH OF THE GREAT RECESSION An economic crisis followed by a recession and austerity, such as that experienced by most of the nations covered, may have differing impacts on national welfare arrangements. One inevitable outcome is an increase in need. As unemployment increases, minimum income programs were affected in two main ways. Young people—​now the largest group among the unemployed in Europe—​resorted to social assistance because they were typically unable to establish entitlement to insurance‐based unemployment benefits. For people who had a work history and contribution records, an increase in their long‐term unemployment means that many reached the end of their entitlement period and thereafter needed to resort to social assistance. The resulting increase in the number of those in need was compounded by the financial crisis turning into a broader sovereign budget crisis in several of the nations. Policymakers were faced with two main options for reform:

108

108  Work and the Social Safety Net   1. Strengthening the safety net (entitlement and benefit levels) and/​or increasing the investment in human resources for those currently in need of minimum income benefits.   2.  Reducing spending on minimum income and activation programs through reductions in the generosity of benefit, tighter eligibility criteria, and a harsher sanctioning regime combined with fewer resources devoted to activation programs. By way of examples rather than systematic comparison, we highlight some of the key activation reforms made in the eight nations that were the subject of our earlier analysis. To assess the nature of the changes that can be observed, we look at changes in the investment in activation programs, benefit levels in national minimum income, and the extent of individualization and personalization of services. In the Czech Republic, the current minimum income scheme (Assistance in Material Need) was first enacted in 2006. The law, introduced by a social democratic government, represented the first shift from protection to conditional activation (Sirovatka 2014). In 2007, a center‐right government introduced further curtailments on eligibility (e.g., canceling the automatic annual revaluation of benefits). Following another reform in 2009, minimum income recipients were automatically transferred to a lower level Existence Minimum after receiving minimum income for six months. The reforms, combined with delayed revaluations after 2009, have resulted in a benefit level of less than half of what was in place in 2000 (Sirovatka 2014, 274). (In 2015, the benefit amounted to only one-​third of the official poverty line in the Czech Republic [Natili 2020].) While this was a long-​standing element in Czech minimum income, the use of in-​kind benefits became more prevalent after 2008, first for recipients accused of misuse of funds and, from 2012, in the form of a Social Card, applied to large segments of those entitled to minimum income. While both positive and negative incentives were in place from the start in 2006, the use of sanctions has moved in a punitive direction over the years. In the first six years from its inception, minimum income in the Czech Republic evolved into an idealized workfare model combined with sharply reduced benefits. While the Czech Republic was not among the nations hardest hit by the Great Recession, ideological change, first within social democracy and later with the change to a center‐right government, appears to have been the main driver toward stricter minimum income—​ and activation regimes (Sirovatka 2014). National discussions focused on benefit abuse (often with references to the Roma population) and the need to shift the balance from rights to more obligations contributed more to the movement for reform than did the impact of the harsher economic climate following the onset of the 2008 crisis (Sirovatka 2014). The evidence suggests that the reforms were hastily planned and poorly implemented. In 2013, this resulted in sharp criticism from the Ombudsman, and, in 2013, both the Social Card and Public Works were dismantled. Minimum income in the Czech Republic therefore embraced

10 9

European Social Assistance Programs  109 the more punitive aspects of activation and returned to a less generous version of the safety-​net approach of the 1990s (Sirovatka 2016). In Denmark, the human resource development approach, described as “workfare with welfare” (Torfing 1999), was gradually changed in the 2000–​2009 period into something resembling a work-​first approach (Kvist and Harsløf 2014). The financial crisis and its aftermath resulted in an increase in both minimum income recipients and numbers enrolled in activation programs. In 2011, a center‐left government took power after a long period of liberal-​conservative control. The government inherited a plan from its predecessor whereby the insured unemployed had their entitlement period halved to two years. First postponed, hoping for an improved economy, this reform was instead implemented in 2013. As a result, a large number of people lost their entitlement and added to the increase in minimum income (Kvist and Harsløf 2014, 68). Activation in minimum income turned to conditional education for young recipients. For those ages thirty and over, activation meant tougher work requirements (Kvist and Harsløf 2014, 68). With the exception of a greater emphasis on education for young people, little changed in Danish activation. In 2016, a new conservative coalition government came into power and has implemented more drastic changes. A new benefit ceiling resulted in lowered benefits for many minimum income recipients, as well as reduced benefits for able-​bodied recipients who fail to complete at least 225 hours of community work annually. Additionally, the emphasis moved away from long-​term and repeated participation in activation programs to more intensive programs in which employers are expected to take a greater role in placements (Danish Ministry of Employment 2014). In France, the minimum income system (RSA) was planned before the crisis and introduced in 2009 by the conservative Sarkozy government. Compared to the earlier RMI, RSA represented a stronger emphasis on obligations and a higher benefit for those combining benefits with low‐paid work (Clegg and Palier 2014). Activation was also increasingly focused on work-​first strategies, and, since 2009, only those deemed incapable of work are served with the preexisting social activation program. The French economy contracted sharply at the beginning of the crisis, and the recovery was slow. Increases in unemployment affected youth, in particular. Because the RSA excluded people under the age of twenty-​five, a new “Youth RSA” for the eighteen-​to twenty-​five-​year-​olds was introduced in 2010. As the target group of the new program was mainly qualified youth with work experience, the focus of activation shifted away from the original target group of RSA. In addition to being crowded out of activation, traditional minimum income recipients, now in greater numbers due to the crisis, were also faced with a less tailored and individualized service as individual counselors were increasingly replaced by self-​service using the internet (Clegg and Palier 2014). In Germany, the current minimum income system was introduced by a social‐democratic government as part of the Hartz IV legislation in 2005. While eligibility criteria became stricter, the merger with Unemployment Assistance resulted in better resourced and more individualized activation programs after

10

110  Work and the Social Safety Net 2005 (Clasen and Goerne 2014). Unlike other European nations, Germany experienced a continued decrease in unemployment in 2008–​2010. The growing austerity in Europe also facilitated, however, a growing support for fiscal consolidation of the state in Germany. For minimum income recipients, this resulted in a curtailment of benefit levels for some, and the volume of activation services was reduced in real terms. Both this reduction and the shortage of staff to provide services has disproportionally affected those most distant from the labor market (Clasen and Goerne 2014, 187). Together with Germany, Norway was one of the European nations least affected by the Great Recession. As part of a decade‐long drive to end poverty (2000–​2010), center‐right and red‐green governments have moved activation in Norway away from the workfare‐like programs in place during the first wave of activation reforms (Gubrium et al. 2014). As part of a major governance reform, the Qualification Program was introduced by a red-​green government in 2007. In 2011, the financing of the program was changed. While it was first introduced as a national program with state funding to ensure universal implementation, this was changed to block grant funding that year. Numbers have declined steadily after its implementation, and the strong emphasis on tailored services was lacking in most of the local authorities of Norway. The election in 2013 brought in a conservative-​right minority government. In 2015, the government secured majority support for a new universal workfare program that replaced the workfare condition that has been a part of the minimum income legislation since 1993. While the preexisting clause stated that local authorities can demand work in exchange for benefits, the new legislation changed this to shall. By way of emphasizing this shift toward universal conditionality, a new clause was added to the law, stating that recipients of minimum income could be sanctioned with a reduced benefit if the new condition was not met. While less resourced and less individualized activation is associated with the harsher economic climate in the other nations, developments in Norway were an expression of a political desire to reduce the rights and strengthen the obligations of minimum income recipients. In a legal analysis of the new universal activation program, Kane and Kohler-​Olsen (2015) concluded that this curtailment was motivated by moralism. The Netherlands entered the recession with an extensive set of activation programs and services in place. While this was a continuation of a position already established in the 1990s, the strategy of activation took a turn in 2004. New legislation changed the content and governance of activation drastically. The turn was toward decentralization and stricter sanctioning, but also an increased emphasis on individual tailoring of interventions (Spies and van de Vrie 2014). The Netherlands was among the nations most strongly affected by the economic crisis. In the period from late 2009 to summer 2010, the number receiving social assistance (WWB/​the Investment in Sustainable Work for Young People Act [WIJ]) increased sharply (Statistics Netherlands 2010). The initial response of

11 1

European Social Assistance Programs  111 the government was to extend the coverage/​placements of activation. In 2010, a right‐wing government came into power and quickly introduced drastic reforms. The budget for activation was cut by more than half, affecting both the insured and recipients of minimum income. The already harsh sanction regime was further strengthened with the sharpening of obligations linked to social assistance (Spies and van de Vrie 2014). The Netherlands was already one of the activation regimes that changed the most from the first to the second wave of reforms, and these changes suggest that it moved further in the direction of a hardened climate for the workless poor. Portugal introduced its first national safety net scheme, the RMG (Guaranteed Minimum Income), in 1996. The scheme included an activation component from the beginning. This component was later strengthened, first with the reform in 2003, when a center‐right government changed the name from RMG to RSI (Social Insertion Income) and introduced stricter entitlement criteria in addition to strengthened work obligations and sanctions (Moreira, Carolo and Nicola 2014). Of the nations covered here, Portugal was hit the hardest by the financial crisis. An initial response to the crisis came in 2009, when the then socialist government introduced two reforms. First, in the form of added protection, the entitlement period for unemployment benefits was extended and a tax incentive was provided for employers hiring unemployed people. Second, facing increasing pressure from the European Union due to the debt problem, entitlements to minimum income were reduced, obligations increased, and sanctions were strengthened (Moreira and Carolo 2014). Portugal was one of the nations forced to accept a crisis package from the International Monetary Fund, European Union, and the European Central Bank as a necessary step in managing its large national debt. In the 2011 election, the socialist government was replaced by a center‐right coalition, which introduced new cost savings in the form of tightened eligibility criteria and lower benefit levels. Activation increasingly took the form of an obligation to engage in “socially useful work.” While Portugal was one of the nations that initially developed a wider focus on social integration as part of the minimum income scheme, events suggest that the nation was moving quickly in the direction of providing a last-​resort safety net (Moreira and Carolo 2014). With the reforms of minimum income and the introduction of national programs for activation in the late 1990s, Britain—​under the leadership of New Labour—​quickly went from being an activation‐laggard to one of the leaders in the first wave of European activation reforms (Lødemel and Trickey 2001). Making work pay—​in the form of tax credits—​initiated Britain’s second wave of reform. The reform process in this period culminated with the introduction of the Flexible New Deal in 2009. While this reform extended the emphasis on delivering personal and tailored services, it also culminated in stronger work‐related conditions and sanctioning tied to the receipt of aid (Griggs et al. 2014). In 2010, the conservative‐liberal government quickly announced that it wished to do away with what it considered to be an overabundance of activation programs

12

112  Work and the Social Safety Net (Griggs et al. 2014, 94). In 2012, the existing activation programs under the Employment and Support Allowance and the Job Seekers Allowance were merged into a single integrated package called the Work Programme. Framed in a language of widespread welfare dependency and distrust of minimum income recipients’ willingness to work, participation in the Work Programme was mandatory for the long-​term unemployed. Minimum income recipients who were not yet required to participate in the Work Programme but who were found to not be looking for work were required to participate in a four-​week work experience called “Mandatory Work Activity.” Failure to participate in the Work Programme or Mandatory Work Activity led to a reduction or termination of benefits (Daguerre and Etherington 2014; Department for Work and Pensions 2011; Her Majesty’s Treasury 2010). In summary, the crisis did not bring out signs of visible added protection for the populations on minimum income, the lowest safety net. On the contrary, benefits were often reduced in a situation when more people needed support. Activation programs were not extended, but were instead contracted as a result of budget cuts. For those who were still offered participation in programs, the positive trend toward more individualized and tailored programs, seen in the years leading up to the crisis, was reversed during the first five years of the Great Recession.

DISCUSSION Activation as a policy form arrived late in Europe. In the mid-​1990s, six European nations (all included here) had enacted provisions for compulsory work‐related activities for recipients of social assistance. The term “activation” was borrowed from medical rehabilitation and used in Europe to distinguish conditional work or training programs from those known as “workfare” in the United States. In the 1990s, “workfare” was used by European policymakers as a foil to describe what the new European programs were not (Lødemel and Trickey 2001, 4). The first systematic comparison of European and US programs suggested that there was some credence to the European claim that, although conditional, the European programs were indeed less corrective and more focused on giving recipients “more” in the form of new opportunities in their striving for self-​ reliance. At the turn of the twenty-​first century, programs in the Netherlands, Denmark, and the UK stood as leaders of this Human Resource Development strategy (Trickey 2001, 279). The second wave of reform, which in most cases took place during 2000–​2008, saw a similar development in the nations covered here. In regards to the strategy pursued in national programs, the main trend was in the direction of a greater focus on work, often described as a work‐first approach and distinguished from the human resource approach found most clearly expressed in the UK, Denmark, and the Netherlands in the late 1990s. Interestingly, two nations which at that

11 3

European Social Assistance Programs  113 time featured stronger work‐first approaches, Norway and Germany, moved in the opposite direction. In both nations, this took place in the context of a partial closing of preexisting social divisions of activation through the merger of services for the minimum income recipients and other workless groups. Work‐first approaches are often associated with extensive sanctioning. Surprisingly, we find that most of the nations did not emphasize the use of sanctioning in the reforms during this second wave. The clearest exceptions here were Portugal and the Czech Republic. These relative “newcomers” to European activation shared the overall trend toward work‐first and, at the same time, differed from the other six nations in an increased use of sanctions during this period. The second important trend observed in the 2000–​2008 period was a restructuring of activation services. This took different forms. We distinguished between old‐style NPM reforms, most clearly found in the increasing use of private providers (the Netherlands) and an increased customer focus (Norway and the Czech Republic). With a possible exception of the French case, the form of “personalization,” however, took the form of formal procedures more than services delivered. Examples of “post‐NPM” reforms were the creation of merged “one‐stop” services for wider groups of unemployed, involving personal social services and employment and social security, as well as combining state and local services providers (the Netherlands, Germany, and Norway). The cluster analysis further demonstrated that the second wave of reforms created a new landscape of activation among the European nations covered here. As a result of the financial crisis and the ensuing austerity in many of the nations covered here, we asked if the response was a return to more protection in the form of improved benefits and access to programs. What we found was the opposite: less protection and less access to resourced programs. Although the timing of reforms is different in each of the nations covered here, they appear to follow a similar trajectory (Lødemel and Gubrium 2014), where they begin with a resourced activation approach in the first national programs, only to move gradually in the direction of corrective workfare over time. This is also true for the US programs (Lødemel and Gubrium 2014, 331). Twenty years ago, this was not yet to be seen or understood, as European programs developed a decade later than their US counterparts. If the idea of trajectories can stand the test of the coming decade, Europe is likely to follow in the path of the United States. This path involves a step away from the income maintenance of the modern welfare state (cash benefits) and a return to the old form of providing benefits in kind (food stamps). In the United States, this form of curtailment has already seen workfare added to it (work for food) (Besharov 2015). As recent research also finds examples of workfare required in US shelters (Gubrium and Gubrium 2017), we are perhaps witnessing a return to the principles guiding the Poor Laws of the nineteenth century. As salaries are reduced for low-​wage workers in Europe and minimum income benefits are also reduced, thereby remaining below the level of the lowest wage, it is perhaps not only the workhouse test, but less eligibility—​the accompanying principle of the UK 1834 Poor Law—​that is in store for the future uninsured workless.

SA

SA

PT-​2008

NO-​2009

UA

Suitable work

Harsh Low Restitutive

Harsh Low Restitutive

Low

Social insertion Strong High Restitutive

Strong High Restitutive

Soft Low Restitutive

Soft High Restitutive

Integrated

Positive incentive

No incentive

Positive incentive

Positive incentive

Positive incentive

Integrated

Co-​located

Integrated

Split

Split

Comprehensive Integrated

Positive incentive

Market-​ oriented

Market-​ oriented

Market-​ oriented

Network-​ oriented

Market-​ oriented

Market-​ oriented

Market-​ oriented

Market-​ oriented

Strong

Strong

Procedural

Procedural

Strong

Procedural

Poor

Procedural

Integration Involvement Individual focus of benefit of private providers in the and provision activation of services services

Comprehensive Split

Importance Financial incentive given to training and approach education

Strong High Restitutive

Sanction regime

Decentralized—​ Strengthening Repressive Weak Central employability Steering

Centralized

Decentralized—​ Suitable work Strong Central Steering

Centralized

SA

FR-​2009

Decentralized—​ Work Strong Central Steering

Suitable work

UK-​2010

SA

DK-​2010

Centralized

Decentralized—​ Work Strong Central Steering

UA

DE-​2008

Decentralized—​ Suitable work Weak Central Steering

NTH-​2009 SA

SA

CZ-​2009

MI Model of Aims scheme benefit administration

Table 5.A1  Description of minimum income schemes

APPENDIX

14

11 5

European Social Assistance Programs  115

LIST OF INDICATORS Type of minimum income scheme SA: Social Assistance scheme. UA: Unemployment Assistance scheme. TARGETED: Targeted scheme. Model of benefit administration Centralized: Central government body/​agency decides on eligibility for minimum income benefits. Decentralized, Strong Central Steering: Sub-​national authority decides on eligibility for minimum income benefits; however, the level of benefits and the use of central funds are conditioned by central government. Decentralized, Weak Central Steering: Sub-​national authority decides on eligibility for minimum income benefits and (where applicable) is fully responsible for determining the use of central funds. However, central government is able to influence the decision on the level of minimum income benefits. Fully Decentralized: Sub-​national authority is responsible for payment of benefits and is fully responsible for determining the level of minimum income benefits and, where applicable, the use of central funds. Aims and objectives Work: Finding work is the most important obligation for the recipient. The recipient must take a job, even if does not match his or her previous occupation or wage. Suitable Work: Finding work is the most important obligation for the recipient. The recipient can refuse a job that does not match his or her previous occupation or wage. Strengthening Employability: Finding work is the main aim for the recipient; however, the scheme recognizes that this might involve other activities (education, training, health, work experience) that are aimed at strengthening his or her employability. Social Insertion: Finding work is part of a series of obligations aimed at reinserting the individual into society. Sanctions Repressive: A first infringement leads to the termination of minimum income benefits. Harsh Restitutive: A first infringement leads to a cut of at the least 66% of the next monthly benefit payment.

16

116  Work and the Social Safety Net Strong Restitutive: A first infringement leads to a cut of between 33% and 65% of the next monthly benefit payment. Soft Restitutive: A first infringement leads to a cut of less than 33% of the next monthly benefit. Importance given to training and education High: A training or education option is available, and the waiting period is under twelve months. Medium: A training or education option is available, but only after twelve months on minimum income benefits. Low: No guarantee of training or education. Financial incentive approach Comprehensive: Minimum income recipients who take up work are eligible for in-​work benefits. This is coupled with policy tools to reduce the level of minimum income benefits or to decouple them from changes in the cost of living. Negative Incentive: Policymakers use policy tools to reduce the level of minimum income benefits or to decouple them from changes in the cost of living. Minimum income recipients who take up work are not eligible for in-​work benefits. Positive Incentive: Minimum income recipients who take up work are eligible for in-​work benefits. Minimum income benefits, more or less, follow changes in the cost of living. No Incentive: Minimum income recipients who take up work are not eligible for in-​work benefits. Minimum income benefits remain unaltered and, more or less, follow changes in the cost of living. Integration of benefit and activation services Integrated: Benefit administration and activation services are delivered by a single body/​agency and leadership at the national or sub-​ national level. Co-​location: Benefit administration and activation services are delivered by different organizations but provided through one-​stop shops. Does not imply a common leadership. Partial Split: A single body/​agency is responsible for administration of minimum income benefits and the delivery of activation services, but some activation services are delivered by a different institution. Split: Benefit administration and activation services are delivered by different bodies/​agencies at the national or sub-​national level.

11 7

European Social Assistance Programs  117 Use of private providers Market-​Oriented: Private organizations have a role in at the least one area of service provision, and this is done through competitive tendering, preferred-​provider contracts, or a voucher scheme. Network-​Oriented: Private organizations have a role in at least one area of service provision, and this is done through formal partnerships. State-​Centered: Private organizations are not involved in service provision. Customer-​focus in the delivery of services Strong: Case managers are in use. Procedural: Personal Action Plans are mandatory. Case managers are not in use. Poor: Personal Action Plans are not mandatory. Case managers are not in use.

REFERENCES Béraud, M. and A. Eydoux. Activation des chômeurs et modernisation du service public de l’emploi: Les inflexions du régime français d’activation. Travail et emploi 119 (2009): 9–​21.Beraud, Mathieu, and Anne Eydoux. Activating the Unemployed and Modernizing the Public Employment Services. French Institutional Dynamics and Current Challenges. Paper presented at Modernizing Labour Market Institutions conference, September 2008.  http://​www.fep.up.pt/​conf​eren​ces/​iwp​lms/​doc​umen​tos/​wp_​pap​ers/​ pape​r_​be​raud​_e​ yd​oux.pdf. Bergmark, Ake, and Renate Minas. Actors and Governance Arrangements in the Field of Social Assistance. In Rescaling Social Policies: Towards Multilevel Governance in Europe, ed. Yuri Kazepov. Surrey, UK: Ashgate, 2010, 241–​274. Besharov, Douglas J. Testimony on SNAP, House Committee on Agriculture, February 25, 2015. https://​leg​acyl​ampt​est-​2.umd.edu/​pubs/​foo​dass​ist/​Besh​ arov​_​SNA​P_​Te​stim​ony_​15_​0​225.pdf Boston, Jonathan. Basic NPM Ideas and Their Development. In The Ashgate Research Companion to New Public Management, ed. Tom Christensen and Per Lægreid. Surrey, UK: Ashgate, 2010:17–​32. Bradshaw, J. Integrated Approaches to Active Welfare and Employment Policies: Coordination in Activation Policies for Minimum Income Recipients. Dublin: European Foundation for the Improvement of Living and Working Conditions, 2000. Capucha, L. Rendimento Mínimo Garantido: Avaliação da Fase Experimental. Lisbon, Portugal: CIES/​DEPP-​MTS, 1998.

18

118  Work and the Social Safety Net Carmel, Emma, and Theodoros Papadopoulos. The New Governance of Social Security in Britain. In Understanding Social Security: Issues for Social Policy and Practice, 1st ed., ed. Jane Millar. Bristol, UK: Policy Press, 2003:93–​110. Christensen, Tom, and Per Lægreid, eds. The Ashgate Research Companion to New Public Management. Surrey, UK: Ashgate, 2010. Clasen, Jochen, and Alexander Goerne. Germany: Ambivalent Activation. In Activation or Workfare? Governance and Neo-​Liberal Convergence, ed. Ivar Lødemel and Amílcar Moreira. Oxford: Oxford University Press, 2014. Clegg, Daniel, and Bruno Palier. Implementing a Myth: The Evolution of Conditionality in French Minimum Income Provision. In Activation or Workfare? Governance and Neo-​Liberal Convergence, ed. Ivar Lødemel and Amílcar Moreira. Oxford: Oxford University Press, 2014. Considine, Mark. Enterprising States: The Public Management of Welfare-​to-​ Work. Cambridge: Cambridge University Press, 2001. Daguerre, Anne, and David Etherington. Workfare in 21st Century Britain. The Erosion of Rights to Social Assistance. Swindon, UK: ESRC, 2014. https://​cdn. evbuc.com/​eve​ntlo​gos/​33091​323/​workf​arei​n21s​tcen​tury​brit​ain.pdf Danish Ministry of Employment. Vejen til varig beskæftigelse—​den enkelte I sentrum. Copenhagen, Denmark: Danish Ministry of Employment, 2014. http://​www.stm.dk/​mul​time​dia/​Veje​n_​ti​l_​va​rig_​besk​aeft​igel​se_​-​.pdf. Department for Work and Pensions. Jobseeker’s Allowance, Labour Market Questions, Special Conditions for JSA(Cont) and Jobseeking Periods. In Decision-​Making Guide 4. London: Department of Work and Pensions, 2015. Department for Work and Pensions. Mandatory Work Activity: Equality Impact Assessment (March). London: Department for Work and Pensions, 2011. http://​www.dwp.gov.uk/​docs/​eia-​mandat​ory-​work-​activ​ity.pdf. Dujol, Jean-​Benoit, and Etienne Grass. La Construction du RSA. Droit Social 3 (2009): 300–​307. Eardley, Tony, Jonathan Bradshaw, John Ditch, Ian Gough, and Peter Whiteford. Social Assistance in OECD Countries: Synthesis Report. London: Her Majesty’s Stationery Office, 1996. Eichhorst, W., Grienberger-​Zingerle, M., and Konle-​Seidl, R. Activation Policies in Germany: From Status Protection to Basic Income Support. In Bringing the jobless into work? Experiences with activation schemes in Europe and the US, ed. W. Eichhorst, O. Kaufmann, R. Konle-​Seidl, and H.-​J. Reinhard. Berlin, Germany: Springer, 2008a:17–​68. Eichhorst Werner, M. Grienberger-​Zingerle, and Regina Konle-​Seidl. Activation Policies in Germany: From Status Protection to Basic Income Support. In Bringing the Jobless into Work? Experiences with Activation Schemes in Europe and the US, ed. Werner Eichhorst, Otto Kaufmann, and Regina Konle-​Seidl. Berlin: Springer, 2008, 17–​68. Eichhorst, Werner, Otto Kaufmann, Regina Konle-​Seidl, and H. J. Reinhard. Bringing the Jobless into Work? An Introduction to Activation Policies. In

11 9

European Social Assistance Programs  119 Bringing the Jobless into Work? Experiences with Activation Schemes in Europe and the US, ed. Werner Eichhorst, Otto Kaufmann, and Regina Konle-​Seidl. Berlin: Springer, 2008, 1–​16. Fenger, H. Welfare Regimes in Central and Eastern Europe: Incorporating Post-​ Communist Countries in a Welfare Regime Typology. Contemporary Issues and Ideas in Social Sciences 3, no. 2 (2007): 1–​30. Fertig, Michael, and Jochen Kluve. A Conceptual Framework for the Evaluation of Comprehensive Labor Market Policy Reforms in Germany. IZA Discussion Paper No. 1099. Bonn: Institute for the Study of Labor (IZA), 2004. http://​ftp. iza.org/​dp1​099.pdf. Gineste, Sandrine, and Bernard Brunhes. European Employment Observatory Review: Adapting Unemployment Benefit Systems to the Economic Cycle, 2011: France. Brussels: European Employment Observatory, 2012. Goul-​Andersen, Jorgen, and Jacob Pedersen. Continuity and Change in Danish Active Labour Market Policy: 1990–​2007. The Battlefield Between Activation and Workfare. CCWS Working Paper No. 54. Aalborg, Denmark: Center for Comparative Welfare State Studies, Aalborg University, 2007. http://​www.dps. aau.dk/​filead​min/​user​_​upl​oad/​ime/​CCWS/​workin​gpap​ers/​2007-​54-​JGA-​ JJP.pdf. Griggs, Julia, Andrew Hammond, and Robert Walker. Activation for All: Welfare Reform in the UK, 1995 to 2009. In Activation or Workfare? Governance and Neo-​ Liberal Convergence, ed. Ivar Lødemel and Amílcar Moreira. Oxford: Oxford University Press, 2014:89–​108. Guibentif, Pierre, and Denis Bouget. As Políticas do Rendimento Mínimo na União Europeia. Lisbon, Portugal: União das Mutualidades Portuguesas, 1997. Gubrium, Erika, Ivar Lødemel. and Ivan Harsløf. Norwegian Activation Reform on a Wave of Wider Welfare State Change: A Critical Assessment. In Activation or Workfare? Governance and Neo-​Liberal Convergence, ed. Ivar Lødemel and Amílcar Moreira. Oxford: Oxford University Press, 2014:19–​46. Hanesh, W., and N. Baltzer. The Role of Social Assistance as Means of Social Inclusion and Activation: A Comparative Study on Minimum Income in Seven European Countries. Helsinki, Finland: STAKES, 2001. Her Majesty’s Treasury. Spending Review 2010. London: Her Majesty’s Treasury, 2010. https://​www.gov.uk/​gov​ernm​ent/​uplo​ads/​sys​tem/​uplo​ads/​atta​chme​nt_​ d​ata/​file/​203​826/​Spend​ing_​revi​ew_​2​010.pdf. Hvinden, Bjørn. Activation: A Nordic Perspective. In Linking Welfare and Work, ed. Matti Heikkilä. Dublin, Ireland: European Foundation for the Improvement of Living and Working Conditions, 1999, 27–​42. Immervoll, H. Minimum-​ income benefits in OECD countries: Policy design, effectiveness and challenges (IZA Discussion Paper No. 4627). Bonn, Germany: IZA, 2009.

120

120  Work and the Social Safety Net Immervoll, Herwing, and Mark Pearson. A Good Time for Making Work Pay? Taking Stock of In-​Work Benefits and Related Measures across the OECD. IZA Policy Paper No. 3. Bonn: Institute for the Study of Labor (IZA), 2009. http://​ ftp.iza.org/​pp3.pdf. Kane, Aina A., and Julia Kohler-​Olsen. Aktivitetsplikt for sosialhjelpsmottakere—​ har lovgiveren funnet opp hjulet på nytt? Tidsskrift for Erstatningsrett 4. Copenhagen, Denmark: Gyldendal, 2015. Kildal, Nanna. Workfare Tendencies in Scandinavian Welfare Policies. Geneva: International Labour Organization, 2001. http://​citese​erx.ist.psu. edu/​view​doc/​downl​oad?doi=​10.1.1.538.7971&rep=​rep1&type=​pdf. Kvist, Jon, and Ivan Harsløf. Workfare with Welfare Revisited: Instigating Dual Tracks Based on Ethnicity. In Activation or Workfare? Governance and Neo-​ Liberal Convergence, ed. Ivar Lødemel and Amílcar Moreira. Oxford: Oxford University Press, 2014:47–​72. Lødemel, Ivar, and Erica Gubrium. Trajectories of Change: Activation Reforms from Inception to Times of Austerity. In Activation or Workfare? Governance and Neo-​ Liberal Convergence, ed. Ivar Lødemel and Amílcar Moreira. Oxford: Oxford University Press, 2014:327–​348. Lødemel, Ivar, and Heather Trickey, eds. “An Offer You Can’t Refuse”: Workfare in International Perspective. Bristol, UK: Policy Press, 2001. Moreira, Amílcar. The Activation Dilemma. Reconciling the Fairness and Effectiveness of Minimum Income Schemes in Europe. Bristol, UK: Policy Press, 2008. Moreira, Amílcar, Daniel Carelo, and Rui Nicola. From Gateway to Safety Net: The Dynamics of Activation Reforms in Portugal. In Activation or Workfare? Governance and Neo-​Liberal Convergence, ed. Ivar Lødemel and Amílcar Moreira. Oxford: Oxford University Press, 2014:229–​254. Moreira, Amílcar, and Ivar Lødemel. Governing Activation in the 21st Century: A (Hi)story of Change. In Activation or Workfare? Governance and Neo-​Liberal Convergence, ed. Ivar Lødemel and Amílcar Moreira. Oxford: Oxford University Press, 2014:289–​326. Mosley, Hugh. Basic Income Support for Jobseekers in Germany. Statements and Comments. Brussels: European Commission, 2007. http://​pdf.mut​ual-​learn​ ing-​emp​loym​ent.net/​Downlo​ads/​Neth​erla​nds2​007/​GE_​N​L07.pdf. Natili, M. Worlds Of Last-​Resort Safety Nets? A Proposed Typology of Minimum Income Schemes in Europe. Journal of International and Comparative Social Policy 36, no. 1 (2020): 57–​75. doi:10.1080/​21699763.2019.1641134 Newman, Janet. The Double Dynamics of Activation: Institutions, Citizens and the Remaking of Welfare Governance. International Journal of Sociology and Social Policy 27, nos. 9/​10 (2007): 364–​376. Pollitt, Christopher. Clarifying Convergence. Striking Similarities and Durable Differences in Public Management Reform. Public Management Review 3, no. 4 (2002): 471–​492.

12 1

European Social Assistance Programs  121 Ferrera, Maurizio, Anton Hemerijck, and Rhodes, Martin. The Future of Social Europe: Recasting Work and Welfare in the New Economy. Oeiras, Portugal: Celta Editora, 2000. Sirovátka, Tomáš. From Protection Towards Activation: Reform of Social Assistance in the Czech Republic. In Activation or Workfare? Governance and Neo-​ Liberal Convergence, ed. Ivar Lødemel and Amílcar Moreira. Oxford: Oxford University Press, 2014:256–​288. Sirovátka, Tomáš. When Workfare Fails: Post-​Crisis Activation Reform in the Czech Republic. International Journal of Sociology and Social Policy 36, nos. 1/​2 (2016): 86–​101. Sol, E., M. Sichert, H. van Lieshout, and T. Koning. Activation as a Socio-​ Economic and Legal Concept: Laboratorium the Netherlands. In Bringing the Jobless into Work? Experiences with Activation Schemes in Europe and the US, ed. Werner Eichhorst, Otto Kaufmann, and Regina Konle-​Seidl. Berlin: Springer, 2008:161–​220. Spies, Henk, and Nico van de Vrie. From Legitimacy to Effectiveness: Developments in Activation in The Netherlands. In Activation or Workfare? Governance and Neo-​ Liberal Convergence, ed. Ivar Lødemel and Amílcar Moreira. Oxford: Oxford University Press, 2014:143–​171. Statistics Netherlands. 300 Thousand Social Assistance Benefits. Statistics Netherlands Press Release PB10-​057, 2010. http://​www.cbs.nl/​NR/​rdonly​res/​ 87A13​2C9-​4885-​4B2D-​9630-​ACA91​45AD​4EF/​0/​pb10e​057.pdf. Streeck Wolfgang, and Kathleen Thelen, eds. Beyond Continuity: Institutional Change in Advanced Political Economies. Oxford: Oxford University Press, 2005. Tergeist, Peter, and David Grubb. Activation Strategies and the Performance of Employment Services in Germany, the Netherlands and the United Kingdom. OECD Social, Employment and Migration Working Papers No. 42. Paris: OECD Publishing, 2006. doi:10.1787/​341116536484. Torfing, J. Workfare With Welfare: Recent Reforms of the Danish Welfare State. Journal of European Social Policy 9, no. 1 (1999): 5–​28. Trickey, H. Comparing Workfare Programmes—​Features and Implications. In An offer you can’t refuse: Workfare in international perspective, ed. I. Lødemel and H. Trickey. Bristol, UK: Policy Press, 2001:247–​294. Van Berkel, Rik. The Provision of Income Protection and Activation Services for the Unemployed in ‘Active’ Welfare States. An International Comparison. Journal of Social Policy 39, no. 1 (2009): 17–​34. Van Berkel, Rik, and Willibrord De Graaf. The Liberal Governance of a Non-​ Liberal Welfare State? The Case of the Netherlands. In The Governance of Active Welfare States in Europe, ed. Rik Van Berkel, Willibrord de Graaf, and Tomáš Sirovatka. Basingstoke, UK: Palgrave Macmillan, 2011, 132–​152.

12

6 LESS ACTIVATION IN US SOCIAL ASSISTANCE PROGRAMS? Matt Weidinger

OVERVIEW Work activation in the US welfare program called Temporary Assistance for Needy Families (TANF) remains a political football over which both US political parties continue to argue. Recent developments in many ways resemble the debate during the now twenty-​five years since enactment of the 1996 welfare reform law that created TANF, with most Republicans seeking to strengthen the degree to which the TANF program expects recipients to work, while most Democrats seek to weaken work requirements, time limits, and funding restraints that were hallmarks of the 1996 law. But those ongoing disputes have proceeded seemingly despite the actual effects of TANF work activation policies, which have continued to weaken with time. That is in significant part due to one of the earliest and most enduring effects of the 1996 reforms: record reductions in welfare caseloads, which remained near historic lows during both the Great Recession and the economic disruptions resulting from the COVID-​ 19 pandemic. As a result of those large and ongoing caseload reductions, states throughout the TANF era have been able to claim “credits” that reduce—​often to zero—​the share of adults on TANF required to work or train as a condition of collecting benefits. Since the 1996 law, the partisan contests to, on the right, reinvigorate a significant work activation requirement for the current TANF caseload and, on the left, increase welfare caseloads and funding, have been largely fought to a draw. Neither party, even during periods when they controlled both the White Matt Weidinger, Less Activation in US Social Assistance Programs? In: Work and the Social Safety Net. Edited by: Douglas J. Besharov and Douglas M. Call, Oxford University Press. © Oxford University Press 2023. DOI: 10.1093/​oso/​9780190241599.003.0006

12 3

Activation in US Social Assistance Programs  123 House and Congress, has been able to significantly alter the program to match their policy objectives. And now, for over a decade, the TANF program has operated largely without change, based only on short-​term extensions of current law that have continued to result in real annual cuts in the value of the TANF block grant, which is not adjusted for inflation.1 Meanwhile, recent years have seen policymakers on the left as well as the right turn their attention toward expanding other benefit programs. Today, Supplemental Nutrition Assistance Program (SNAP, formerly food stamps), unemployment, and especially Child Tax Credit benefits flow to far more recipient families with children and in far greater amounts than TANF ever provided. Those expanded programs have largely eclipsed the importance of the TANF program and its work activation requirements. In particular, if revived, the expanded and monthly Child Tax Credit payments made in 2021 will supplant the work activity–​backed TANF program with a benefit that, in many states, is more generous than the pre-​TANF Aid to Families with Dependent Children (AFDC) program and is also payable without TANF’s work requirements, time limits, or other restrictions. The net effect will be to further diminish the importance of both the TANF program and its work activation policies among the array of US benefit programs. And by once again paying regular government checks to parents regardless of their engaging in work or other activity, this policy also threatens to extinguish any remaining pretense that work activation—​a key theme of the major welfare reforms enacted in 1996—​remains a significant feature of US welfare policy.

WORK ACTIVATION POLICY UNDER ADC AND AFDC While the history of work activation in the United States extends to before the country’s founding, the lineage of the current TANF program and its posture toward work activation stems directly from the Aid to Dependent Children (ADC) program, created in 1935, as part of President Franklin Roosevelt’s New Deal.2 As detailed in a 1934 report of the Roosevelt administration’s Committee on Economic Security, the authors of the ADC program described its proposed benefits as designed specifically to forestall work by recipient mothers: “The theory of the system of aid to dependent children is that the families will be given enough assistance to meet their minimum budgetary needs, without necessitating gainful employment for mothers of young children, which would be detrimental to home life.” (Committee on Economic Security 1934). Especially 1. For a discussion of TANF extensions and their effect in terms of real TANF funding, see Weidinger (2018). 2. For a useful history of American welfare policy, including on work activation themes, see Olasky (1995).

124

124  Work and the Social Safety Net compared with the far broader eventual scope of the New Deal’s payroll tax–​ funded social insurance programs like unemployment and social security, ADC benefits were paid to a relative handful of mothers—​160,171 in 1936 (Bucklin 1939). The program’s original provision of benefits primarily to widows, along with the attempts by New Deal programs to reserve benefits for prime-​age and primarily male breadwinners, explains ADC’s aversion to gainful employment for its original recipients. Other design features, along with the systematic exclusion of minority families, further served to keep recipiency low (Gordon and Batlan 2011). In the ensuing decades, the ADC program expanded to include some two-​ parent households and also came to include incentives (and eventually even some limited requirements) for recipient parents to work (Falk 2021). Nonetheless, driven by expanding eligibility—​ including the ability of black families to enroll—​and rising family breakdown, caseloads more than tripled between 1964 and 1973, reaching 3.1 million families (Falk 2021). And as rising numbers of women entered the workforce in the 1970s and beyond, political pressure grew for the single and often never-​married mothers who increasingly dominated the caseload of the now AFDC program to engage in—​or at least prepare for—​work as a condition of collecting benefits. By the 1980s, the Reagan administration sought to encourage more state control over welfare programs, for example, by allowing states to require recipients to engage in work experience or community service in exchange for benefits as well as by authorizing demonstrations permitting local testing of welfare-​ to-​work programs. A 1982 proposal to completely devolve cash assistance to states was not enacted (Falk 2021). But the White House was also interested in federal reforms, spurred in part by an influential 1983 study by Mary Jo Bane and David Ellwood (both later welfare officials in the Clinton administration) that noted that while most spells on AFDC were relatively short, 65% of current recipients would eventually spend eight or more years on the rolls (Bane and Ellwood 1985). When the Senate changed hands following the 1986 election, Democrats, now in control of both branches of Congress, took a more active role in reform efforts, crafting what became the Family Support Act (P.L. 100-​485), which President Reagan signed into law in October 1988. Among its many provisions, that legislation created the Job Opportunities and Basic Skills (JOBS) program (which provided employment services, education, and training for cash assistance recipients), removed a putative barrier to leaving welfare by establishing the Transitional Medical Assistance program that continued temporary Medicaid coverage for recipients moving from welfare to work, and guaranteed childcare for AFDC recipients “engaged in work activities” along with transitional childcare for those who left AFDC for work (Peskin, Isaacs, and Fairbanks 1989). That legislation included optional activity requirements, which provided a rhetorical marker on the path to more concrete requirements to come. For

12 5

Activation in US Social Assistance Programs  125 example, the JOBS program included state “participation rate” targets that rose from 7% of nonexempt adults in 1990 to 20% in 1995. But, in practice, the law included broad exceptions to those nominal targets. For example, 38% of recipients were expected to be exempt from participation because they were caring for a child under age three. Other exceptions were made for those who were ill, among other categories (Peskin, Isaacs, and Fairbanks 1989). Data from the Department of Health and Human Services (HHS) suggest the years between 1988 and 1994 saw modest increases in full-​time work (which rose from 2.1% to 3.2% of recipient mothers) and part-​time work (which rose from 4.2% to 4.5%). In keeping with the focus of the program, participation in education and training rose from 2.2% to 13.8%, but that was more than offset by significant declines in the share of mothers seeking work (which fell from 27.6% to 1.7%) (US House of Representatives 1996). As a result, in the years immediately following the enactment of the Family Support Act, the share of AFDC mothers labeled as “needed at home or not actively seeking work” swelled from 55.5% to 71.7%. And instead of helping to drive caseloads down as the Congressional Budget Office’s assessment of the Family Support Act anticipated, the AFDC caseload increased to record levels.3 The nonpartisan Congressional Research Service (CRS) notes that “Between July 1989 and August 1992, almost 1 million families were added to the AFDC basic (primarily one-​parent) caseload, a 27-​percent increase; and by 1992 about one in seven US families with children was enrolled in the program.” (US House of Representatives 1996). A modest intervening recession and the effects of the crack cocaine epidemic are sometimes given as causes for this sharp caseload rise, while studies also point to program and demographic shifts.4 Whatever its cause, the rapid caseload growth following enactment of the Family Support Act contributed to calls for more aggressive reforms, such as making eligibility for AFDC more widely conditioned on work or job preparation by able-​bodied adults. Building off Bane and Ellwood’s earlier work, other studies that noted lengthy stays on benefits contributed to higher caseloads amplified such calls for reform. For example, a 1995 study by LaDonna Pavetti of the Urban Institute found that the average lifetime stay on welfare, counting repeat spells, for families currently enrolled was thirteen years (Pavetti 1995). Foundational to those post-​FSA calls for change were governors from both parties, including Governor Bill Clinton of Arkansas, who in 1992 made his pledge to “end welfare as we know it” a central element of his successful campaign

3. Peskin, Isaacs, and Fairbanks (1989) noted: “By CBO’s estimates, 50,000 families will be off AFDC by the end of the five-​year period, a reduction of about 1 percent in caseloads.” 4. See, for example, Blank (2001) and Congressional Budget Office (1991).

126

126  Work and the Social Safety Net for President (Washington Post 1992). Clinton argued that the AFDC program and its system of monthly checks for millions of low-​income adults with children had become “a way of life” for too many. We will say to those on welfare: “You will have, and you deserve, the opportunity, through training and education, through child care and medical coverage, to liberate yourself. But then, when you can, you must work, because welfare should be a second chance, not a way of life.” (New York Times 1992). Thus, sixty years after the ADC program was originally designed to replace the need for earnings for a small number of low-​income mothers, the US welfare program was now poised to attempt the opposite: to condition receipt of benefits by millions of low-​income mothers on work or related activities (Box 6.1).

Box 6.1  The AFDC program in 1994 and 1995 The following are key characteristics of the AFDC program in 1994 and 1995, immediately prior to the 1996 reforms:a • AFDC benefits were provided to 15% of all US families with children. • Five million families, including 14.2 million total recipients and 9.6 million children, collected AFDC in fiscal year 1994, both all-​time highs. • The average monthly AFDC benefit per family in 1995 was $377 (the equivalent of $671 in 2021 dollars). • Federal and state AFDC benefit expenditures in 1994 totaled $22.8 billion (the equivalent of almost $42 billion in 2021 dollars) • The most common cause (at 74%) for first entering AFDC was having a baby within the last six months. Ninety percent of the youngest children on AFDC had no father in the household. • Of families then collecting AFDC, the average time on the program over the adult’s lifetime was thirteen years. Overrepresented among those long-​term recipients were parents who entered the program before age twenty-​four, had less than twelve years of education, were Hispanic or black, were never married, or had no recent work experience. • Only 9% of households on AFDC reported any earnings at all. • The share of adults on AFDC participating in the JOBS program was 11%. The most common “JOBS” activity was participation in postsecondary education.b US House, Ways and Means Committee, Section 8, “1996 Green Book.”

a

General Accounting Office, “Welfare to Work: Most AFDC Training Programs Not Emphasizing Job Placement,” May 1995, https://​www.gao.gov/​ass​ets/​hehs-​95-​113.pdf. b

12 7

Activation in US Social Assistance Programs  127

CREATION OF TANF “WORK PARTICIPATION REQUIREMENTS” Despite his campaign pledge to reform welfare, once in the White House, President Clinton instead initially focused on controversial budget and healthcare reform legislation. His proposal to reform welfare was unveiled only in June 1994. The Clinton administration proposal “would have phased in a two-​year limit on AFDC receipt without work, followed by required participation in a wage paying work program after two years.” (Falk 2021). This proposal immediately drew the ire of liberal Democrats in Congress, who attacked it in committee and blocked its consideration. (Pianin 1994). One key Democratic policymaker, Rep. Harold Ford, Sr. of Tennessee, who chaired the House Ways and Means welfare subcommittee, vowed to oppose any reform legislation that restricted eligibility for AFDC unless it also guaranteed recipients employment paying at least $9 per hour (Haskins 2007). That impasse left the AFDC system intact, despite Clinton’s promises of reform, ahead of the pivotal 1994 congressional elections. In their campaigns in 1994, House Republican candidates adopted Clinton’s “end welfare as we know it” pledge and made work-​oriented welfare reforms a central plank of their ten-​point “Contract with America.” Like the Clinton proposal, it would have required work by adult recipients within two years of enrolling in AFDC. But, unlike the Clinton plan, and in a nod to the Bane and Ellwood studies mentioned above showing long-​term dependence was a reality for many families, it also would have imposed a five-​year lifetime limit on benefits per adult. The legislation also proposed putting limits on the eligibility of unwed teens to welfare checks and a “family cap” that prevented benefits from rising when recipients had more children while on the rolls. In yet another contrast with the Clinton approach, funding for AFDC and childcare would be capped, and states would have the option to receive AFDC funds in the form of a block grant (Falk 2021). After Republicans claimed the House majority in the November 1994 elections for the first time in forty years (along with control of the Senate for the first time since the mid-​1980s), they worked closely with Republican governors, including those who had previously embraced welfare reforms in their states, to advance federal reform legislation. Consideration of the resulting House bill included dozens of contentious committee hearings and markup sessions, which featured public demonstrations and opponents’ suggesting the legislation’s supporters were cruelly using the nation’s children as “guinea pigs” (Haskins 2007, 150). The legislation ended the entitlement to welfare checks, replaced AFDC with a block grant, ended the JOBS program, required states to ensure that at least 50% of adult recipients participated in work activities by 2003, and included provisions ending benefits for unwed teen parents and requiring states to adopt the “family cap.” It also included new child protection, nutrition, and supplemental security

128

128  Work and the Social Safety Net income block grants; significant work requirements for food stamp benefits; and broad restrictions on benefits for noncitizens (Work Opportunity Act of 1995). Floor debate matched the intensity of committee consideration (Haskins 2007). And, demonstrating the bipartisan appeal of work requirements, Democrats offered a substitute bill, sponsored by Nathan Deal of Georgia (who within a month would switch parties), which supporters argued was “real” welfare reform because it raised the nominal target work participation rate to eventually 52% while increasing funding for education, training, and childcare (Individual Responsibility Act of 1995). All House Democrats who voted supported the Deal substitute, which was defeated before the majority Republican plan was approved on a 234–​199 vote on March 24, 1995. In the end, all House Members save one (Democrat Walter Tucker of California, absent during the vote on the Deal substitute) supported overturning the welfare status quo and creating a new system of work requirements under either the Deal substitute or the Republican-​sponsored bill. The Senate, led by Republican Majority Leader Robert Dole of Kansas, spent the next six months crafting its alterative. Key issues included federal funding for states, the required level of state funding (called “maintenance of effort” or MOE) for the new TANF program, additional funding for childcare, and the nature of restrictions on benefits for unwed teens and the family cap. The Senate ultimately rejected what some (including some governors) derided as “conservative strings” on the block grant (such as ending benefits for teen parents and requiring the family cap) and included $3 billion in additional funding for childcare before approving its legislation in a 87 to 12 vote on September 19, 1995. The subsequent conference agreement was presented to President Clinton twice: once as part of broad budget savings legislation and a second time as a freestanding welfare reform bill. Clinton vetoed both versions in December 1995 and January 1996, respectively, arguing the legislation did “too little to move people from welfare to work” and made excessive cuts in spending on benefits (New York Times 1996b). Following those vetoes, Congressional Republicans, again working with the nation’s governors, assembled a third version of the legislation. Along the way, presidential politics became more closely enmeshed in the welfare reform debate as Dole emerged as the frontrunner and eventual presidential nominee of the Republican party. Eager to run against a Clinton record of twice vetoing welfare reform, Dole and his advisers were reluctant to offer Clinton a third chance to sign a bill. As negotiations on a third bill unfolded, a key flashpoint became whether to sever the main welfare reform legislation, including its popular work requirements, from policies such as proposed Medicaid and food stamp block grants, which were strongly opposed by the administration. On the other side of the debate were many of the nation’s governors, including Republicans who supported those block grants. But House Speaker Newt Gingrich jettisoned the idea of a food stamp block grant, insisting that benefit would remain an

12 9

Activation in US Social Assistance Programs  129 open-​ended federal entitlement. And a revolt led by the large House Republican freshman class—​who feared that without a welfare reform law they would have little record to run on that fall—​forced the leadership to drop the proposed Medicaid block grant as well. In the end, the House rank and file chose their own political fortunes over those of the Dole campaign, and leadership was forced to go along. With those changes, and further increases in funding for childcare, the path was cleared for final legislative action. The legislation took the form of a reconciliation bill, which passed the House and Senate in late July prior to an accelerated conference that negotiated the final agreement. On July 31, Clinton, advised by his pollster that a third veto of welfare reform could be “politically catastrophic” to his re-​election prospects, issued a public statement that he would sign the legislation (Harris and Yang 1996). The conference report was then quickly approved by large bipartisan majorities in the House (voting 328–​101 later on July 31) and Senate (voting 78–​21 on August 1). In addition to nearly unanimous support among Republicans, a majority of Democrats in the Senate and half of Democrats in the House supported the conference agreement. Clinton signed the “Personal Responsibility and Work Opportunity Reconciliation Act” into law (P.L. 104-​193) on August 22, 1996. The legislation ended AFDC, marking the first time a major New Deal program had been terminated. It replaced the individual entitlement to AFDC with the TANF program, a broad new block grant to states funding welfare checks but also a wide range of benefits, services, and activities that states could deploy to assist families with leaving welfare or avoiding it in the first place in favor of work. Compared with the vetoed bills, the legislation included $4 billion in additional funding for childcare, a key administration aim (Government Publishing Office 1996). The legislation included far-​reaching changes in child support, children’s disability benefits, eligibility of noncitizens for major benefits, and more. Supplemental Security Income (SSI) and food stamp changes contributed the bulk of the legislation’s $54 billion in projected savings over seven years (Congressional Budget Office 1996) In terms of work activation, or what policymakers crafting the new law termed “work participation requirements,” on its face, the new law created the strongest work requirements ever included in federal law. Those included work participation requirements on states, under which states were expected to engage a steadily rising share of all families on welfare with an adult recipient (and, after 2006, what was dubbed a “work-​eligible individual”) in one or more of twelve specific “work activities” including unsubsidized employment, job search, and participation in specified education and training activities. By 2002, the share would reach 50% (minus a “caseload reduction” credit). States that failed to meet that target would lose federal funds (Haskins and Weidinger 2019). The loss of funds for failing to satisfy state work participation requirements was designed to be meaningful and increasingly painful for states—​and thus

130

130  Work and the Social Safety Net encourage them to take effective measures to satisfy the requirements. For the first year of failure, a state could lose up to 5% of its TANF block grant (minus any transfers to other programs), rising in 2 percentage point increments for each subsequent year of failure. HHS could reduce the penalty based on the degree of noncompliance or if noncompliance was due to “extraordinary circumstances.” That financial penalty for failure to satisfy the work requirements reflected the flip side of the financial advantages that states realized from the new block grant itself. Not only did they have new-​found freedom to allocate federal and state resources as they saw fit to assist families and otherwise satisfy the federal law’s terms: if they were successful in meeting those objectives without spending the full federal block grant, they also could save—​or, as displayed later, divert—​ federal and some state funds to other purposes. Alongside the apparent 50% work participation requirement on states, the expressed requirement that any individual adult on TANF must work in exchange for benefits applied only after the adult was on the caseload for two years (or less at state option), and individual states could define “work” for this purpose. The legislation also included “pro rata” penalties against individuals who failed to work as required by states. Subsequent research by the Government Accountability Office revealed that, in 1998, about 5% of TANF families were typically subject to sanction for failing to satisfy work and other program requirements, in most cases experiencing partial sanctions instead of the loss of all TANF benefits (General Accounting Office 2000).

EFFECTS OF THE 1996 REFORMS Not surprisingly, supporters of the new law predicted a coming era of rising work and personal responsibility and declining poverty and dependence. Reform opponents, on the other hand, predicted that changes in the law would result in disaster. For example, the New York Times opined, “This is not reform, it is punishment. . . . The effect on cities will be devastating” (New York Times 1996a). The late Sen. Daniel Moynihan (D-​NY) predicted children “sleeping on grates” in a dystopian “Grate Society” (US Congress 1996). Future Speaker Nancy Pelosi (D-​ CA) ventured, “The Republican welfare reform proposal will make the problems of poverty and dependence much worse because it refuses to make work the cornerstone of welfare reform” (US House of Representatives 1996b). Several senior Clinton administration officials resigned in protest of the president’s signing the new law. The years after 1996 suggest that key effects of the new law were closer to those anticipated by the law’s supporters than its detractors. Contrary to Rep. Pelosi’s forecast of “much worse” dependence, welfare caseloads dropped sharply following enactment of the new law.

13 1

Activation in US Social Assistance Programs  131 Mar 1994 Historic Peak: 5.1 million families

Millions of Families 6 5 4

Jun 2020 1.1 million families

3 2 1

2020

2015

2010

2005

2000

1995

1990

1985

1980

1975

1970

1965

1959

0

Through June 2020 Note: Gray shading represents recessions Figure 6.1  AFDC/​TANF caseload, 1959–​2020. Source: Congressional Research Service (CRS), with data from the US Department of Health and Human Services (HHS).

As Figure 6.1 displays, the AFDC/​TANF caseload rose sharply following the enactment of the 1988 Family Support Act, peaked in 1994, began to decline as federal welfare reforms were being debated in 1995 and 1996, and then plummeted at unprecedented rates in the years following enactment of those reforms. Only once before had welfare caseloads declined even two years in a row (US Congress 2006). But after reaching its peak, the welfare caseload proceeded to decline for fourteen consecutive years, from just over 5 million families in 1994 to 1.6 million families in 2008, an unprecedented drop of nearly 70% (Haskins and Weidinger 2019). Between August 1996 when the new law was enacted and just five years later in August 2001, the number of families on welfare fell from 4.4 million to under 2.2 million, a decline of more than 50%.5 The declines were ultimately interrupted by the Great Recession but have continued in the years since, with the TANF caseload falling to 1.1 million families in 2019, immediately prior to the pandemic. As Figure 6.2 displays, caseloads remained at that 1.1 million level in June 2020, before subsequently falling to 984,000 in December 2020 (Falk and Landers 2021). As University of Maryland scholar Doug Besharov, writing in 2008 for the American Enterprise Institute noted, caseload declines were generally unrelated to the nature of a state’s work participation requirements:

5. See https://​www.acf.hhs.gov/​sites/​defa​ult/​files/​docume​nts/​ofa/​1996_​1​5mon​ths.pdf and https://​ www.acf.hhs.gov/​sites/​defa​ult/​files/​docume​nts/​ofa/​2001_​15mo​nths​_​tan​ssp.pdf. The 1996 figure is AFDC cases, and 2000 is TANF and Separate State Program/​MOE families.

132

132  Work and the Social Safety Net 90 85 80

Percent

75 70 65 60 55 50 45 40 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Year All women

All single mothers

Never-married mothers

Figure 6.2  Labor force participation rates for selected groups of working-​age women, 1990–​2018. Source: Current Population Survey, US Census Bureau and the Bureau of Labor Statistics. See https://​cps.ipums.org/​cps/​sda.shtml.

Caseloads fell sharply in all states, yet they did so seemingly without regard to whether states developed ambitious programmes or not. They fell in states with strong work-​first requirements and those without them; in states with mandatory work experience (workfare) programmes and those without them; in states with job training programmes and those without them; and in states with generous child care subsidies and those without them. They just fell (Besharov 2008) The widespread caseload declines suggest that federal policies promoting personal responsibility and work over dependence on benefits, while leaving implementation and key policy details to states, had their intended effects. The declines were consistent with fundamental incentives included in the new law, which rewarded states for having smaller welfare caseloads in several ways. First, the provision of fixed federal block grants reversed a prior financial incentive for larger caseloads embedded in the former open-​ended entitlement system—​under which states received greater federal funds when caseloads rose. Instead, the new block grant maintained federal funds at roughly 1995 levels, removing the financial incentive for greater caseloads. In its place, and as caseloads fell, states experienced rapidly rising federal funds per family remaining on welfare. Such funds allowed states to expand spending on work supports such as childcare or nontraditional uses such as child welfare, as the payment of welfare checks fell sharply (Congressional Budget Office 2015).

13 3

Activation in US Social Assistance Programs  133 Second, rapid caseload declines led to many states’ receiving large “caseload reduction credits” toward the new and gradually rising work participation rate requirements (which, for single-​parent households that formed the bulk of the caseload, rose in 5 percentage point increments from 25% in 1997 to 50% in 2002). The result was that most states had little trouble satisfying the new law’s work participation requirements in the years following enactment. State policy choices encouraged or required by the new law often pulled in the direction of smaller caseloads as well. As the Urban Institute noted in 2006, following enactment, “States moved quickly to design welfare-​to-​work policies that emphasized getting recipients into jobs by shifting to ‘work-​first’ welfare systems, modifying program rules to allow more earned income, imposing shorter lifetime limits, and adopting tougher sanctions for families that do not comply with work requirements” (Urban Institute 2006). The years following enactment of the 1996 reforms also saw a significant acceleration in the already growing share of never-​married single mothers (most of whom were not on welfare but comprised the group most likely to enter and stay on the rolls) who participated in the labor force. Figure 6.2 displays how this trend began in 1992 as the economy emerged from the shallow 1990–​1991 recession, continued after Earned Income Tax Credit (EITC) expansions enacted in 1993, and then accelerated sharply in the years following enactment of the 1996 reforms. The figure suggests that mothers who left welfare and similar low-​income mothers with backgrounds that made them most likely to experience welfare spells increased their participation in the labor force and then maintained a historically high level of work in the years following enactment of the 1996 reforms (Haskins and Weidinger 2019).6 As Robert Moffitt and Stephanie Garlow noted in a 2018 review, “In the initial years after reform, many more women joined the labor force than even the reform’s most ardent supporters had hoped” (Moffitt and Garlow 2018). Earnings from work similarly showed significant improvement in the years following the 1996 reforms, especially when coupled with rising work support benefits such as larger EITC payments that more than offset declining welfare payments.7 As shown in Figure 6.3, the mean wage and salary income of   6. The data in Figure 6.1 are labor force participation (LFP) rates, defined by the Department of Labor as the share of mothers either employed or actively looking for work. The seven-​year period leading up to 2000 saw strong increases in the LFP rates of all single mothers, never-​married single mothers (the most disadvantaged group and the most likely to be on welfare), and married mothers. Despite far smaller caseloads and zero effective work participation requirements in many states, LFP rates remained well above their averages prior to welfare reform in the years before the pandemic. 7. As Ron Haskins of the Brookings Institution noted in congressional testimony in 2006, “Census Bureau data for female-​headed families in the bottom 40 percent of the income distribution for female-​headed families (those below about $21,000 in 2000) show that their pattern of income shifted dramatically between 1993 and 2000. In 1993 earnings accounted for about 30 percent of

134

134  Work and the Social Safety Net $40,000

Real 2017 Dollars

$35,000 $30,000 $25,000 $20,000 $15,000 $10,000 $5,000 $0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Year Married mothers

Never-married mothers

Figure 6.3  Mean wage and salary income for working-​age mothers, 1990–​2018. Source: Current Population Survey, US Census Bureau and the Bureau of Labor Statistics. See https://​cps.ipums.org/​cps/​sda.shtml.

never-​married mothers ages 18 to 54 rose sharply after the 1996 reforms and reached nearly double pre-​reform levels prior to the pandemic, in real terms (Haskins and Weidinger 2019). As a number of observers have noted, the strong economy in the late 1990s and policies designed to make work pay, including expansions to the EITC and the introduction of the Child Tax Credit, contributed to this improving work and earnings picture.8 As displayed in Figure 6.4, the increases in work and earnings by unmarried mothers depicted above, accompanied by growing work support benefits, contributed to significant declines in the poverty rates of children following the enactment of the 1996 reforms, as measured by the official as well as the supplemental poverty measures. As Figure 6.4 shows, 20.8% of children were in poverty under the official poverty measure in 1995. But the rate fell over the next five years to around 16.2% the income of low-​income, female-​headed families, while welfare payments, including cash, food stamps, housing, and school lunch, accounted for nearly 55 percent. By 2000 this pattern had reversed: earnings had leaped by an astounding 136 percent, to constitute nearly 60 percent of income, while welfare income had plummeted by over half, to constitute only about 23 percent of income [Figure 6.2]. As a result of the growth in earnings and legislated expansions of the EITC, income from the EITC more than tripled. Thus with earnings and EITC payments leading the way, the total income of these low-​income families increased by about 25 percent over the period (in dollars adjusted for inflation)” (US Congress 2006).” 8. As noted in Haskins and Weidinger (2019), “reviewers across the political spectrum (Blank 2009; Haskins 2017; Grogger and Karoly 2005) have attributed improvements in the rates of work, income, and poverty following the 1996 reforms to a combination of a strong economy, other reforms like the increases in the EITC, and welfare reform policies themselves.”

13 5

Activation in US Social Assistance Programs  135 25%

Poverty Rate

20% 15% 10% 5%

2020

2018

2016

2014

2012

2010

2008

2006

2004

2002

2000

1998

1996

1994

1992

1990

0%

Year Supplemental Poverty Measure

Official Poverty Measure

Figure 6.4  Child poverty rates using two measures of poverty, 1990–​2020.

in 2000. It then rose following the 2001 recession, leveled off until the Great Recession that began in late 2007, and then rose to 22% in 2010. Child poverty declined after 2012 to about 18% in 2016. In 2019, 14.4% of children were poor according to the official poverty measure, a decline of more than 30% from the 1995 level (US Census Bureau 2020). The Supplemental Poverty Measure (SPM), which includes the effect of benefits such as the EITC and food stamps, shows an even greater and more consistently positive improvement in child poverty over the period, falling from 28% in 1993 to 15.2% in 2016 (US Census Bureau 2017). Child poverty as measured by the SPM has since declined further, reaching 12.5% in 2019, marking a more than 50% decline from the 1993 level (US Census Bureau 2020). Despite such significant declines in child poverty and rising government spending on the working poor, critics of TANF express concern about the continued incidence of deep poverty, often pointing to a declining share of poor families receiving TANF benefits as a contributing factor.9 Such concerns spotlight the effects of the changing nature of TANF benefits from mostly cash assistance to increasingly work supports and other forms of assistance. As the General Accounting Office (GAO) noted in congressional testimony in 2002, The increased emphasis on work support and other services for recipients of cash assistance and those not receiving cash assistance represents a significant departure from previous welfare policy that focused on providing monthly cash payments. While the goals and target populations of welfare spending have changed, the key measure of the number of people served

9. For a discussion of rising spending on the working poor, see Hoynes and Schanzenbach (2018).

136

136  Work and the Social Safety Net remains focused solely on families receiving monthly cash assistance. Although this measure provides important information for administrators and policymakers, it does not provide a complete picture of the number of people receiving benefits or services funded at least in part with TANF/​ MOE funds (US General Accounting Office 2002). That changing nature of benefits suggests that the SPM (which includes growing tax benefits such as the EITC and the Child Tax Credit payable before 2021 only when parents work, as well as food stamp benefits) may offer a more complete depiction of need in the post-​reform era. As shown in Figure 6.4, child poverty as measured under the SPM was declining or stable in nearly all years since 1996, reaching an all-​time low in 2019. This indicates that many parents were able to make up for a loss of cash assistance income from TANF by other means, including wages from work, tax benefits related to work, and other benefits, especially food stamps (Tiehan, Jolliffe, and Smeeding 2013). Further, 2019 research by Bruce Meyer and his colleagues finds that more than 90% of those believed to be in extreme poverty are not so once the value of in-​kind transfers such as significant food stamp transfers is added to income, survey reports of earnings and transfer receipt are replaced with more accurate administrative records, and the ownership of assets is accounted for (Haskins and Weidinger 2019).

CHANGES IN THE DEFICIT REDUCTION ACT By authorizing the new TANF program and its many policy changes only through fiscal year 2002, the authors of the 1996 welfare reform law intended for Congress to carefully review its effects before continuing the program. In considering such effects, Republican policymakers focused much of their attention on the steep caseload declines following the 1996 law and their implications for the state work participation rate requirements. For example, as Congress was considering reauthorization legislation in 2002, the GAO testified that thirty-​one states had effectively no work participation rate requirement in fiscal year 2000, when the nominal requirement was 40% (Fagnoni 2002). Republican policymakers thus included policies in welfare reauthorization legislation designed to restore the dynamic the authors of the 1996 law intended states to face by 2002. That is, that states either needed to engage 50% of current adults on TANF in countable work activities each month, reduce their current caseload by an additional 50%, or a combination of the two factors equaling 50%. The House in 2002 and 2003 passed Republican-​authored TANF reauthorization legislation that included such updates (Falk 2021). But the legislation failed to gain traction in a narrowly divided Senate (where 60 votes are normally required to pass such legislation), including because states resisted efforts to strengthen the work requirements absent additional federal funding.

13 7

Activation in US Social Assistance Programs  137 The House in early 2005 again initiated action on similar TANF reauthorization legislation, which Republican proponents argued was needed because “the share of current welfare recipients engaged in work or training has fallen in three of the last four years” (Committee on Ways and Means 2005). That legislation was included in budget reconciliation legislation called the Deficit Reduction Act (DRA), which could advance with just a simple majority of 50 votes in the Senate. The DRA included an extension of current TANF funding—​continuing the federal block grant without an inflation adjustment—​through fiscal year 2010 (Falk 2021). The DRA also left the 50% state work participation requirement unchanged but updated the caseload reduction credit to provide credit for caseload declines since 2005 rather than 1995. It increased mandatory childcare funding by $200 million per year, even as other welfare provisions in the legislation were projected to save $7.2 billion over ten years, driven mainly by reductions in funding for federal child support administration (Congressional Budget Office 2006). But even though the DRA welfare policy tracked the 1996 law in key respects (fixed funding for TANF, strengthened work activity requirements, additional funding for childcare, and other federal savings), this time support for the legislation was highly partisan. The legislation was narrowly approved in December 2005 by the Senate (51–​50, with Vice President Cheney joining 50 Republicans in support) and the House (212–​206, with all Democrats voting no) before President Bush signed it into law in February 2006 (US Congress 2006). The bipartisan support that characterized the welfare reform law in 1996, bolstered by a Democratic president who promised to “end welfare as we know it,” was clearly a thing of the past.

TANF DURING AND AFTER THE GREAT RECESSION Critics of welfare reform and its generally fixed federal funding (the program also included a modest “contingency fund” for states experiencing increased need) have often predicted trouble in the event of a recession and presumed rising demand for benefits. While caseloads remained stable during the generally mild 2001 recession, many were concerned that the Great Recession that began in December 2007 would result in significantly increased demand for benefits. Indeed, anticipating such greater demand for TANF assistance, the American Recovery and Reinvestment Act (ARRA, also known as the Obama stimulus law) in February 2009 changed TANF program funding rules as well as work activation policies. ARRA . . . included $5 billion for a new TANF Emergency Contingency Fund (ECF) available to be spent in FY2009 and FY2010. The ECF supplemented funding for the regular TANF contingency fund, which itself was depleted in early FY2010. The ECF reimbursed states for 80% of

138

138  Work and the Social Safety Net the cost of increased expenditures for basic assistance, short-​term emergency aid, and subsidized employment. ARRA also temporarily froze the TANF caseload reduction credit at prerecession levels, through its application to the FY2011 work participation standards (Falk 2021). ARRA allowed states to substitute the lower of their fiscal year 2007 or 2008 caseloads for the actual caseloads used to calculate the caseload reduction credit in fiscal years 2009, 2010, and 2011 (Office of Family Assistance 2010). In effect, this allowed states to engage fewer current TANF recipients in work activities than they otherwise would have needed to in order to satisfy the work participation requirements and avoid risking federal financial penalties. If this policy was designed to promote significant increases in TANF cash assistance caseloads during the recession, a review of TANF caseload dynamics during this period suggests it had only modest success. The average number of families on TANF programs nationwide fell from 2.1 million in fiscal year 2005 to 1.7 million in fiscal year 2008, a 19% drop. The national caseload then rose only modestly to 1.8 million in fiscal year 2009 and 1.9 million in fiscal years 2010 and 2011. The GAO reported in 2015 that, in fiscal year 2011, forty-​nine of fifty states received some caseload reduction credit, and twenty-​two states saw their work participation rate requirement reduced to zero as a result of that credit.10 By fiscal year 2012 (i.e., once the temporary policy allowing states to freeze caseload reduction credits at their FY 2007 or 2008 levels had expired), the national caseload was declining once again.11 Meanwhile, other programs saw far greater increases in recipients and benefits provided. For example, the New York Times in 2012 noted “soaring” food stamp receipt during and after the Great Recession, even as “the number of Americans receiving cash welfare has fallen since the 1990s” (De Parle 2012). Unemployment benefits were similarly enlarged, with significant expansions in eligibility for, the amount of, and the duration of weekly checks. As one review summarized, “During the Great Recession, SNAP and Unemployment Insurance programs acted as a counterforce to the economic shock: greatly increasing government spending and expanding income support to families in a way that was at least equal to the cushioning role that they have played in previous recessions, if not more” (Bitler 2019). What accounts for the difference in demand for benefits? The TANF program’s financing—​including its shared financial responsibility with the states—​likely played a significant part. As the New York Times noted in 2012, “Since the states get fixed federal grants, any caseload growth comes at their own expense. By 10. The GAO also reported that other states made increasing use of “excess MOE credits” as well as “token checks” to satisfy work participation requirements (US Government Accountability Office 2015). 11. See Office of Family Assistance (2016).

13 9

Activation in US Social Assistance Programs  139 Work Participation Rate

51.9% 53.0% 48.4%

28.9%

27.5%

29.4%

30.3% 30.6% 29.7% 29.4% 29.4% 29.0% 29.5%

34.4% 33.5%

36.6%

48.1%

47.1% 39.6%

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Figure 6.5  National average TANF work participation rate for all families, FY2002–​FY2020. Source: Congressional Research Service (CRS) based on data from the US Department of Health and Human Services (HHS).

contrast, the federal government pays the entire food stamp bill no matter how many people enroll; states encourage applications, and the rolls have reached record highs” (De Parle 2012). For example, officials in Westchester County, New York, described being rewarded for placing families on the fully federally funded food stamp program rather than the partially state-​funded TANF program.12 As noted above, many states were able to satisfy the 50% work participation requirement in the early years of the TANF program through the use of caseload reduction credits. As depicted in Figure 6.5, in each year prior to FY 2016, the share of the national TANF caseload engaged in work activities for enough hours to count consistently fell short of the nominal 50% goal set by the authors of the 1996 law. During the decade from FY 2002 through 2011, the national average work participation rate fell within a narrow band of between 27.5 and 30.6 each year, before rising considerably in the years following the Great Recession, according to the CRS (Falk and Landers 2021). The apparent rise in work participation rates in recent years reflects changing policy choices by states, which some describe as work-​arounds designed to avoid engaging more recipients in work activities. As the CRS notes, “The increase in the work participation rate has not come from an increase in the number of recipients in regular TANF assistance programs who are either working or in job preparation activities. This increase stems mostly from states creating new ‘earnings supplement’ programs that use TANF funds to aid working parents in the Supplemental Nutrition Assistance Program (SNAP, formerly food

12. Personal communication with Douglas Besharov.

140

140  Work and the Social Safety Net 50% standard WPR for all states before credit reductions

WPR

Total Official WPR

Unsubsidized employment

50% Earnings Supplement Regular TANF

25%

Welfare-to-Work Activities 2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

0%

Figure 6.6  TANF work participation rate, FY2002–​FY2016. Source: Congressional Research Service (CRS) tabulations of the TANF National Data Files, 2002–​2016.

stamps) or who have left the regular TANF assistance programs for work” (Falk and Landers 2021). Figure 6.6 displays how those new “earnings supplement” (sometimes called “token payment”) programs are responsible for “(a)lmost all of the increase” in the work participation rate since FY 2012, as CRS describes (Falk 2018). Congress tried to fix some work participation loopholes in its 2006 reauthorization legislation, but states have used a combination of caseload reduction and additional loopholes since then to significantly diminish the share of current TANF recipients expected to engage in work activities. Since 2010, TANF has operated under a series of short-​term extensions without significant policy changes to the underlying program (Weidinger 2018). That has left the program, and especially its intent of engaging a significant share of adults in work and other activities, largely adrift for now over a decade. States continued to claim caseload reduction credits and when necessary tap other loopholes like “excess maintenance of effort (MOE) credits” and the provision of “token payments” to allow them to satisfy the work participation requirements without engaging a significant share of adults in work activities.13 With Democrats, who as discussed below are increasingly focused on ending work requirements altogether in other programs, presently in control

13. For example, the GAO in 2012 found that in 2009 “16 of 45 states that met the TANF work participation rate would not have done so without the credit they received for excess state MOE spending.” See US Government Accountability Office (2012).

14 1

Activation in US Social Assistance Programs  141 of both Congress and the White House, that status seems highly unlikely to change in the near future. About half of all states also currently avoid the separate and higher work requirement for two-​parent families by putting these cases in a solely state-​funded program in which work participation requirements do not apply. Data based on new reporting requirements imposed by HHS show that only 22.3% of TANF funds were spent on basic assistance (i.e., cash payments to families) in 2019. Another 11% was spent on work activities and 3% on work supports. Even if the 16% spent on childcare is included under work supports, only a total of 53% of TANF funds is spent on activities directly related to basic assistance and work (Haskins and Weidinger 2019). To address such perceived flaws, a number of groups and individuals offered TANF reform proposals in the years before the pandemic. These included a comprehensive five-​year reauthorization bill crafted by senior Republicans on the House Committee on Ways and Means, which was approved on a partisan vote in committee in May 2018 but was never considered on the House floor. Senior Republicans in the House and Senate introduced related legislation in early 2019, but it was not acted on in either body; a revised version was reintroduced in both bodies in July 2021, but has not advanced. This legislation would transition from the process-​oriented work participation requirement system to an outcome-​focused accountability system for states, require states to engage all work-​eligible individuals in work and activities (as defined by each state), and phase out the ability of states to count nongovernmental third-​party spending as state spending, designed to minimize the degree to which “excess MOE credits” could reduce the work participation requirements. As a result, immediately prior to the pandemic in FY 2019, half of all states had effectively no work participation rate requirement (Administration for Children and Families 2020). This was due to the caseload reduction credit, now based on caseload declines since 2005. Participation by adults on TANF in activities other than unsubsidized employment remained rare (Administration for Children and Families 2020). The TANF caseload in 2019 numbered around 1 million families with 3 million total recipients (Administration for Children and Families 2020) In contrast, the SNAP (often called food stamps) caseload before the pandemic numbered 35 million recipients—​more than one in ten Americans and ten times the number collecting TANF assistance (US Department of Agriculture 2021). That is evidence of decades of efforts to expand federal food stamp receipt while TANF caseloads, funded by both state and federal funds, continued to drift ever lower.

TANF IN THE COVID-​19 PANDEMIC Since the pandemic struck the US economy in March 2020, the number of TANF recipients and families has fallen slightly (through December 2020), in sharp

142

142  Work and the Social Safety Net contrast with other benefit programs.14 As with prior declines, this likely owes to state TANF policies as well as the fixed block grant funding of the program—​to which states are expected to contribute their own funds. That contrasts with rapidly growing and entirely federally funded programs such as repeated large stimulus checks, growing SNAP benefits, and extended and significantly expanded unemployment checks enacted at the start of the pandemic and extended in December 2020 and March 2021. Similarly, and in contrast with the Great Recession, federal legislative activity involving TANF since the pandemic struck has been remarkably minimal. For example, the March 2020 Coronavirus Aid, Relief, and Economic Security (CARES) Act included only a straight reauthorization of the TANF program through November 2020. Meanwhile, that legislation included historic expansions in eligibility for and the amount of benefits provided to tens of millions of families under unemployment, food, housing, health, and other programs, at a total cost of $1.7 trillion (Congressional Budget Office 2020). Later, the March 2021 American Rescue Plan created the Pandemic Emergency Assistance Fund providing states a share of $1 billion to be used through the end of fiscal year 2022, generally for non-​recurrent short-​term TANF benefits (Office of Family Assistance 2021). Again, this level of additional assistance paled in comparison with other benefits provided to those in need under that and related legislation. For example, between March 2020 and September 2021, an estimated $700 billion in temporary federal unemployment benefits were provided through several major bills (Weidinger 2021b). In terms of work participation requirements, the absence of legislative activity was similarly telling. In contrast with the Great Recession, since the pandemic struck, Congress has not even bothered to act on legislation temporarily minimizing the effect of TANF’s work participation requirements. However, HHS can waive financial penalties for any states that do not satisfy the requirements due to “reasonable cause,” and it has said it would use this authority “to the maximum extent possible” (Falk and Landers 2021). Recent experience suggests that will not be necessary, including because the TANF caseload remains remarkably low. In FY 2020, all states satisfied the main work participation rate requirement, including twenty-​eight states whose 50% target rate was reduced to zero due to the caseload reduction credit (Administration for Children and Families 2021). Few adults participate in activities other than unsubsidized employment, which includes those collecting token checks; nonparticipating adults rarely complete any hours of activity at all. Absent legislative changes, states will continue to claim such credits allowing them to satisfy the work participation requirements

14. Unemployment insurance, for example, grew from under 2 million recipients prior to the pandemic to over 30 million by mid-​2020. See Administration for Children and Families (2020).

14 3

Activation in US Social Assistance Programs  143 without engaging a significant share of adults now on the rolls in productive activities.

CONCLUSION Work activation in the TANF program remains a policy football over which the political parties have continued to struggle in recent years, ultimately resulting in few changes despite repeated opportunities to improve the program. Under current program rules, most states have found it easy to satisfy work participation requirements without placing a significant share of current adult recipients in work, education, training, or other activities. Indeed, some states have recently taken to offering token TANF payments to working adults not on the program in an apparent effort to avoid engaging adults on the program in required activities. That development, alongside the greater effects of the decline in the TANF caseload leading to a large caseload reduction credit, has continued to neuter the practical effect of TANF work participation requirements. It is telling that the TANF caseload has remained near historic lows—​with fewer than 1 million families on the rolls, only about half of whom include an adult subject to work requirements—​despite significantly increased needs stemming from the pandemic and the resulting economic disruptions since March 2020. That is likely a result of the shared responsibility of paying for TANF benefits between state and federal governments, in contrast with new and expanded federal benefits crafted in response to the pandemic that generally relieved states of significant benefit costs. Those federal programs ramped up quickly and remained significantly larger than pre-​pandemic levels in mid-​2021 in terms of benefit receipt and spending even after the worst of the pandemic appeared to have passed. For example, in May 2021, federal SNAP benefits were payable to more than 42 million persons in some 22 million households. In August 2021, the Biden administration announced the largest increase in food stamp benefits ever (USDA 2021; Nova 2021). State and federal unemployment benefits, including unprecedented expansions in eligibility for and the amount of federal benefit checks, are likely to total $900 billion since the start of the pandemic and, at their peak, were paid to more than 30 million recipients.15 And fully refundable monthly federal Child Tax Credit payments were flowing to 39 million households with children in mid-​2021—​nine times the 4.4 million families on the TANF program when it was created in 1996, much less the far smaller caseloads of today.

15. See Weidinger (2021b) for a discussion of benefit spending, and US Department of Labor (2021), for peak claims.

14

144  Work and the Social Safety Net The recent changes to the Child Tax Credit program reflect the sharpest contrast with the TANF program. The Child Tax Credit program’s benefits and number of recipients have significantly increased, including through the elimination of the former requirement that parents work to collect benefits. That policy, enacted for 2021, and which President Biden and other senior Democrats would extend permanently, effectively recreates limitless monthly federal benefit checks for parents without any expectation of work—​in effect, reviving the defining trait of the pre-​reform AFDC program. These developments reflect the growing ascendence of progressives—​who generally opposed the work-​based 1996 welfare reforms and have resisted efforts to strengthen the TANF work requirements ever since—​among the current governing Democratic coalition. That includes President Biden, who was a supporter of the 1996 reforms but now supports eliminating the prior work requirement for adults claiming Child Tax Credits. But instead of directly attacking the TANF program and its work requirements, they have instead converted the Child Tax Credit—​formerly a benefit reserved for workers—​into the equivalent of welfare checks for nearly all parents, regardless of whether they work or not. If extended for additional years beyond 2021, such payments would effectively replace TANF and its expectation of work activity with a new federal benefit paid to almost all parents regardless of their participation in work, training, or other constructive activity. In many states, the Child Tax Credit benefit for a parent with two young children already exceeds the real value of welfare checks paid under AFDC for similar families (Weidinger 2021b). It remains to be seen how state TANF programs react to these federal benefit checks, for which all current TANF recipients and millions of additional low-​income parents were eligible. Especially if this policy is made permanent, some states may further restrict the payment of welfare checks using TANF funds, which could reduce the apparent TANF caseload even more. But regardless of how state TANF programs react, these recent developments only continue the now decades-​long trend of the TANF program becoming a diminishing component of the American social welfare system as a shrinking share of families collect its cash assistance benefits and few of those who do are expected to engage in activities in exchange for that assistance. The absence of significant TANF-​related legislation in the pandemic, a period when other programs were significantly expanded to benefit tens of millions of families with children, has already underscored the TANF program’s growing irrelevance within that broader policy landscape.

REFERENCES Administration for Children and Families. Temporary Assistance for Needy Families (TANF) and Separate State Programs—​Maintenance of Effort (SSP-​ MOE) Work Participation Rates and Engagement in Work Activities Fiscal Year 2019. Washington, DC, 2020.

14 5

Activation in US Social Assistance Programs  145 Administration for Children and Families, Office of Family Assistance, “Work Participation Rates: Fiscal Year 2020,” Washington, DC, 2021, https://​www. acf.hhs.gov/​ofa/​data/​work-​partic​ipat​ion-​rates-​fis​cal-​year-​2020. Bane, Mary Jo, and David Ellwood, “The Dynamics of Dependence: Routes to Self-​Sufficiency,” US Department of Health and Human Services, 1985. Bitler, Marianne. “Responsiveness of the Safety Net During Downturns: Lessons From the Great Recession,” EconoFact, 2019, https://​econof​act.org/​res​pons​ iven​ess-​of-​the-​saf​ety-​net-​dur​ing-​downtu​rns-​less​ons-​f rom-​the-​great-​recess​ion Blank, Rebecca M. “What Causes Public Assistance Caseloads to Grow?” Journal of Human Resources 36, 1 (2001): 85–​118. doi:10.2307/​3069671 Besharov, Douglas J. “Two Cheers for American Welfare Reform: Lessons Learned, Questions Raised, Next Steps,” American Enterprise Institute, Washington, DC, 2008, https://​www.aei.org/​wp-​cont​ent/​uplo​ads/​2011/​10/​ 20081​030_​Besh​arov​Welf​are.pdf?x91​208 Bucklin, Dorothy. “Public Aid for the Care of Dependent Children in Their Own Homes, 1932–​38,” Social Security, 1939, https://​www.ssa.gov/​pol​icy/​docs/​ ssb/​v2n4/​v2n4​p24.pdf. Committee on Economic Security, “Social Security in America,” Washington, DC, 1934, https://​www.ssa.gov/​hist​ory/​repo​rts/​ces/​ces​book​c13.html. Committee on Ways and Means, Subcommittee on Human Resources, markup documents for HR 240, Washington, DC, 2005. Congressional Budget Office, “CBO Staff Memorandum,” December 1991, https://​www.cbo.gov/​sites/​defa​ult/​files/​102nd-​congr​ess-​1991-​1992/​repo​rts/​ 1991_​1​2_​ap​reli​mana​lysi​siof​grow​ing.pdf. Congressional Budget Office, “Federal Budgetary Implications of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996,” Washington, DC, 1996, https://​www.cbo.gov/​sites/​defa​ult/​files/​104th-​congr​ ess-​1995-​1996/​repo​rts/​1996do​c32.pdf. Congressional Budget Office. Letter from Phillip Swagel, Director of Congressional Budget Office, to Honorable Mike Enzi, Chairman of the Committee on the Budget, Washington, DC, 2020, https://​www.cbo.gov/​sys​ tem/​files/​2020-​04/​hr748.pdf. Congressional Budget Office, “S. 1932: Deficit Reduction Act of 2005,” Washington, DC, 2006, https://​www.cbo.gov/​sites/​defa​ult/​files/​109th-​congr​ ess-​2005-​2006/​coste​stim​ate/​s19​32co​nf0.pdf. Congressional Budget Office, “Temporary Assistance for Needy Families: Spending and Policy Options,” Washington, DC, 2015, https://​www.cbo.gov/​sites/​defa​ ult/​files/​114th-​congr​ess-​2015-​2016/​repo​rts/​49887-​tanf.pdf. De Parle, Jason. “Welfare Limits Left Poor Adrift as Recession Hit,” The New York Times, 2012, https://​www.nyti​mes.com/​2012/​04/​08/​us/​welf​are-​lim​its-​left-​ poor-​adr​ift-​as-​recess​ion-​hit.html.

146

146  Work and the Social Safety Net Fagnoni, Cynthia. “States Are Using TANF Flexibility to Adapt Work Requirements and Time Limits to Meet State and Local Needs,” Before the Subcommittee on Human Resources, Committee on Ways and Means, 107th Cong. (2002) Falk, Gene. “The Temporary Assistance for Needy Families (TANF) Block Grant: A Legislative History,” Congressional Research Service, 2021, https://​ greenb​ook-​waysa​ndme​ans.house.gov/​sites/​greenb​ook.waysa​ndme​ans.house. gov/​files/​R44​668. Falk, Gene. “Temporary Assistance for Needy Families: Work Requirements,” Congressional Research Service, Washington, DC, 2018, https://​www.eve​rycr​ srep​ort.com/​files/​2018-​03-​27_​IF10856_​03e12b7c0bf63​e9d8​a450​990d​3efe​ 9260​d437​d66. Falk, Gene, and Patrick Landers, “The Temporary Assistance for Needy Families (TANF) Block Grant: Responses to Frequently Asked Questions,” Congressional Research Service, Washington, DC, 2021, https://​fas.org/​sgp/​ crs/​misc/​RL32​760.pdf. Linda Gordon and Felice Batlan, “The Legal History of the Aid to Dependent Children Program,” 2011, https://​social​welf​are.libr​ary.vcu.edu/​pub​lic-​welf​ are/​aid-​to-​depend​ent-​child​ren-​the-​legal-​hist​ory/​ Government Publishing Office, “Public Papers of the Presidents of the United States: William J. Clinton,” 1996, https://​www.govi​nfo.gov/​cont​ent/​pkg/​PPP-​ 1996-​book2/​pdf/​PPP-​1996-​book2-​doc-​pg1​328.pdf. Harris, John F. and John E. Yang, “Clinton to Sign Bill Overhauling Welfare,” The Washington Post, 1996, https://​www.was​hing​tonp​ost.com/​wp-​srv/​polit​ics/​ spec​ial/​welf​are/​stor​ies/​wf080​196.htm Haskins, Ron. Work Over Welfare: The Inside Story of the 1996 Welfare Reform Law (Washington, DC: Brookings Institution Press, 2007 Haskins, Ron and Matt Weidinger, “The Temporary Assistance for Needy Families Program: Time for Improvements,” Annals of the American Academy of Political and Social Science, 686: 286–​309, 2019, doi:10.1177/​ 0002716219881628. Hoynes, Hilary W. and Diane Whitmore Schanzenbach, “Safety Net Investments in Children,” Brookings Papers on Economic Activity, Washington, DC, 2018, https://​www.brooki​ngs.edu/​bpea-​artic​les/​saf​ety-​net-​inve​stme​nts-​in-​ child​ren/​. Moffitt, Robert and Stephanie Garlow, “Did Welfare Reform Increase Employment and Reduce Poverty?” 2018, https://​ine​qual​ity.stanf​ord.edu/​sites/​defa​ult/​files/​ Pat​hway​s_​Wi​nter​2018​_​Emp​loym​ent-​Pove​rty.pdf. The New York Times, “In Their Own Words: Transcript of Speech by Clinton Accepting Democratic Nomination,” July 17, 1992, https://​www.nyti​mes. com/​1992/​07/​17/​news/​their-​own-​words-​tra​nscr​ipt-​spe​ech-​clin​ton-​accept​ ing-​dem​ocra​tic-​nom​inat​ion.html.

14 7

Activation in US Social Assistance Programs  147 The New York Times, “A Sad Day for Poor Children,” August 1, 1996a, https://​w ww.nyti​mes.com/​1996/​08/​01/​opin​ion/​a-​s ad-​day-​for-​p oor-​child​ ren.html. The New York Times, “Welfare Bill Vetoed,” January 10, 1996b, https://​times​ mach​ine.nyti​mes.com/​times​mach​ine/​1996/​01/​10/​058​696.html?pag​eNum​ ber=​1. Nova, Annie. “Biden Administration Announces the Biggest Increase to Food Stamps Ever,” CNBC, 2021, https://​www.cnbc.com/​2021/​08/​16/​biden-​adm​ inis​trat​ion-​announ​ces-​bigg​est-​incre​ase-​to-​food-​sta​mps-​ever.html. Olasky, Marvin. The Tragedy of American Compassion (Washington, DC: Regnery Publishing, 1995). Office of Family Assistance, “Q&A: The American Recovery and Reinvestment Act of 2009 (Recovery Act): Caseload Reduction Credit/​ TANF Work Participation Rates,” Washington, DC, 2010, https://​www.acf.hhs.gov/​ofa/​ faq/​qa-​ameri​can-​recov​ery-​and-​reinv​estm​ent-​act-​2009-​recov​ery-​act#_​Cas​ eloa​d_​Re​duct​ion_​Cred​it_​T​anf. Office of Family Assistance, “TANF-​ACF-​PI-​2021-​02 (The Pandemic Emergency Assistance Fund),” Washington, DC, 2021, https://​www.acf.hhs.gov/​ofa/​pol​ icy-​guida​nce/​tanf-​acf-​pi-​2021-​02. Office of Family Assistance, “TANF Caseload Data 1996–​2015,” 2016, https://​ www.acf.hhs.gov/​ofa/​data/​tanf-​casel​oad-​data-​1996-​2015 Pavetti, LaDonna “Who Is Affected by Time Limits?” in Welfare Reform: An Analysis of the Issues, edited by Isabel Sawhill, 1995, http://​web​arch​ive.urban. org/​publi​cati​ons/​306​620.html#cha​p07. Peskin, Janice, Julia Isaacs, and Alan Fairbank, “Work and Welfare: The Family Support Act of 1988,” Congressional Budget Office, Washington, DC, January 1989, https://​www.cbo.gov/​sites/​defa​ult/​files/​101st-​congr​ess-​1989-​1990/​ worki​ngpa​per/​198​9_​01​_​wor​k_​0.pdf Pianin, Eric. “Similarities, Conflicts Arise at Welfare Reform Hearings,” The Washington Post, July 28, 1994. Tiehen, Laura, Dean Jolliffe, and Timothy Smeeding, The Effect of SNAP on Poverty. University of Kentucky, 2013. Urban Institute, “A Decade of Welfare Reform: Facts and Figures,” Washington, DC, 2006, https://​www.urban.org/​sites/​defa​ult/​files/​publ​icat​ion/​51136/​900​ 980-​a-​dec​ade-​of-​welf​are-​ref​orm-​facts-​and-​figu​res.pdf. US Census Bureau, “Income and Poverty in the United States: 2019,” Washington, DC, 2020, https://​www.cen​sus.gov/​libr​ary/​publi​cati​ons/​2020/​demo/​p60-​ 270.html. United States Census Bureau, “The Supplemental Poverty Measure: 2016,” Washington, DC, 2017, https://​www.cen​sus.gov/​cont​ent/​dam/​Cen​sus/​libr​ ary/​publi​cati​ons/​2017/​demo/​p60-​261.pdf.

148

148  Work and the Social Safety Net US Congress, “Proceedings and Debates of the 104th Congress, Second Session, Vol. 142, No. 106,” Washington, DC, July 18, 1996, https://​www.congr​ess.gov/​ 104/​crec/​1996/​07/​18/​CREC-​1996-​07-​18-​sen​ate.pdf. US Congress, “S.1932: Deficit Reduction Act of 2005,” Washington, DC, 2006, https://​www.congr​ess.gov/​109/​plaws/​publ​171/​PLAW-​109​publ​171.pdf. US Congress, “To Review Outcomes of 1996 Welfare Reforms,” Washington, DC, July 19, 2006 (testimony of Ron Haskins), https://​www.govi​nfo.gov/​cont​ent/​ pkg/​CHRG-​109hh​rg30​623/​pdf/​CHRG-​109hh​rg30​623.pdf. US Department of Agriculture, Food and Nutrition Service, “SNAP Data Tables,” Washington, DC, 2021, https://​www.fns.usda.gov/​pd/​suppl​emen​tal-​nutrit​ ion-​ass​ista​nce-​prog​ram-​snap US Department of Agriculture, Food and Nutrition Service, “Supplemental Nutrition Assistance Program Participation and Costs,” Washington, DC, 2021, https://​fns-​prod.azuree​dge.net/​sites/​defa​ult/​files/​resou​rce-​files/​SNAP​ summ​ary-​8.pdf US Department of Labor, “COVID-​19 Impact,” Department of Labor Blog, Washington, DC, 2021, https://​oui.dol​eta.gov/​press/​2020/​070​920.pdf US General Accounting Office, “States Provide TANF-​Funded Work Support Services to Many Low-​Income Families Who Do Not Receive Cash Assistance,” Washington, DC, 2002, 13, https://​www.gao.gov/​ass​ets/​gao-​02-​615t.pdf. US General Accounting Office, “Welfare Reform: State Sanction Policies and Number of Families Affected,” Washington, DC, 2000, https://​www.gao.gov/​ ass​ets/​hehs-​00-​44.pdf. US Government Accountability Office, “Temporary Assistance for Needy Families: An Overview of Spending, Federal Oversight, and Program Incentives,” Washington, DC, 2015, https://​www.gao.gov/​ass​ets/​gao-​15-​ 572t.pdf. US Government Accountability Office, “Temporary Assistance for Needy Families: State Maintenance of Effort Requirements and Trends,” Washington, DC, 2012, https://​www.gao.gov/​produ​cts/​gao-​12-​713t United States House of Representatives, Ways and Means Committee, “1996 Green Book,” Washington, DC, 1996a, Section 8, table 8-​29, https://​aspe.hhs. gov/​1996-​green-​book. US House of Representatives, “Welfare and Medicaid Reform Act of 1996; Congressional Record Vol. 142, No. 106,” July 18, 1996b, https://​www.congr​ ess.gov/​congre​ssio​nal-​rec​ord/​1996/​07/​18/​house-​sect​ion/​arti​cle/​H7796-​1. The Washington Post, “Bill Clinton in 1992 Ad: ‘A Plan to End Welfare As We Know It,’ ” https://​www.was​hing​tonp​ost.com/​video/​polit​ics/​bill-​clin​ton-​in-​ 1992-​ad-​a-​plan-​to-​end-​welf​are-​as-​we-​know-​it/​2016/​08/​30/​9e635​0f8-​6ee0-​ 11e6-​993f-​73c​693a​8982​0_​vi​deo.html. Weidinger, Matt. “ ‘Child Allowances’ Revive Welfare as We Knew It,” AEIdeas, February, 2021a, https://​www.aei.org/​pove​rty-​stud​ies/​child-​all​owan​ces-​rev​ ive-​welf​are-​as-​we-​knew-​it/​.

14 9

Activation in US Social Assistance Programs  149 Weidinger, Matt. “Headed for $900 Billion in Unemployment Benefits,” AEIdeas, Washington, DC, 2021b, https://​www.aei.org/​pove​rty-​stud​ies/​hea​ded-​for-​ 900-​bill​ion-​in-​unemp​loym​ent-​benef​i ts/​ for a discussion of benefit spending, Weidinger, Matt. “While Congress Differs on Border Wall, House and Senate Bills Agree on Trimming Welfare Block Grant,” AEIdeas, 2018, https://​www. aei.org/​econom​ics/​while-​congr​ess-​diff​ers-​on-​bor​der-​wall-​house-​and-​sen​ ate-​bills-​agree-​on-​trimm​ing-​welf​are-​block-​grant/​.

150

7 FIVE DECADES OF DISABILITY BENEFIT POLICIES IN FIVE OECD COUNTRIES Duncan McVicar, Roger Wilkins, and Nicolas R. Ziebarth

INTRODUCTION This chapter describes the evolution of public disability benefit programs in five countries of the Organisation for Economic Cooperation and Development (OECD)—​the Netherlands, Sweden, Great Britain, Germany, and Australia—​ placing these programs in the context of the broader social welfare system in each country.1 We select these countries as examples of nations with similar policy goals but very different approaches to achieving them. Sweden represents the Scandinavian welfare state model and Great Britain and Australia, the Anglo-​ Saxon tradition. Germany and the Netherlands are corporate welfare states with social insurance systems based on a social partnership between employers and employees whose costs are explicitly shared. All five countries are highly developed industrialized countries that have carried out extensive reforms of their public disability programs over the past five decades. Together, they provide a heterogeneous mixture of case studies that illustrate how to implement structural program reforms while protecting the working-​age population from the economic consequences of a work-​limiting 1. We thank Jan Maarten van Sonsbeek for data on the Netherlands, Marten Palme and Lisa Laun for data on Sweden, and Dr. Lueg as well as Ms. Kühnapfel from the German Insurance Association (Gesamtverband der Deutschen Versicherungswirtschafte) for their data on private disability insurance policies and other background information on Germany. In particular, we thank Richard Burkhauser and Mary Daly for excellent input on an earlier version. Duncan McVicar, Roger Wilkins, and Nicolas R. Ziebarth, Five Decades of Disability Benefit Policies in Five OECD Countries In: Work and the Social Safety Net. Edited by: Douglas J. Besharov and Douglas M. Call, Oxford University Press. © Oxford University Press 2023. DOI: 10.1093/​oso/​9780190241599.003.0007

15 1

Five Decades of Disability Benefit Policies  151 impairment. These countries also illustrate how they managed to slow program growth and reduce disability benefit recipiency rates. We begin by comparing trends in rates of public disability benefit receipt, finding that all five countries have experienced pronounced fluctuations in their disability recipiency rates since the early 1970s. We then describe policies in each country and show correlations with country-​specific disability recipiency rates. These correlations are interpreted as suggestive of a link between policies and recipiency rates. Although our chapter is a descriptive comparative analysis based on case studies, it highlights the similarities of experiences across five OECD countries in their efforts to manage program growth and support the labor market integration of individuals with work limitations.

AN OVERVIEW OF DISABILITY RECIPIENCY RATES IN FIVE OECD COUNTRIES The number of working-​aged people receiving social insurance benefits for disability has increased substantially in most OECD nations over the past five decades. Part of this increase is simply based on a growing working-​age population. However, Figure 7.1 shows that, even when controlling for population growth by plotting the number of beneficiaries as a share of the working-​age population, the disability recipiency rates in all five countries peaked above the level first observed in 1970. In all five countries, however, it also fell below that peak in our most recent year of data. Figure 7.1 shows that all five countries have in common a pattern of initially rising recipiency rates followed by a leveling off and subsequent declines. There are, however, substantial differences in initial rates and in the timing and magnitudes of both the increases and decreases in recipiency rates. Table 7.1 summarizes these differences, showing the timing (year) and level (recipiency rate) of the initial year, the peak year, and the final year observed over the five decades of our analysis. At the beginning of the 1970s, the disability rates were, in ascending order, 1.6% in Australia, 2.7% in Great Britain, 3.1% in the Netherlands, 3.5% in Sweden, and 4.2% in Germany. As can be seen in Figure 7.1, this ordering changed over the next five decades. The Netherlands and Sweden experienced their peak levels in 2003 and 2005, respectively. Growth in the recipiency rate in the Netherlands was substantial between the mid-​1970s and mid-​1980s. After a first peak in the early nineties, in 2003, the rate peaked at 7.4% before falling over the next fifteen years to 4.8% in 2018. In contrast, Sweden’s growth was steadier and more sustained between 1970 and 2000, but was then followed by a sharp rise to 10.2% in peak year 2005 before falling substantially over the next nine years to 5.6% in 2018.

1974

1978

1982

1986

1990

1994

1998

2002

2006

2010

2014

0 2018

2

4

6

8

10

Figure 7.1  Disability recipiency rates across countries. Source: Social Security Administration, US Census Bureau, Australian Government Department of Social Services, Australian Bureau of Statistics, Department of Work and Pensions, Office for National Statistics, Social Security Administration, US Census Bureau, Statistics Sweden and Swedish Social Insurance Agency yearbooks, Statistics Netherlands, and the Institute of Employee Benefit Schemes.

1970

Australia Great Britian Netherlands Sweden Germany United States

Percentage 12

15 3

Five Decades of Disability Benefit Policies  153 Table 7.1  Disability recipiency rates for five OECD countries Australia

Great Britain Netherlands Sweden

Germany

Initial year Year

1970

1971

1973

1970

1970

DSP recipiency rate

1.6

2.7

3.1

3.5

4.2

Year

2011

2003

2003

2007

1984

DSP recipiency rate

5.4

6.7

7.4

10.2

5.8

Peak year

Final year Year

2020

2018

2018

2018

2013

DSP recipiency rate

5.2

5.8a

7.4

5.9

3.3

Sources: Australia: Department of Social Services, Australian Bureau of Statistics; Germany: Deutsche Rentenversicherung, Statistisches Bundesamt; Great Britain: Department of Work and Pensions, Office for National Statistics; Netherlands: Statistics Netherlands, Institute of Employee Benefit Schemes; Sweden: Statistics Sweden, Swedish Social Insurance Agency yearbooks. OECD SOCR database a This does not include Universal Credit recipients who, prior to 2013, would have received ESA.

As compared to the Netherlands and Sweden, Great Britain and Australia not only had lower recipiency rates in the early 1970s, but also relatively little growth until the late 1980s (Great Britain) and the early 1990s (Australia). In Great Britain, growth was then very rapid until 1996, but much less thereafter—​the rate peaked at 6.7% in 2003. Over the next fifteen years it fell back to 4.9%. For Australia, the rate of growth was noticeably higher in the 1990s than in the previous two decades, but slowed thereafter. In the peak year, 2011, the recipiency rate was 5.4% and thereafter fell, to 4.0% in 2020. While Germany’s disability recipiency rate was the highest of all these countries in the early 1970s, growing to 5.8% in peak year 1984, it fell substantially between 1984 and 1990, and again immediately following reunification in 1990. Since 1991, it has remained relatively steady and was 3.4% in 2018. Germany has had the lowest disability recipiency rates of all five countries since 1992. In what follows, we discuss the evolution of each country’s disability programs. We focus on the relationship between policy changes and recipiency rates in the context of wider social policy and economic conditions in these countries. By extending the set of countries considered and/​or the time span over which they are studied, this chapter builds on and updates cross-​country reviews such as Burkhauser et al. (2014), Burkhauser et al. (2016), and Böheim and Leoni (2016).

DISABILITY POLICIES IN FIVE OECD COUNTRIES In the industrialized world, social insurance against income losses due to work disability is just one pillar of a broader social welfare system designed to protect

154

154  Work and the Social Safety Net working-​age individuals from the loss of labor market income. Because labor market work is the primary source of income for most families, OECD nations have generally built complex social protection schemes to support individuals who are unable to earn wages in the labor market. All OECD countries provide some form of universal state-​guaranteed social insurance coverage for people who are considered long-​term work disabled. Because long-​term work disability is typically the outcome of a longer process of having health issues, this chapter also discusses closely related social insurance schemes which cover employees unable to work on health grounds before they become eligible for longer-​term disability benefits. These programs include government-​regulated or -​provided short-​term and long-​term sickness benefits and accident and medical rehabilitation programs, as well as workplace accommodation programs. One complication that arises with disability programs is the lack of a precise definition or easily verifiable marker for work disability. Work disability is not a static concept, and social conceptualizations of disability evolve over time. For example, over the past twenty years, the medical model of disability underlying categorical disability programs has been replaced by a conceptualization that recognizes the social environment as an important determinant of an individual’s ability to participate in society (World Health Organization [WHO] 2001). (There is no clear consensus on the most appropriate conceptualization of disability, although the most widely used is the World Health Organization’s [WHO] International Classification of Disability, Health, and Functioning [WHO 2001].) Under this model, work disability is a changeable state that depends on several factors, including an individual’s health impairment, the level of accommodation offered in the workplace, and the relative economic payoffs associated with working or exiting the labor force to receive disability benefits. Below, we discuss how changes in disability policies in the Netherlands, Sweden, Great Britain, Germany, and Australia have been associated with changes in disability recipiency rates in each country. Figure 7.2 (Panels A through E) shows disability recipiency rates along with major policy changes during this period for each country.

The Netherlands The disability system in the Netherlands contains both a social insurance program that protects workers against lost labor earnings (Wet op de arbeidsong eschiktheidsverzekering [WAO]/​Wet werk en inkomen naar arbeidsvermogen [WIA]) and a program that provides a social assistance minimum income for disabled adults with little or no work history (Wet arbeidsongeschiktheidsvoo rziening jonggehandicapten [“Wajong”]). A separate social minimum scheme for the disabled self-​ employed (Wet arbeidsongeschiktheidsverzekering zelfstandigen [WAZ]) was closed to new entrants from 2004. Together with sickness benefits which cover the initial part of a disability spell, the Dutch social

Replacement rates upto 80% with a 15% impairment standard

Replacement rate cut from 80% to 70% of after tax income

Courts rule that unless proven otherwise partially disabled workers are unemployed due to discrimination

Firms now responsible for first year of employee’s sick pay

Further tightening of eligibility criteria, firms now responsible for first 6 weeks of employee’s sick pay

Labor market rule abolished

New insurance scheme introduced (WIA) with various reforms

Figure 7.2  Disability recipiency rates by country. A: The Netherlands. B: Sweden. C: Great Britain. D: Germany. E: Australia. Sources: A: Statistics Netherlands, Institute of Employee Benefit Schemes, provided by Jan Maarten. Data since 2007 are taken from the OECD SOCR database, Wajong benefits are not considered in this graph. B: Statistics Sweden and Swedish Social Insurance Agency yearbooks, provided by Lisa Laun; data since 2007 from OECD SOCR database. C: British Social Security Statistics, Department of Work and Pensions, and Office for National Statistics; data since 2007 from OECD SOCR database. D: Deutsche Rentenversicherung (2014), Statistisches Bundesamt (2014); data since 2007 from OECD SOCR database. E: : Department of Social Services and Australian Bureau of Statistics.

2

3

4

5

6

6

8

9

10

Percentage

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Reforms implemented designed to increase employer cost of employee sickness

Sweeping reforms implemented focusing on work support rather than cash assistance

Reforms introduced easing return to benefits for exiters

Reforms introduced focusing on strengthening individual’s incentives to work

Figure 7.2 Continued

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Eligibility standards loosened for long term unemployed, sickness benefits replace up to 90% of earnings throughout this period

Replacement rates lowered, employer’s now pay first 14 days of sickness, pure labor market considerations removed

Sickness & disability programs merged and eligibility screening process standardized

2

4

6

8

10

12

Percentage

156

Employment and Support Allowance replaces IB for new claims: new tougher medical screening

Figure 7.2 Continued

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Universal Credit rolls out, replacing income-based ESA

ESA rolls out of existing claimants; increasing state pension age for women

Pathways to Work program rolled out: various measures introduced focused on returning disabled workers to the labor force

Unemployment Benefit replaced with Jobseeker’s Allowance: generosity of program further reduced

Invalidity Benefit replaced with Incapacity Benefit: eligibility criteria tightenined and medical screening streamlined

Supplementary Benefit replaced by Income SupportS

Restart reform: eligibility criteria tightened for unemployment benefits

0

1

2

3

4

5

6

7

8

Percentage

15 7

Eligibility standards generous

Self-employed and housewives become eligible for Statutory Pension Insurance. Disabled workers can officially retire without deductions at age 62

Legal obligation for employers to implement “workplace reintegration management”

Major structural reform: work potential assessed for all jobs, not just in recent or comparable occupation

Reduction in generosity of benefit level and introduction of max. earning limits

Standards substantially tightened Structural reforms and focus on accomodation

October 1990 Reunification

Tightening of eligibility conditions: must have paid contributions in last 3 out of 5 years

Figure 7.2 Continued

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

0

1

2

3

4

5

6

7

Percentage

158

Gradual increase in male and female retirement age

Increase in relative benefit levels due to average wage indexation

Gradual increase in female retirement age

Waiting period of 2 years introduced for most new claimants

Welfare to work reforms: Work capacity test tightened from 30 to 15 hours a week

Closure of various non-disability programs

Disability Reform Package: 85 percent permanent work incapacity eligibility requirement changed to ‘unable to work 30 hours at minimum wage for next 2 years’

50% of incapacity must now be due to impairment rather than socioeconomic factors

Screening process tightened to put greater emphasis on medical rather than socioeconomic factors

Figure 7.2 Continued

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

0

1

2

3

4

5

6

Percentage

15 9

160

160  Work and the Social Safety Net insurance program provides a comprehensive system of both partial and total disability benefits to workers, based on lost labor earnings, regardless of how or where their disability occurred. One likely reason for the rapid growth in the Dutch disability program over the 1970s was the relatively generous benefits that the system provided (Figure 7.2, Panel A). The first level of disability protection for Dutch workers was universal sickness benefits. In the 1970s, government payments from this program replaced up to 80% of net-​of-​tax wage earnings for up to one year. Moreover, most workers had the rest of their net-​of-​tax earnings replaced under collective-​bargaining agreements with their employers. Sickness benefits were payable for up to twelve months, and, after one year, employees still receiving benefits were eligible for disability benefit screening. Workers with chronic conditions that caused a reduction in their work capacity were eligible for disability benefits. Those judged fully disabled were eligible for benefits, again equal to 80% of their previous before-​tax earnings. Those judged partially disabled (with some residual earnings capacity) were eligible for partial benefits; the minimum degree of impairment for eligibility was 15%. In the mid-​1970s, in a significant loosening of access to full disability benefits, Dutch courts determined that unless disability evaluators could prove otherwise, they were required to attribute a partially disabled worker’s lack of employment to discriminatory behavior. The result was that it became “administrative practice” to treat partially disabled unemployed persons as if they were fully disabled. That interpretation of the law made assessing lost earnings capacity unnecessary beyond the minimum 15% since it became sufficient to entitle a person to full benefits. This practice essentially made the Dutch partial disability system a very generous full disability benefit system. Changes in eligibility, together with the generosity of the system, are prime candidates for explaining the rapid growth in Dutch disability benefit recipiency rates during the 1970s. The serious recession of the early 1980s and the growing costs of disability benefits put pressure on the Dutch government to reduce the growth of disability-​ based transfers. Reforms initiated between 1982 and 1987 were the first of three major efforts over the next two decades to regain control of the Dutch disability benefit program. By 1985, a series of cuts in the replacement rate effectively lowered it from 80% of before-​tax income to 70% of after-​tax income for both new entrants and existing beneficiaries. In subsequent years, system growth slowed but did not halt completely. In 1987, the labor market consideration rule was abolished. However, disability adjudicators still tended to either grant full benefits or deny any benefits. Denial rates remained quite low, suggesting that the legal change did not stop the de facto use of labor market considerations in the adjudication process. Nonetheless, the slower growth in disability recipiency rates in the second half of the 1980s brought the Netherlands more in line with disability growth in Australia, Great Britain, and Sweden over the decade. After peaking at 7.7% in 1992, disability recipiency rates began to drop. This just preceded 1994 reforms that included further tightening of eligibility criteria.

16 1

Five Decades of Disability Benefit Policies  161 Additionally, in a new policy, firms were made responsible for an employee’s first six weeks of sickness benefits. The introduction of this structural change of the system, although echoing a similar reform introduced in Great Britain in the mid-​1980s, was unprecedented in the Netherlands. It represented a deliberate change in policy intended to encourage firms to provide accommodation, rehabilitation, and continued employment opportunities to workers as an alternative to moving them on to long-​term cash benefits. The mandate that firms would bear the full responsibility for sickness benefits was extended from six weeks to one year in 1996. However, the decline in the Dutch disability recipiency rate stopped in 1997, when the rate began to slowly climb again, ending the decade slightly above where it started. In 2002, the Dutch disability system began to phase in the third and most significant set of reforms. These reforms culminated in the establishment of a new disability benefit scheme in 2004 (WIA) which, for new claimants, replaced the WAO scheme that had been in place since 1967. These systemic reforms fundamentally changed disability policy in the Netherlands. The reforms made work rather than cash benefits the expectation and enforced this by increasing the incentives of both employees and their employers to invest more time and effort in accommodation and rehabilitation following the onset of a disability. Foremost was the extension from one to two years of the mandate that firms (including small employers) bear full responsibility for employees’ sickness benefits. These changes effectively meant that, during the first two years following a health shock, workers were the responsibility of the firm and not eligible for long-​term, government-​provided disability benefits. During these two years, employers must allow workers receiving sickness benefits to remain with the firm and can only dismiss employees who refuse to cooperate with a reasonable work-​resumption plan. The reforms also gave firms a list of prescribed rehabilitation and accommodation activities that they (via a private occupational health agency) must provide to workers to assist them in remaining on their job or finding alternative employment. This new set of responsibilities on firms is known as the “Gatekeeper protocol.” After two years, workers become eligible to apply for long-​term disability benefits but have to provide documentation regarding return-​to-​work efforts during the two-​year period. In 2007, nearly 14% of disability benefit claims were returned to employers, and the employer continued to be responsible for employing the worker until the claim was processed or the worker had returned to work. Reforms at the front end of the disability process were accompanied by significant reforms in the longer-​term disability benefit program. All employers were required to pay for the fully disabled (permanent) disability benefit program through a uniform pay-​as-​you-​go premium. Employers were also required to fund the publicly run partial disability benefit program but could opt out by instead enrolling their workers with a private insurer.

162

162  Work and the Social Safety Net Either way, employers have to pay experience-​rated premiums that cover the first ten years of partial disability benefit receipt. After ten years, benefits are covered by the uniform pay-​as-​you-​go premiums. Reintegration services for disability benefit recipients were also enhanced at this time, with a move to more individually tailored packages of schooling, training, interviewing, and/​ or work placements, either purchased by the benefits agency from the private sector (“Trajectories”) or designed by the individuals themselves given a budget from the benefits agency (“Individual Reintegration Plans”). For an initial period, those finding work could also receive wage supplements. Based on these reforms, the Dutch disability benefit system, long seen as out of control, is considered by Prinz and Tompson (2009) as one that has learned from its mistakes and provides an example for other OECD countries to follow. Böheim and Leoni (2016) similarly identify the Netherlands as the country with the most extensive disability benefit reforms over the period from 1990 to 2007, along both “integration” (activation) and “compensation” (payment generosity) dimensions. A review by Koning and Lindeboom (2015) draws a similar conclusion. It is likely no coincidence that the Dutch disability recipiency rate peaked in 2003 and has been falling ever since, with a new record low at 4.8% in 2018. Indeed, the disability benefit recipiency rate is now back to the level of the early 1980s. Van Sonsbeek and Gradus (2011) presents micro-​simulation evidence on the consequences of the post-​2002 round of policy changes discussed above. They estimate that the combined impact of the introduction of experience rating, together with the introduction of the statutory Gatekeeper protocol and stricter examinations, will reduce the projected long-​term number of disability beneficiaries by 600,000. They also estimate that the introduction of the new WIA scheme will further reduce that number by 250,000 by 2040, as compared to a “no-​change scenario.” To put this in perspective, the number of disability benefit recipients in the Netherlands peaked at 1 million in 2003. Koning and Lindeboom (2015) raise the possibility that the increased responsibility borne by employers under the reformed system may reduce incentives to hire workers with discernible health conditions, or at least to hire such workers on permanent contracts given that temporary workers do not initially impact employers’ experience-​rated premiums. Although there is as yet little hard evidence of such effects, they have the potential to take some of the shine off the Dutch reforms. Furthermore, Koning and Lindeboom (2015) note that although employment rates among men with health impairments have increased following the mid-​2000 reforms, the proportion of men with health impairments who are neither working nor in receipt of disability benefits has also increased.

16 3

Five Decades of Disability Benefit Policies  163

Sweden Sweden provides sickness and disability benefits through a combination of programs. For those with an earnings history, the bulk of the protection is provided based on a social insurance program financed by statutory employer and employee contributions. Many employers in Sweden also pay into occupational-​based insurance programs on behalf of their employees. Like most European nations, Sweden additionally has a long-​standing universal needs-​based cash transfer program that provides a guaranteed social minimum income floor to all its citizens. This protection is funded out of general revenues and is available to everyone who lives or works in Sweden. Although benefits provide a minimum income to anyone in need, applicants apply for benefits based on income and particular circumstances, such as disability, parental needs, or old age. Benefits are set nationally and indexed to keep pace with the price level. As in the Netherlands, a key reason for the rapid growth in the Swedish disability program over the 1970s (Figure 7.2, Panel B) was the relatively generous benefits that the system provided. This generosity was apparent in both the ease of entry onto the program and the benefit replacement rate. The first level of protection for Swedish workers with health problems are sickness benefits. In the 1970s, sickness benefits replaced about 90% of expected earnings for individuals with “abnormal physical or mental conditions” that reduced their normal work capacity by at least 25%. Workers claiming sickness absence for more than eight days were required to obtain a certificate from a doctor. This was primarily facilitated by primary-​care physicians with no centralized screening. After one year, employees still receiving benefits could apply for long-​term disability insurance. Workers with functional limitations that reduced their work capacity were eligible for disability benefits. Benefits were awarded for partial (50%) and full disability. Those younger than sixty were also offered support for rehabilitation and vocational training. Like sickness benefits, disability benefits were very generous, replacing the vast majority of expected lost earnings. Over the course of the 1970s, standards for obtaining long-​term disability benefits were loosened to make it easier for the long-​term unemployed to move onto the program. For workers of all ages, unemployment spells of more than one year were added to the list of criteria considered in the disability screening process. For workers older than sixty, long-​term unemployment became a sufficient condition for moving onto the disability benefit rolls, even without a certifiable functional limitation. Similar to the Dutch case, these changes meant that the disability benefit program was increasingly being used as a very generous long-​term unemployment benefit program. Econometric studies of the Swedish system support this view (see Rebick 1994; Larsson 2006).

164

164  Work and the Social Safety Net Generous benefits and easier access resulted in continued steady growth in disability recipiency rates throughout the 1980s. These features also left the program vulnerable to rapid growth related to the triple-​dip recession in the early 1990s. As shown in Figure 7.2, Panel B, following the foreign exchange crisis in 1990 and ensuing deep recession, disability recipiency rates surged. Policymakers responded by lowering the replacement rates on sickness benefits, by making employers pay for the first fourteen days of sickness absence, and by removing the labor market criteria for disability benefits for older workers. With these policy changes and an improving economy, recipiency rates stabilized for most of the rest of the decade. However, they remained high and at a level that policymakers believed unsustainable. As such, policy reforms in the 1990s increased the cost to employers of worker sickness absence and tightened eligibility criteria for sickness and disability benefits. (The Swedish government made numerous changes to sickness benefit replacement rates, the number of days the employer paid for employee sickness absence, and the number of days the worker had to wait before receiving sickness benefits; see Andren 2003. In addition, policymakers removed most of the special allowances for disability insurance afforded to unemployed and older workers; see Jönsson et al. 2011.) Facing increasing fiscal pressures and a renewal of disability benefit recipiency rate growth at the end of the 1990s, the Swedish government proposed much more sweeping reforms to the sickness and disability system in 2000. Despite considerable opposition from various advocacy groups, significant reforms were put into place over the remainder of the decade. The driving principle was that work support, rather than cash assistance in lieu of work, should be the primary goal of disability policy. This general principle translated into a number of specific reforms. In 2003, the government merged the sickness benefits and disability systems and began a series of changes to standardize and enforce the administration of these now joint systems. By centralizing the screening process and developing standardized protocols for granting cash benefits, policymakers were better able to regulate the gatekeepers and enforce the strategy of promoting participation in work before offering cash benefits. As part of the merger of the sickness and disability programs, vocational and rehabilitation experts were required to be actively involved at the sickness benefit stage, which policymakers intended would stem the flow of new applicants to the long-​term disability program. To aid in this process, sickness benefits were capped at one year, and beneficiaries were evaluated for work ability at 180 days of absence. In addition, employers were required to work with disability administrators to create rehabilitation plans. Furthermore, gatekeepers were given the power to demand that employers provide certification of the steps taken to accommodate the worker. Following these reforms, the disability recipiency rate began to fall rapidly. In 2008, the Swedish government undertook an additional series of reforms to its sickness and long-​term disability programs (Hartman 2011; Organisation for

16 5

Five Decades of Disability Benefit Policies  165 Economic Cooperation and Development [OECD] 2009). These reforms were meant to further curb growth in the rolls and more actively return newly impaired workers back to the labor market. New rules strengthened the work incentives for individuals with disabilities. The principal reform was the establishment of a new timeline for the provision of rehabilitation services under the sickness absence program. Checkpoints were closely aligned with assessment of work capacity and a reduction of the cash value of sickness benefits for those who did not return to work. In addition to adding more checkpoints, the reforms also front-​loaded the evaluations to three-​, six-​, and twelve-​month increments. In essence, the earlier checkpoints moved existing rehabilitation, counseling, and assessment interventions much closer to the onset of impairment, when return to work was more likely. These reforms increased return to work of new sickness program entrants and reduced time on the program. In contrast, few of those already on the sickness program when the new reforms were initiated ever returned to work. When their sickness benefits ended, some beneficiaries moved onto other social assistance programs (Hartman 2011). These findings suggest that early intervention matters. Waiting even one year following the onset of impairment significantly reduces the chance of rehabilitation and return to work. However, further reforms in 2013 made it easier for existing beneficiaries to return to work without fear of losing their right to return to benefits. Nevertheless, the reductions in new beneficiaries were sufficient to see the disability recipiency rate continue to fall in the latter part of the decade and throughout the 2010s, including during the recession of 2008–​2009. Figure 7.2, Panel B illustrates the sharp decline in recipiency rates from 10.2% in 2007 to 5.8% in 2018. The recipiency rate is now back to where it was in the mid-​1970s.

Great Britain Böheim and Leoni (2016) identify Great Britain as the second most extensive reformer within the OECD over the period 1990 to 2007 in terms of the integration aspects of its disability benefits. Further major reforms have also been introduced more recently. Our discussion focusses on Great Britain rather than the United Kingdom (thereby excluding Northern Ireland) because the underlying data refer to Great Britain. Nevertheless, with the exception of a handful of recent welfare reforms whose introduction has been delayed in Northern Ireland, Great Britain and Northern Ireland share a common welfare system. In terms of population, the United Kingdom is dominated by Great Britain. Our conclusions are therefore likely to carry over to the United Kingdom as a whole. In 1971, Great Britain provided universal needs-​based cash transfers via its Supplementary Benefit program and somewhat higher cash transfers via its Unemployment Benefit program for those expected to work. The main social insurance program for working-​age people with disabilities between 1971 and 1995 was Invalidity Benefit (IVB). All those of working age who were deemed unable to work in their usual occupation on grounds of ill health or disability

16

166  Work and the Social Safety Net (determined largely by the claimant’s family doctor) and who had a record of sufficient social insurance contributions (paid during employment) were eligible, initially for Sickness Benefit (the first twenty-​eight weeks) and subsequently for IVB. Both Sickness Benefit and IVB recipients were counted in the IVB register. Individuals with an insufficient record of social insurance payments were eligible for “credits only” payments (the IVB system would pay their social insurance contributions) generally alongside means-​tested Supplementary Benefit payments. This group was also counted in the IVB register. IVB paid a slightly higher flat rate to beneficiaries, which was more generous than unemployment benefits for longer-​duration claimants. Some (generally older recipients with a sufficiently long work history) also received a small earnings-​related premium known as the Additional Pension. Even for those receiving the Additional Pension, replacement rates were still considerably less generous than those in the Swedish and Dutch disability systems. This regime was in place throughout the period of slow but steady growth in disability benefit recipiency rates over the 1970s and early 1980s (see Figure 7.2, Panel C). The 1980s brought several major changes that decreased IVB recipiency rates. During the early to mid-​1980s, one particular change that held back growth was the introduction of Statutory Sick Pay (SSP) in 1983, which, as in the Netherlands, made employers responsible for paying sickness benefits, in this case for the first eight weeks of a claim. The maximum duration was extended to twenty-​eight weeks, the full duration of sickness benefits, in 1986. Employees receiving SSP were not counted by the IVB register, so even if this reform did not impact behavior (which seems unlikely given the change to employer incentives) it did remove many short-​ duration claims from the roll (Anyadike-​Danes and McVicar 2008). Labor market factors, including deep recessions in the early 1980s and early 1990s, and rapid structural change away from mining and heavy industry throughout the 1980s and into the 1990s, acted in the opposite direction. Indeed, the rapid growth in disability recipiency rates during the second half of the 1980s and the first half the 1990s has been widely interpreted as a form of hidden unemployment (Beatty et al. 2000). Furthermore, a large share of IVB claims over this period were related to hard-​to-​measure mental illness or musculoskeletal conditions, despite no explicit change in the medical screening regime for IVB over this period (McVicar 2008; Banks et al. 2015). The Restart reforms of 1987–​1988 intensified both financial and “hassle-​ avoidance” incentives to shift to IVB from unemployment benefits. Restart imposed compulsory work-​focused interviews for long-​term unemployment benefits claimants, reduced the generosity of unemployment benefit payments, and introduced a requirement to show evidence of job search activity at fortnightly signing interviews. Huddleston (2000) suggests “there is clear evidence of a ‘structural break’ around 1987” in moves from unemployment to IVB (for which no such reforms had been introduced), coinciding with the in-​step increase in disability recipiency rates.

16 7

Five Decades of Disability Benefit Policies  167 Another factor likely contributing to growth in the IVB rolls occurred in 1988, when the Income Support (IS) program replaced Supplementary Benefits as the primary means-​tested social assistance payment for those out of work with insufficient work history to qualify for IVB or other insurance-​based benefits. Although there were various elements to this reform, the most relevant change was that those claiming IS on grounds of disability could now receive a higher level of payment (the Disability Premium) than those claiming IS on other grounds. Because yearly increases in IVB fixed rates (since the 1980s), unemployment, and social assistance payment benefits are tied to inflation, they have generally declined relative to real wages, hence lowering their real replacement rate even at the bottom of the wage distribution. But for those IVB beneficiaries who were also eligible for the Additional Pension, this was less the case since their benefit levels were tied to real wages. This was especially true in the recessionary years of the early 1990s, when increases in Additional Pension benefits temporarily made IVB more than twice as generous as unemployment benefits for many older recipients (Huddleston 2000). The period of rapid growth in disability recipiency rates came to an abrupt end in 1995 with a set of major reforms that ended the IVB program and replaced it with the Incapacity Benefit (IB) program for all new beneficiaries. Like other European countries, this reform attempted to slow inflows into the disability insurance system, which had been particularly high during the years of rapid growth leading up to 1995 (Anyadike-​Danes and McVicar 2008). Means-​tested beneficiaries of the IS program with disabilities continued to receive a Disability Premium and be counted as part of the IB program. But IB was both less generous than IVB (the earnings-​related Additional Pension was scrapped for new claimants and payments were made taxable), and, most importantly, the medical eligibility system was tightened. Medical screening was now carried out by government doctors rather than family doctors. This type of standardization is similar to that adopted by Sweden in 2006. The work capacity bar was also set higher with the move to assessments of claimants’ capacity to carry out any work rather than work in their usual occupation. In addition, IB’s status as an insurance payment was blurred in 1999, with the introduction of limited means-​testing for new claimants with significant (private) pension income, even those who had made sufficient contributions to be otherwise fully covered for IB benefits. There were also further reforms tightening the conditions for unemployment benefit receipt and reducing its generosity over this period, most notably the replacement of the old regime of unemployment benefits with Jobseekers’ Allowance in 1996. This tightening of unemployment benefits might, in part, explain why disability recipiency rolls began to rise again in the late 1990s and early 2000s. Disability recipiency rates only began to fall again in the mid-​2000s, coincident with a new set of work-​first reforms, called “Pathways to Work,” aimed at slowing the inflow of disability beneficiaries and boosting outflows for those having recently

168

168  Work and the Social Safety Net joined the roll. This program was piloted in 2003 and rolled out nationally beginning in 2005. It made movement onto the IB program (including receipt of social assistance on disability grounds) conditional on attendance at work-​focused interviews, with the aim of steering at least some recipients into employment support services and ultimately back into the labor market. It also introduced a “back to work” bonus payment and provided additional in-​work condition-​management health support for those returning to employment from IB. Finally, medical assessments (now relabeled “Personal Capability Assessments”) were brought forward, taking place three months into the IB claim rather than six months into the claim. Early evaluation evidence from the pilots suggested that Pathways to Work made a significant contribution to falling (local) disability rolls at the time, although the extent to which this was reflected at the national level has subsequently been questioned (Adam et al. 2010; National Audit Office 2010). Disability recipiency rates have continued to decline since then, even through the Great Recession and its aftermath, to levels last seen in the early 1990s (although note the Universal Credit–​related caveat on data since 2013, explained below). In part, this is likely to reflect the inflow-​constraining effects of the earlier reforms to disability benefits described above. But there have also been further reforms to disability benefits more recently, which are likely to have further restrained growth despite the at times difficult macroeconomic conditions. In 2008, the new Employment and Support Allowance (ESA) program replaced IB as well as IS on grounds of disability for new applicants. This new program of insurance-​based benefit for those with sufficient work history and means-​tested social assistance (income-​based) benefit for those without sufficient work history included a new, tougher Work Capability Assessment, with fewer exemptions, in place of the existing system of Personal Capability Assessments (see Sissons 2009). The requirement to attend work-​focused interviews introduced under Pathways to Work was, for all but the most severely disabled, extended into a requirement to engage in work-​related activity that was linked explicitly to payments, initially with around one-​quarter of the existing benefit payment made conditional upon compliance. The higher rate of payment for longer-​duration claims was also removed. Sissons (2009) interprets the lack of growth in disability recipiency rates over the period 2008–​2009 as evidence that they have ceased to play a role as a major destination for the hidden unemployed. In a break with the tradition of reforms largely targeted on inflows to disability rolls, and echoing similar efforts at activation of existing recipients in Sweden, rolling out from 2011 onward existing IB recipients have also been reassessed under the new ESA eligibility criteria. Many have been judged ineligible as a result of medical rescreening under the stricter Work Capability Assessment, although some have since successfully appealed these decisions. Banks et al. (2015) provides descriptive evidence that suggests the introduction of ESA and the initial period of its roll-​out to existing claimants most likely led to a fall in disability recipiency rates, albeit one that has been partly masked by demographic changes

16 9

Five Decades of Disability Benefit Policies  169 (specifically, the aging of the baby boomers). (Increases in the state pension age for women, which began to roll out in 2010, have also kept disability recipiency rates higher than they would otherwise have been over this period.) They also argue, however, that there appears to have been little significant shift into employment or, in partial contrast with Sweden, onto alternative (unemployment) benefits, which begs the question of what has happened to those previously but no longer receiving disability benefits. Subsequent changes to ESA have included the 2012 introduction of a one-​year time limit on the insurance-​based element for all but the most severely disabled, after which recipients shift to the more heavily means-​tested social assistance (income-​based) ESA payment, if they are eligible. Finally, from 2013, a new benefit called Universal Credit has gradually started to replace new claims for income-​ based (social assistance) ESA and will in time replace ESA for many existing claimants (the roll-​out is currently expected to be complete by 2024). Because Universal Credit is replacing not just income-​based ESA but six legacy benefits including Jobseeker’s Allowance, and because data are not published separately for Universal Credit recipients who would otherwise have received ESA, one implication of this is that it is now more difficult to count income-​replacement disability benefit recipients in Britain. It is therefore likely that Figures 7.1 and 7.2, Panel C increasingly undercount disability benefit recipients in Britain from 2013 onward. This blurring of the distinction between unemployment benefit recipients and disability benefit recipients for all but the most severely disabled is also reflected in the potential for further tightening of work preparation and even work search requirements for some Universal Credit claimants with disability.

Germany German employees are eligible for both short-​and long-​term statutory sickness insurance benefits. (Germany also has a separate Statutory Accident Insurance [SAI] program covering temporary and permanent work absences in case of work accidents or diseases.) Employers are required to provide short-​term sickness benefits of 100% of the wage for up to six weeks (Ziebarth and Karlsson 2010, 2014). Workers with longer spells are reevaluated for access to long-​term sickness benefits. Long-​term sick leave benefits are funded by the health insurance benefit package. Statutory Health Insurance, which covers 90% of the population, replaces 70% of net wages and can be paid for up to seventy-​eight weeks (see Ziebarth 2009, 2013). In Germany, the statutory Old-​Age Pension Scheme (OAP) and the Work Disability Pension (WDP) for both partially and totally disabled workers are integrated in the pension insurance pillar of the social insurance system (Deutsche Rentenversicherung 2020b). Both programs pay benefits to workers who have paid contributions into the systems during their work life. Employers and employees are each subject to a payroll tax (9.35%) of the monthly gross wage

170

170  Work and the Social Safety Net up to the social contribution ceiling. At the end of 2019, total WDP benefits per month were €1.2 billion, or 8% of total spending by the SPI (Deutsche Rentenversicherung, 2020a).2 As shown in Figure 7.1, in the early 1970s, Germany had the highest recipiency rate of any of the OECD countries in our comparison. One reason for this high rate was a change in WDP rules in 1969 that allowed partially disabled workers to receive full WDP benefits if they were unable to find a job (Burkhauser and Hirvonen 1989). Further expansions in 1972 extended coverage to housewives and the self-​employed, and allowed disabled workers to transition to the retirement program at age sixty-​two without an actuarial reduction in benefits. As seen in Figure 7.2, Panel D, in the aftermath of the reforms, disability benefit recipiency rose significantly, peaking at 5.8% in 1984. This rapid growth in the recipiency rate led to a substantial tightening of WDP eligibility criteria. WDP reforms in the early 1980s limited coverage to workers who had paid payroll taxes in at least three of the past five years and had accumulated at least five years of market work experience. Because many housewives did not meet these eligibility work criteria, the reforms greatly curtailed their WDP coverage. Consequently, a large fraction of the decline in recipiency rates during this period was a result of the reduction in access for women working outside the paid labor market (see Robert Koch Institut [RKI] 2006; Börsch-​ Supan and Jürges 2012 for a more detailed discussion). These reforms turned growth in disability recipiency negative, more than offsetting the increases experienced in previous decades. Germany undertook additional policy reforms in the 1990s and 2000s. In 1996, actuarial reductions and caps on the earnings of WDP beneficiaries were introduced. Börsch-​Supan and Jürges (2012) argue that these caps reduced the inflow of males onto the WDP. The data show that the number of new male beneficiaries fell from about 150,000 per year prior to the reforms to just 75,000 per year after the reform. Figure 7.2, Panel D shows that this large reduction in the inflow of new male beneficiaries onto the program put steady downward pressure on the disability recipiency rate over the rest of the decade. (Note that the figures reflect the stock of all beneficiaries. As such, even large declines in the flow of new beneficiaries affect the overall disability recipiency rate only gradually.) In 2001, another round of structural WDP reforms became effective. Most important among the reforms was a change in the eligibility standard from being unable to work in the occupation in which one was trained, effectively in the last job or a comparable job in terms of the skills it required, the wages it paid,

2. The figure of €1.165 billion (14 billion per year) is based on 72,301 partial WDI beneficiaries with average monthly cash benefit of €553 and 1,415,295 full WDI beneficiaries with average benefits of €842 (Deutsche Rentenversicherung 2020a). Note that WDP benefits are converted to an old-​ age pension of the same amount when reaching the statutory retirement age. Expenditures are calculated for WDI beneficiaries younger than the statutory retirement age.

17 1

Five Decades of Disability Benefit Policies  171 and its prestige, to being unable to work in any job. In other words, public occupational disability insurance was abolished and, henceforth, only applicants who were not able to work six (three) hours per day in any occupation would be eligible for partial (full) DI benefits. In addition, the benefit level was cut by 2.7% for those younger than fifty-​five, and for those between fifty-​five and sixty years even more to disincentivize using the disability program as an alternative to early retirement. As a result, total inflows (men and women combined) onto the WDP program decreased further. This slow but steady decline in new beneficiaries put additional downward pressure on the overall disability recipiency rate (Krause et al. 2013). Geyer and Ziebarth (2022) comprehensively evaluate this reform in a causal effects framework and find average inflow rate decreases of 15% per year and cohort since 2001 (where the decrease for men was twice as large than for women.) From 2010 to 2017, the decrease in inflows stabilized with a long-​term reform effect of minues 35% compared to the pre-​2001 era. WDP reforms in 2004 continued to focus on reducing the inflow of new recipients onto the program. However, the attention of these reductions shifted away from tightening WDP eligibility requirements and toward promoting worker accommodation on the job. Specifically, the reforms made it a legal obligation of employers to provide workplace reintegration in the event of a work-​limiting impairment. Indeed, the law mandates that when impaired workers have exhausted their short-​term sickness benefits of six weeks and are being considered for long-​ term sickness benefits, employers must act to coordinate a plan that includes input from the sick-​listed employee, WDP experts, the work council, and the workplace physician. The plan is meant to ensure that the worker’s temporary disability can be overcome and to prevent future reductions in work capacity. Another point is worth mentioning in this context. The reduced DI benefit level increased the share of DI recipients who also apply for means-​tested social assistance. In 2019, only 2.6% of all pensioners received social assistance but among people with DI benefits, the rate was nearly 15% (Deutsche Rentenversicherung 2020c). A series of studies using different data sources documents this increasing risk of poverty (Krause et al. 2013; Märtin et al., 2012, 2014; Geyer, 2022). As a consequence, in a series of reforms after 2014, policymakers responded to this development and structurally increased the benefit level of DI benefit stepwise through 2031. Between 2018 and 2019, benefit levels of new DI recipients increased by about 10 percent. Generally, the experience of Germany over the past five decades highlights the role that policy decisions play in the dynamics of disability recipiency rates. Pre-​1970 policies meant that German disability recipiency rates were higher in 1970 than in the other countries observed in this study. The expansion of coverage and generosity resulted in another round of rapid increases in recipiency rates in the late 1970s and early 1980s. Subsequent reforms limited access, made benefits less attractive, and mandated that employers implement a workplace

172

172  Work and the Social Safety Net reintegration management. These reforms significantly altered the trajectory of the WDP program. The movement of the German public disability support system toward a limited access cash transfer program c growth in the market for private occupational disability insurance (PODI). PODI provides benefits to workers who have a reduced work capacity in their current (or a comparable) occupation, a less stringent level of work incapacity than the statutory WDP now provides. Today, among households whose household head is an employee, 38% hold an individual private DI policy (Statistisches Bundesamt 2018). PODI plans are experience-​rated and individually underwritten—​premiums depend on age, medical diagnoses, and occupation. As a result, premiums can be unaffordable for high-​risk occupations and applicants may be denied coverage. Private disability insurance follows private insurance law and is based on a private contract between the insurer and the insured, which specifies the conditions for the insured risk individually. Importantly, since WDP is not means-​tested, private PODI benefits can be received in addition to WDP benefits, which prevents a crowd-​out of private policies. Private work disability insurance (PWDI) also exists in Germany. It provides benefits in case of general work disability and is similar to WDP in terms of disability eligibility criteria. While, in principal, these benefits could also be received while receiving WDP benefits, there is a much smaller private market for this type of insurance. Figure 7.3 plots the total number of new private disability insurance policies (PODI and PWDI) in Germany from 1976 to 2013 (DGV 2015). The graph shows the dramatic increase in new policies between 1997 and 2001, the years of the two latest WDP reforms. While 44,000 new contracts were signed in 1997, this number quadrupled to 184,000 in 2000, and then again more than doubled to 473,000 in 2001. Since 2001, the numbers have stabilized at this high level. It is plausible to conclude that this uptick in private market coverage has been the result of a decline in the coverage and generosity of the public WDP program. In fact, Geyer and Ziebarth (2022) analyze this question in a causal effects framework but find only very limited evidence that the uptick in Figure 7.3 is specifically driven by the “notch” cohorts for whom public occupational disability insurance was abolished. (As seen in Figure 7.3, demand already started to increase substantially around 1996 when a first structural pension reform was implemented.) Although there are multiple possible explanations for this finding (including the affected cohorts not valuing occupational disability insurance, not being aware of the 2001 reforms, and market regulation), Geyer and Ziebarth (2002) find that the explanation with the most compelling evidence is that demand for private insurance for the elderly has increased because of structural reforms that cut the generosity of social insurance and aggressive promotion by private insurers and policymakers. Because private disability insurance is not guaranteed issue and individually underwritten in Germany, a substantial share of the population is unable to obtain

17 3

Five Decades of Disability Benefit Policies  173

Figure 7.3  Development of new private disability insurance policies in Germany. Source: Association of German Insurers (DGV).

coverage. Others are liquidity constrained and yet others do not think it would be necessary. Geyer and Ziebarth (2022) find that sick people are significantly less likely to have private coverage. Those who believe that they will die younger than the average person in their age-​gender group are also significantly less likely to have coverage, as are people who live paycheck-​to-​paycheck and those who say that they want to enjoy life now. When categorizing those without private ODI coverage using k-​means clustering, Geyer and Ziebarth (2022) find that a first group cannot obtain coverage due to sickness and low income, despite their young age. A second group is relatively wealthy and healthy but close to retirement and thus has no incentive to purchase coverage. A third group is between forty and fifty years of age and could be a target group for policies to further increase private coverage. However, more than a quarter of those say they have no means to save, and others are in bad health. This analysis illustrates little potential to increase private coverage much further without structural regulatory reforms.

Australia Australia’s disability income support program, the Disability Support Pension (DSP), is a universal “flat rate” benefit available to all people aged sixteen and older who meet specified incapacity criteria, regardless of employment history. There is no requirement to have been employed or in any way “paid for coverage,” and benefits payable do not depend on workers’ past earnings. It is most similar to a means-​tested guaranteed minimum income program whose benefits are greater than those provided by other Australian welfare or unemployment benefit programs. Hence, DSP is closer in concept to a disability-​based welfare program

174

174  Work and the Social Safety Net than a traditional social insurance program. However, for purposes of this analysis, we considered beneficiaries of this program as receiving public disability benefits. While benefits are substantially less generous than those of Germany, the Netherlands, and Sweden, the DSP program is nonetheless susceptible to the same risks to growth since it is based on the malleable concept of disability and pays benefits higher than those of other welfare benefits. Indeed, one of the features of the DSP benefit level over the 1970–​2020 period is its growth relative to other benefits, particularly since the mid-​1990s. Most welfare benefits have been indexed to inflation, whereas the DSP has been indexed to a measure of average wages, which has grown substantially in real terms since the mid-​1990s. In 1996, the DSP benefit was only 8% higher than the unemployment benefit, but by 2020 was 52% higher. Unsurprisingly, given its indexation to wages over much of the 1970–​2020 period, there has been little or no trend increase in the level of the DSP benefit relative to average weekly earnings. This would suggest that increases in the replacement rate are not an important explanation for DSP growth. Moreover, over the first decade of the 2000s, the income “disregard” (the maximum market income a recipient may receive before the benefit begins reducing) increased more slowly than average earnings, corresponding to a “tightening” of eligibility criteria for DSP and coinciding with a flattening out of the cumulative DSP recipiency rates over the period (McVicar and Wilkins 2013). The DSP has, however, become more generous relative to earnings at the lower end of the earnings distribution, at least over the period since 1993, and this may be a more relevant comparison for low-​skilled workers with disabilities. For example, between 1994 and 2018, adult full-​time weekly earnings at the 10th percentile increased by a factor of 2.4, whereas the maximum DSP payment has increased by a factor of approximately 2.9 (Australian Bureau of Statistics [ABS] 1994, 2019). Similarly, the level of DSP payments has increased substantially relative to the level of the Australian minimum wage over the period from 2000 to 2020, and particularly from 2008 to 2020, when the maximum DSP payment increased by 66% and the minimum wage increased by 39%. Nonetheless, it is likely that the major changes in DSP recipiency rates shown in Figure 7.2, Panel E, are driven by changes in macroeconomic conditions and disability eligibility criteria rather than the relative generosity of benefits. This view is supported by work by Cai and Gregory (2004), who suggest that the small drop in DSP recipiency rate between 1980 and 1982 was the result of a tightening of eligibility rules by the administrative authority, which began placing greater emphasis on medical factors and less weight on socioeconomic factors. This change in approach, however, was largely reversed in 1983, when the Labor Government came to power. The most important increase in the DSP recipiency rate occurred in the 1990s (Figure 7.2, Panel E). Prior to the arrival of the COVID-​19 pandemic, Australia experienced its last official and most serious recession in the early 1990s. In 1991, in the midst of this recession, there was a major change in the DSP eligibility criteria. To be eligible for DSP prior to the 1991 reform, a worker had to be incapable of working at more than 15% capacity. In 1991, this was replaced with a

17 5

Five Decades of Disability Benefit Policies  175 requirement that the claimant be incapable of working thirty or more hours per week. If a forty-​hour work week is considered normal, this effectively replaced an 85% impairment standard with a 25% impairment standard, changing DSP from a long-​term total disability program to a long-​term partial disability program with no reduction in benefits. Hence, it required DSP gatekeepers to decide if an unemployed worker with a partial (as low as 25%) disability was unemployed because of his or her disability or because of economic conditions in the midst of a period of slow economic growth and high unemployment rates. It is not surprising that DSP recipiency rates increased substantially over this period. While DSP benefits were then not significantly greater than first-​tier social minimum benefits, DSP provided a more permanent income floor with no work requirement. (There is no tier-​two unemployment insurance benefit program in Australia; rather, unemployed workers are covered by a tier-​one universal minimum benefit currently called the JobSeeker Payment. Benefit levels are needs-​based and do not require past work experience. However, recipients are expected to return to work.) As Cai et al. (2007) show, few entrants to DSP leave the program to return to work, so that the increase in inflows of beneficiaries during the recession led to increases in disability recipiency rates that lasted over many additional years. As the Australian economy recovered and then expanded in the 1990s, growth in DSP recipiency rates slowed, although there was no decline in the rate of receipt until the 2005–​2007 period, when economic growth was exceptionally strong on the back of the mining boom (Figure 7.2, Panel E). Australia did not experience a recession between 1991 and 2020, but did experience small rises in unemployment in each of the two worldwide recessions that have occurred over this period. The rate of growth in DSP recipiency rates increased temporarily in the wake of the 2001 worldwide recession. Following the onset of the global financial crisis in 2008, there was first a leveling off in DSP receipt between 2007 and 2009 (thus arresting the decline between 2005 and 2007), followed by a sharp rise in DSP receipt between 2009 and 2011. The DSP recipiency rate then fell substantially over the next nine years, in 2020 reaching its lowest level since 1995. Cai and Gregory (2004) and McVicar and Wilkins (2013) also argue that reforms to non-​DSP welfare payments over the 1990s and 2000s had unintended consequences for DSP receipt. Certain types of payments, such as for temporary sickness, were discontinued in the 1990s, and, more importantly, over the 1990s and 2000s, welfare benefits for the unemployed and for lone parents became increasingly conditional on verifiable job search and participation in active labor market programs, reducing the relative attractiveness of these benefits. Consistent with these changes, and also with the increased relative generosity of DSP, McVicar and Wilkins (2013) show that over the period from 1993 to 2011, receipt of non-​DSP welfare benefits by people with disability declined appreciably but that this was largely offset by the increase in DSP receipt. Thus, welfare reforms that made the benefits for unemployment and lone parents less generous were, to some extent, thwarted by the shift of significant numbers of beneficiaries from these non-​DSP welfare programs to the DSP.

176

176  Work and the Social Safety Net The increased incentives for unemployed low-​skilled applicants to apply for disability benefits puts additional pressure on DSP gatekeepers to only admit those who are unemployed because their impairment reaches the DSP standard. However, because the relationship between impairment and disability is mutable, this is hard to do. Thus, in economic downturns, the increased pool of potentially eligible unemployed workers is likely to result in program growth. Australian governments have been acutely aware of these pressures and have therefore implemented a string of welfare reforms since 1999 that have consistently moved to tighten eligibility requirements for DSP. Initially, these reforms primarily targeted new claimants, with most existing recipients grandfathered under existing rules. However, more recently, reforms targeting existing recipients have begun to be introduced. The two most significant reforms affecting new claimants were in July 2006 and September 2011. The 2006 reform restricted DSP eligibility to claimants with a work capacity of less than fifteen hours per week, rather than thirty hours per week, effectively tightening program eligibility rules from a 25% to a 62.5% impairment standard. The 2011 reform involved implementation of a new DSP assessment procedure which, among other things, effectively introduced a two-​ year waiting period prior to DSP eligibility. In July 2014, requirements to participate in education, training, or work experience programs were introduced for DSP recipients younger than thirty-​five assessed as capable of working at least eight hours per week. The July 2014 changes also included reassessment, under the new assessment criteria introduced in September 2011, of DSP eligibility for all recipients aged younger than thirty-​five who were granted the benefit after 2007 (and before September 2011). The 2006 reform may have had some role in mitigating the rise in the relative generosity of DSP program benefits, but it seems more likely that the strong economic growth prevailing at the time was responsible for arresting DSP growth in the 2000s. This is because the turn-​around in DSP receipt predates the reform by several years and, moreover, DSP receipt again grew strongly between 2009 and 2011. However, the combined effects of the 2006 and 2011 reforms seem to have been an important contributor to the decline in DSP receipt that occurred between 2011 and 2020, particularly when viewed in light of the relative weakness in economic growth over this period. The 2014 reforms appear to have had only minimal impacts on aggregate DSP receipt. Indeed, the proportion of working-​age DSP recipients younger than thirty-​five rose slightly between 2014 and 2020, from 17.6% to 19.2%.

CONCLUSION Over the five decades of our study, all five of the countries examined have experienced increases in their disability recipiency rates, followed by plateauing and eventual declines. This pattern reflects a commonality in the evolution of

17 7

Five Decades of Disability Benefit Policies  177 their disability benefit policies—​periods of expanding eligibility or benefits were followed by rising recipiency rates. These rising rates, in turn, triggered policy reforms which tightened eligibility standards again, slowing down the growth of receipt and, in four of the five countries, also reducing the level of receipt. Although focusing on a narrower time span from 1990 to 2007, Böheim and Leoni (2016) suggest that this trend toward tighter eligibility criteria is common to many other OECD countries. Naturally, policy reforms tend not to impact on disability recipiency rates rapidly, uniformly, or in isolation of the broader social policy and economic context (and vice versa). Changes in disability recipiency rates often considerably lag behind policy changes. Most policy reforms target inflows onto disability benefit programs and, given that benefit spells are typically of long duration, changes in inflows take time to work through into material changes in stocks. Inflows are also sensitive to economic conditions. In contrast, disability outflows are relatively insensitive to economic conditions. The net result is that the effects of policy changes of the “loosening” variety were mostly felt when economic conditions deteriorated and unemployment rose. In contrast, the effects of policy changes of the “tightening” variety tended to appear only gradually over time as accumulated effects of reduced inflows slowly translated into reductions in the rate of growth of benefit receipt. Nevertheless, the effects of “tightening” policies in reducing benefit receipt are intuitively greatest in recessions when, absent the tightening, inflows would otherwise be much higher. Indeed, one interpretation of the policy tightening is that recessions have been rendered less important determinants of benefit receipt. In particular, the late 2000s economic downturn and its aftermath did not produce a sharp rise in recipiency rates in any of the five countries examined here. Despite the common pattern described in this chapter, we found substantial differences in the initial levels of disability benefit receipt across our five countries at the beginning of the 1970s, and in the dynamics over the four subsequent decades. This variation reflects differences in the timing and nature of both “loosening” and “tightening” policies, together with differences in the timing and severity of economic downturns and other contextual factors. At one end of the spectrum is Germany, which tightened its policies much sooner than any of the other countries. The Netherlands and Sweden both sustained growth and high levels of disability benefit receipt for much longer than the other countries before introducing policies that were effective in reducing inflows. Great Britain experienced relatively late growth and was relatively quick to introduce policies that arrested this growth. At the other end of the spectrum, Australia experienced most of its program growth in the 1990s and only experienced a significant decrease in disability benefit receipt from 2011 onward. Finally, despite the extensive reforms discussed in this chapter, all countries (other than Germany) now have substantially higher disability recipiency rates than at the beginning of the 1970s. To the degree that the secular growth in rates

178

178  Work and the Social Safety Net has been policy driven, this means that workers with disabilities were not employed who might in fact have been employed in the absence of these policies. That is, in addition to concerns about the fiscal sustainability, disability programs themselves can contribute to lower labor force participation and lower employment rates among people with disability (Autor and Duggan 2003; Staubli 2011; Maestas et al. 2013; Kostøl and Mogstad 2014; Moore 2015; Haller et al. 2020; Geyer and Ziebarth 2022). Of course, for individuals with work limitations for whom even part-​time work may simply not be possible, disability systems must provide effective support. But the recent experience of the countries reviewed here suggests that it is possible to change the culture and social expectations of and for individuals with work limitations, and hence for disability recipiency rates to go down as well as up.

APPENDIX Table 7.A1  Data description and sources Summary of DI data availability across countries Australia

Great Britain Netherlands Sweden

Germany

Initial year

1970

1971

1973

1970

1970

Final year

2020

2018

2018

2018

2018

Missing years

(–​)

(–​)

(–​)

1984

gaps until ‘00

Age range of working population

16–​64

16–​64

15–​65

16–​64

16–​64

Data Sources Australia: Historical population estimates are from the Australian Bureau of Statistics www.abs.gov.au. DSP caseload data are from Australian Government Department of Social Services Statistical Papers 1–​12 and DSS Payment Demographic Data: https://​data.gov.au/​data/​data​set/​cff2a​e8a-​55e4-​47db-​a66d-​e177f​e0ac​6a0. Great Britain: Historical population estimates are mid-​year population estimates from the Office for National Statistics http://​www.ons.gov.uk. Disability benefit caseload data (combining IVB, IB and ESA caseloads) are from Social Security Statistics and, from 1999 onward, from the Department of Work and Pensions, http://​dwp.gov.uk. Netherlands: Historical population data are from Statistics Netherlands http://​www.cbs.nl/​en-​GB/​menu/​home/​ defa​ult.htm. Disability insurance caseloads data are from the Institute of Employee Benefit Schemes, courtesy of Jan Maarten van Sonsbeek. Sweden: Historical population estimates are from Statistics Sweden http://​scb.se. Disability Insurance prevalence data are from the Social Insurance Agency yearbooks, courtesy of Lisa Laun and Marten Palme. Germany: Deutsche Rentenversicherung (2014) on absolute numbers of WDP beneficiaries. Statistik der Deutschen Rentenversicherung (2014): “Rentenversicherung in Zeitreihen 2014,” http://​forsch​ung.deuts​che-​ren​ tenv​ersi​cher​ung.de, and upon request. Statistisches Bundesamt (2014) on population between fifteen and sixty-​five, unemployment rates, and people out of the labor force https://​www-​gene​sis.desta​tis.de. Since 2007: Except for Australia, all data are taken from the OECD SOCR database,https://​www.oecd.org/​els/​soc/​rec​ipie​nts-​ socr-​by-​coun​try.htm#Trend.

17 9

Five Decades of Disability Benefit Policies  179

REFERENCES Adam, S., Bozio, A., and C. Emerson. Reforming Disability Insurance in the UK: Evaluation of the Pathways to Work Programme. London: Institute for Fiscal Studies, 2010. Andren, D. Sickness-​Related Absenteeism and Economic Incentives in Sweden: A History of Reforms. CESifo DICE Report 1, no. 3 (2003). Anyadike-​Danes, M., and D. McVicar. Has the Boom in Incapacity Benefit Claimant Numbers Passed its Peak? Fiscal Studies 29, no. 4 (2008): 415–​434. Australian Bureau of Statistics (ABS). Distribution and Composition of Employee Earnings and Hours. Catalogue No. 6306.0. Canberra: ABS, 1994. Australian Bureau of Statistics (ABS). Employee Earnings and Hours. Cat. No. 6306.0. Canberra: ABS, 2019. Autor, D. H., and M. G. Duggan. The Rise in Disability Recipiency and the Decline in Unemployment. Quarterly Journal of Economics 118, no. 1 (2003): 157–​205. Banks, J., R. Blundell, and C. Emmerson. Disability Benefit Receipt and Reform: Reconciling Trends in the United Kingdom. Journal of Economic Perspective 29, no. 2 (2015): 173–​190. Beatty, C., S. Fothergill, and R. Macmillan. A Theory of Employment, Unemployment and Sickness. Regional Studies 34 (2000): 617–​630. Böheim, R., and T. Leoni. Disability Policies: Reform Strategies in Comparative Perspective. NBER Working Paper No. 22206. Cambridge, MA: National Bureau of Economic Research (NBER), 2016. Börsch-​ Supan, A., and H. Jürges. Disability, Pension Reform, and Early Retirement in Germany. In Social Security Programs and Retirement around the World: Historical Trends in Mortality and Health, Employment, and Disability Insurance Participation, ed. D. H. Wise. Chicago: University of Chicago Press, 2012, 277–​300. Burkhauser, R. V., M. Daly, D. McVicar, and R. Wilkins. Disability Benefit Growth and Disability Reform in the US: Lessons from Other OECD Nations. IZA Journal of Labor Policy 3, no. 4 (2014): 1–​30. doi:10.1186/​2193-​9004-​3-​4. Burkhauser, R. V., M. Daly, and N. R. Ziebarth. Protecting Working-​Age People with Disabilities: Experiences of Four Industrialized Nations. Journal of Labour Market Research (Zeitschrift fuer Arbeitsmarktforschung) 49 (2016): 367–​386. Burkhauser, R. V., and P. Hirvonen. United States Disability Policy in a Time of Economic Crisis: A Comparison with Sweden and the Federal Republic of Germany. Milbank Quarterly 67, no. 2 (1989): 166–​194. Cai, L., and R. Gregory. Inflows, Outflows and the Growth of the Disability Support Pension (DSP) Program. Australian Social Policy 2002–​03 (2004): 121–​143. Cai, L., H. Vu, and R. Wilkins. Disability Support Pension Recipients: Who Gets Off (and Stays Off) Payments? Australian Economic Review 40, no. 1 (2007): 37–​61.

180

180  Work and the Social Safety Net DGV (Association of German Insurers). Statistical Yearbooks of German Insurance 2014. Berlin: Association of German Insurers, 2015. http://​www. en.gdv.de/​2014/​11/​prope​rty-​casua​lty-​insura​nce-​with-​rec​ord-​cla​ims-​expe​ ndit​ure/​. Deutsche Rentenversicherung. Rentenversicherung in Zahlen 2020. 2020a. https://​ w ww.deuts ​ c he- ​ ren​ t env​ e rsi ​ c her ​ u ng.de/ ​ DRV/ ​ D E/ ​ E xper ​ t en/ ​ Z ah​ len-​und-​Fak​ten/​Stat​isti​ken-​und-​B eric​hte/​stat​isti​ken-​und-​berich​te_​n​ode. html?https=​1, last accessed on June 11, 2021. Deutsche Rentenversicherung. Erwerbsminderungsrente: Das Netz für alle Fälle, 15th edition., 2020b. Deutsche Rentenversicherung. 2020c. Rentenversicherung in Zeitreihen, DRV-​ Schriften, 22nd edition, 2020c. Geyer, J., and N. R. Ziebarth. Fundamentally Reforming the Public DI System: Evidence from German Notch Cohorts, mimeo, 2022. Geyer, J. Der Einfluss von Rentenreformen auf Zugänge und Zahlbeträge in Erwerbsminderungsrenten –​Modell-​rechnungen bis 2050: Forschungsbericht; Forschungs-​projekt gefördert durch die Hans-​Böckler-​Stiftung. 164. DIW Berlin: Politikberatung Kompakt. Politikberatung Kompakt. DIW Berlin, German Institute for Economic Research, 2021. Haller A., Staubli, S., and J. Zweimueller. Designing Disability Insurance Reforms: Tightening Eligibility Rules or Reducing Benefits? NBER Working Paper No. 27602. Cambridge, MA: National Bureau of Economic Research (NBER), 2020. Hartman, L. Swedish Policies for the Disabled: What Do We Know About the Effects? Paper presented at the Labour Activation in a Time of High Unemployment conference, Paris, November 13–​15, 2011. http://​umdc​ipe.org/​conf​eren​ces/​ Labor​Acti​vati​onPa​ris/​Pap​ers/​Laura%20He​artm​ans%20dr​aft%20pa​per_​11_​1​ 108.pdf. Huddleston, T. Explaining the Growth in the Number of People Claiming Incapacity Benefits. London: Department for Work and Pensions, 2000. Jönsson, L., M. Palme, and I. Svensson. Disability Insurance, Population Health and Employment in Sweden. NBER Working Paper No. 17054. Cambridge, MA: National Bureau of Economic Research (NBER), 2011. Koning, P., and M. Lindeboom. The Rise and Fall of Disability Insurance Enrollment in the Netherlands. Journal of Economic Perspectives 29, no. 2 (2015): 151–​172. Kostøl, A. R., and M. Mogstad. How Financial Incentives Induce Disability Insurance Recipients to Return to Work. American Economic Review 104, no. 2 (2014): 624–​655. Krause, P., U. Ehrlichand, and K. Möhring. Erwerbsminderungsrentner: Sinkende Leistungen und Wachsende Einkommensunterschiede im Alter. DIW Wochenbericht 80, no. 24 (2013): 3–​9.

18 1

Five Decades of Disability Benefit Policies  181 Larsson, L. Sick of Being Unemployed? Interactions Between Unemployment and Sickness Insurance in Sweden. Scandinavian Journal of Economics 108, no. 1 (2006): 97–​113. Maestas, N., K. J. Mullen, and A. Strand. Does Disability Insurance Receipt Discourage Work? Using Examiner Assignment to Estimate Causal Effects of SSDI Receipt. American Economics Review 103, no. 5 (2013): 1797–​1829. Märtin, S., Zollmann, P., and R. Buschmann-​ Steinhage. Sozioökonomische Situation von Personen mit Erwerbs-​minderung: Projektbericht I zur Studie. DRV Schriften 99. Berlin: Deutsche Rentenversicherung Bund, 2012. Märtin, S., Zollmann, P., and R. Buschmann-​Steinhage “Sozioökonomische Situation von Personen mit Erwerbsminderung: Projektbericht II zur Studie.” DRV Schriften 105. Berlin: Deutsche Rentenversicherung Bund, 2014. McVicar, D. Why Have UK Disability Benefit Rolls Grown So Much? Journal of Economic Surveys 22 (2008): 114–​139. McVicar, D., and R. Wilkins. Explaining the Growth in the Number of Recipients of the Disability Support Pension in Australia. Australian Economic Review 46, no. 3 (2013): 345–​356. Moore, T. J. The Employment Effects of Terminating Disability Benefits. Journal of Public Economics 124, no. C (2015): 30–​43. National Audit Office. Support to Incapacity Benefits Claimants Through Pathways to Work. London, UK: National Audit Office, 2010. Organisation for Economic Cooperation and Development (OECD). Sickness, Disability and Work: Keeping on Track in the Economic Downturn. Paris: OECD Publishing, 2009. www.oecd.org/​datao​ecd/​42/​15/​42699​911.pdf. Prinz, Christopher, and William Tompson. Sickness and Disability Benefit Programs: What Is Driving Policy Convergence. International Social Security Review 62 (October/​December 2009): 41–​61. Rebick, M. Social Security and Older Workers’ Labor Market Responsiveness: The United States, Japan and Sweden. In Social Protection versus Economic Flexibility: Is There a Trade-​Off?, ed. R. Blank. Chicago: University of Chicago Press, 1994:189–​221. Robert Koch Institut (RKI). Gesundheitsbedingte Frühberentung, Gesundheitsbericht-​erstattung des Bundes, Heft 30, Statistisches Bundesamt, 2006. http://​www.rki.de/​DE/​Cont​ent/​Gesund​heit​smon​itor​ing/​Gesun​dhei​tsbe​ rich​ters​tatt​ung/​Them​enhe​fte/​fruehb​eren​tung​_​inh​alt.html. Sissons, P. Welfare Reform and Recession: Past Labour Market Responses to Job Losses and the Potential Impact of Employment Support Allowance. People, Place & Poverty 3, no. 3 (2009): 171–​182. Statistisches Bundesamt. Wirtschaftsrechnungen Einkommens-​ und Verbrauchsstichprobe Ausstattung privater Haushalte mit ausgewählten Gebrauchsgütern und Versicherungen, Fachserie 15 Heft 1, 2018. Staubli, S. The Impact of Stricter Criteria for Disability Insurance on Labor Force Participation. Journal of Public Economics 95, nos. 9–​10 (2011): 1223–​1235.

182

182  Work and the Social Safety Net Van Sonsbeek, J. M., and R. Gradus. Estimating the Effects of Recent Disability Reforms in the Netherlands. Tinbergen Institute Discussion Paper No. 11–​ 121/​3. Amsterdam: Tinbergen Institute, 2011. World Health Organization (WHO). International Classification of Disability, Health and Functioning. Geneva: World Health Organization, 2001. Ziebarth, N. R. Langzeitkranke verlieren durch Kürzung des Krankengeldes fünf Milliarden Euro. DIW Wochenbericht German Institute for Economic Research 76, no. 20 (2009): 326–​332. Ziebarth, N. R. Long-​Term Absenteeism and Moral Hazard–​Evidence from a Natural Experiment. Labour Economics 42 (2013):277–​292. Ziebarth, N. R., and M. Karlsson. A Natural Experiment on Sick Pay Cuts, Sickness Absence, and Labor Costs. Journal of Public Economics 94, nos. 11–​ 12 (2010): 1108–​1122. Ziebarth, N. R., and M. Karlsson. The Effects of Expanding the Generosity of the Statutory Sickness Absence Insurance. Journal of Applied Econometrics 29, no. 2 (2014): 208–​230.

 183

8 LESSONS FOR US DISABILITY POLICY FROM OTHER OECD COUNTRIES Richard V. Burkhauser and Mary C. Daly

INTRODUCTION All countries of the Organisation for Economic Cooperation and Development (OECD) provide some form of public protection to working-​age people with disabilities.1 In their early manifestations, these programs focused on easing the financial consequences of disability by providing cash assistance in lieu of full-​time work. Over time, a number of countries have initiated fundamental reforms which moved away from simple cash assistance to emphasize pro-​work programs designed to help people with disabilities maintain their labor market connections. On balance, these changes have reflected concerns about rapid growth in program rolls as well as increased awareness that many people with disabilities can remain productive in the labor market. Fundamental reforms in these nations resulted in substantial declines in disability recipiency rates (disability beneficiaries as a share of the working-​age population). For the most part, the reforms focused on slowing entry onto long-​ term disability cash transfers by keeping newly impaired workers in the labor market. Although there were some efforts to reduce the number of existing beneficiaries through disability reassessments and work incentives, these were both 1. The contents of this paper do not necessarily reflect the view of the Federal Reserve Bank of San Francisco or the Federal Reserve Board of Governors. We thank Catherine van der List and Olivia M. Lofton for excellent research assistance. We have no financial interests that constitute a conflict of interest with this research. Richard V. Burkhauser and Mary C. Daly, Lessons for US Disability Policy from Other OECD Countries In: Work and the Social Safety Net. Edited by: Douglas J. Besharov and Douglas M. Call, Oxford University Press. © Oxford University Press 2023. DOI: 10.1093/​oso/​9780190241599.003.0008

184

184  Work and the Social Safety Net less important and less successful. In the United States, proposals to reform its primary long-​term disability cash transfer program (Social Security Disability Insurance [SSDI]) have been discussed periodically over the past three decades with only modest progress despite a growing caseload and funding pressures. As was the case in the Netherlands, Sweden, Great Britain, Germany, and Australia before their reforms (as discussed in the previous chapter), disability recipiency in the United States was rising at an unsustainable pace through the Great Recession and its aftermath. The SSDI rolls rose from 1.2 million in 1967 to 9 million in 2014, and, beginning in 2009, the SSDI program began paying out more in annual benefits than it receives in taxes and interest from its trust fund. In the light of this growth, the SSDI program was projected to be insolvent in 2016, but this was postponed with the passage of the Bipartisan Budget Act of 2015 (Social Security Administration 2015) which allowed funds from the Social Security Old-​Age and Survivors (OASI) Trust to be borrowed by the SSDI Trust Fund. In this chapter we describe the factors driving rapid SSDI program growth in the United States and show that they are similar to those faced by other OECD nations. We then discuss some shared lessons emerging from the reform experiences in other countries, described in the previous chapter. We argue that these shared lessons provide an important foundation for discussions about how best to protect and sustain protection for working-​age Americans with disabilities. We focus exclusively on the SSDI program, which is administered by the Social Security Administration. This program is the only federal government social insurance program in the United States targeting working-​age people with disabilities. However, the United States also provides categorical means-​ tested benefits to working-​age people with disabilities through the Supplemental Security Income (SSI)–​disabled adults program and to the families of children with disabilities through the SSI–​disabled children program. The Social Security Administration also administers these two means-​tested programs. The disability determination for the SSI–​disabled adults program is exactly the same as that for the SSDI program. The disability determination for the SSI–​disabled children program is similar but in an age-​appropriate way. But the benefits of both these programs are considerably smaller and are not related to past earnings. In addition, unlike SSDI, both have income and assets tests that ensure that funds are targeted to low-​income adults with disabilities who have little experience in the labor market or on children with disabilities in low-​income families. The histories, policy changes, and target populations of these two needs-​based programs are sufficiently different that we do not discuss them here. For a discussion of these two programs see Burkhauser and Daly (2011, 2016). In sum, the lessons are as follows: first, disability is not an immutable state. Rather, it is a product of health and the social, cultural, and economic environment faced by individuals with impairments. This leads to the second

18 5

Lessons for US Disability Policy  185 lesson: disability policy affects the behavior of individuals with impairments. As reforms in many OECD countries demonstrate, generous programs conditioned on not working lead to lower levels of work and higher disability recipiency rates than do programs focused on maintaining individuals with impairments in the labor market. The third lesson is that incentivizing individuals with impairments to stay in the labor market is far easier than incentivizing existing disability beneficiaries to return to work. This means that gaining control of disability rolls is best done by stemming the flow of new beneficiaries rather than trying to reduce existing SSDI caseloads. The final lesson is this: hurdles to reform are surmountable. Together, the experiences of other nations suggest that it is possible to balance the competing goals of providing social insurance against adverse health shocks during working-​age and maximizing the work effort of all working-​age adults with and without disabilities. Past disability policies in both the United States and other OECD countries have focused more on the former than the latter, resulting in rapid growth in disability transfer populations that outpaced growth in the economy. Efforts to shift to more pro-​work policies over the past four decades in Europe suggest that fundamental disability reforms, if done well, can lower projected long-​term costs for taxpayers, make the job of disability administrators less difficult, and, importantly, improve the short-​and long-​run opportunities of people with disabilities.

DISABILITY PROGRAM GROWTH IN THE UNITED STATES The US SSDI program provides earnings replacement insurance for workers whose recent earnings histories are sufficient for them to obtain coverage. Benefits are based on a worker’s earnings history but with a progressive adjustment that replaces a greater share of the earnings of lower-​earning workers and are funded by a payroll tax equally assigned to employer and employee. SSDI defines disability as an inability to engage in any substantial gainful activity because of a medically determinable physical or mental impairment that can be expected to result in death or has lasted or can be expected to last for a continuous period of not less than twelve months. Hence this program is intended only for those who are judged “totally and permanently” disabled. US employers are not required to provide sickness benefits to their workers. The US government also has no mandatory sickness benefit program nor are US firms required to pay into a temporary disability insurance program at the federal level. But temporary disability programs are mandatory in five states: California, Hawaii, New Jersey, New York, and Rhode Island. However, all states have mandatory Workers Compensation programs that provide both short-​and long-​term disability benefits to workers with job-​related partial or total disabilities. Here we focus only on SSDI, which is the only categorical federally run social insurance program in the United States for working-​age people with disabilities.

186

186  Work and the Social Safety Net Applicants for disability benefits move through a multistep and often lengthy process in which their health-​based impairments and the degree to which these impairments affect their ability to work are gauged. This process of determining whether an applicant is eligible for benefits has several layers. The first criterion for denial—​the earnings test—​is factually easy for Disability Determination Stage (DDS) gatekeepers to establish. While the level of earnings is somewhat arbitrary, it is easy to verify whether a person is working and earning above the limit. Virtually no one is denied benefits based on this criterion. The next stage is more difficult: determining whether the applicant’s impairment is sufficient to prevent the applicant from working is a key challenge for any disability program. The Social Security Administration attempts to set the level of severity according to current medical research on the condition in question. Unfortunately, there is no single answer about the effect most impairments will have on work ability. The work limitations associated with impairments vary by individual and by the social and physical environments in place. Some impairments are easier to judge than others, but where it is more difficult to establish the severity of an impairment or its potential effect on an applicant’s ability to work, administrative discretion plays a greater role in determining the applicant’s outcome, with mental illness and musculoskeletal conditions among the most difficult to measure. (For a more thorough discussion, see Institute of Medicine 2007.) For applicants who do not meet or exceed the medical listings, DDS gatekeepers consider a set of vocational criteria (e.g., age, education, skill level, etc.). Presently, the majority of initial SSDI judgments by these DDS gatekeepers are based on vocational criteria rather than on the severity of an applicant’s health condition alone (increasingly so in regard to mental illness and musculoskeletal conditions). Those denied benefits initially can ask to be reconsidered. Failing reconsideration, they can appeal at the Administrative Law Judge (ALJ) level. At the ALJ level, applicants can, for the first time in this process, be represented by counsel and present expert witnesses. In contrast, the Social Security Administration can do neither at this level. A sign that the disability determination process is becoming not only more discretionary but also more time-​consuming is that appeals at the ALJ level substantially increased over the history of the program. Once on the system, few beneficiaries are reevaluated unless they work beyond a relatively low earnings threshold, and very few beneficiaries exit the program other than through death or reaching retirement age. At normal retirement age, an SSDI beneficiary officially leaves the program but continues to receive the same benefit now paid out of the OASI Trust Fund. From the 1970s through the Great Recession and its aftermath, the number of disabled worker beneficiaries increased nearly six-​fold, rising from 1.5 to 9 million in 2014. This rapid growth in the rolls put increasing pressure on program finances. Inflation-​adjusted SSDI expenditures rose more than four-​fold,

18 7

Lessons for US Disability Policy  187 growing from $33 to $145 billion (in 2014 dollars) over this same period. Based on that growth, in 2015, the SSDI program was projected to be insolvent by 2016 (Social Security Administration 2015). This technical insolvency was postponed when President Obama signed the Bipartisan Budget Act of 2015 in November 2015, allowing OASI Trust Fund dollars to be borrowed by the SSDI Trust Fund. While this temporarily resolved the issue of technical insolvency, in itself it did nothing to alleviate the shortfall between projected revenues coming into the SSDI system and its projected expenditure. The rapid rise in caseloads and costs are even more worrisome when put in the context of the broader goals of the SSDI program: to protect the economic well-​being of people with disabilities. Since the passage of the Americans with Disabilities Act of 1990 (ADA), the employment rate of those with disabilities has declined considerably and their household income has remained flat. Increasingly, people with disabilities are substituting SSDI benefits for labor market earnings, making them net withdrawers rather than net contributors to the tax base during their working age. This outcome challenges the finances of the SSDI program and is at odds with the view of disability codified in the ADA that people with disabilities are able and willing to participate in the labor market.

WHY HAS THE SSDI CASELOAD INCREASED? Possible explanations for SSDI program growth can be broadly classified into two groups: (1) those that are focused on events not directly related to the program itself (i.e., the aging of the population, changes in the underlying severity of disability, the entry of women into the labor force, or the increase in the normal retirement age from sixty-​five to sixty-​six in the Social Security [OASI] retirement program), and (2) those that are focused on changes in program rules or their application (the cyclicality of application rates, the growth in SSDI benefits relative to wage earnings, and specific changes in rules and their interpretation and implementation over time). Work by Daly, Lucking, and Schwabish (2013), using shift share analysis argues that, after accounting for these one-​time unrelated factors, 43.8% of the growth in SSDI recipiency rates (SSDI beneficiaries as a share of the working-​age population) between 1980 and 2011 was program-​related and will continue to affect program growth in the future. Research by Duggan and Imberman (2009) and Autor and Duggan (2006, 2010), looking at somewhat different periods, finds that these program-​ related changes affected individual behavior and accounted for an even larger share of program growth. Most importantly, all of these researchers expect that these program-​related factors will continue to play a large and increasing role in program growth going forward.

18

188  Work and the Social Safety Net

Non–​Program-​Related Growth Factors The most obvious potential driver of SSDI growth is the aging of the population. Since SSDI benefits are conditioned on having a disability, and disability generally rises with age, the aging of the baby boom generation will, on net, push up the SSDI rolls. A simple way to gauge the impact of this change is to fix SSDI recipiency rates by age group in some period and let growth in the rolls evolve based on changes in the age structure of the population. Autor and Duggan (2006, 2010) do this and find that, between 1984 and 2003, changes in age structure accounted for about 6% of the increase in SSDI receipt among the non-​ elderly population over the period. Daly, Lucking, and Schwabish (2013), find that between 1980 and 2011 the aging of the population accounted for 17.9% of SSDI recipiency growth (see Figure 8.1). Another potential driver is health. To qualify for SSDI benefits, individuals must have a medically determinable ailment expected to last for at least twelve months or result in death. If the health of the insured population has declined over time, this would influence program enrollment and growth. Surveys asking about activity and work limitations point to a relatively stable pattern in these measures over the past two decades. Although work and activity limitations rise with age, there is little evidence that the prevalence within an age group of such limitations has increased over time. (For additional discussion, see Burkhauser and Daly 2011.) % 50 43.8 40 29.2

17.9

Recipiency Catch-Up

30

12.8

9.1 16.5

20

10

SSDI Eligibility

0 Increase in retirement age

Aging of the population

Women’s labor force

Residual

Figure 8.1  Sources of SSDI prevalence rate growth, 1980–​2011. Source: Social Security Administration, Bureau of Labor Statistics, and FRBSF Staff Calculations. https://​www.frbsf.org/​econo​mic-​resea​rch/​publi​cati​ons/​econo​mic-​let​ter/​2013/​june/​fut​ure-​soc​ial-​ secur​ity-​dis​abil​ity-​insura​nce-​ssdi/​. (Daly, Lucking, and Schwabish 2013.)

18 9

Lessons for US Disability Policy  189 Changes in the labor force participation of women have also influenced program growth. Since SSDI is an insurance program, eligibility for benefits requires a fixed number of quarters of covered employment. The substantial increase in the labor force participation of women has increased both their SSDI coverage and their receipt of disability benefits. It is straightforward to compute the magnitude of this change on the total growth in SSDI rolls. Researchers have found that the increased number of women in the paid labor force can only explain a fraction of the total rise in SSDI caseloads since the mid-​1980s. For instance, Daly, Lucking, and Schwabish (2013) find that, between 1980 and 2011, the increase in the number of women in the paid labor force can explain 16.5% of the growth of the SSDI caseload (see also Autor and Duggan 2006, 2010). They recognize, however, that in 1980, women’s SSDI recipiency rate was well below that of men, even after accounting for the lower eligibility of women. Analysts do not agree on what explains this gap. Some argue it reflects underlying health differences between men and women. Others maintain that women eligible for SSDI were not representative of the entire population of women in 1980 and that a more representative sample of women would have had a recipiency rate similar to men. They quantify this assumption by setting the recipiency rate for women equal to that of men in 1980. As Figure 8.1 shows, this adds another 12.8% to their estimates of how much the greater eligibility of women has contributed to rising recipiency rates. Daly, Lucking, and Schwabish (2013) add a final non–​program-​related factor: the increase of the normal retirement age to sixty-​six over this period. Doing so accounts for another 9.1% of SSDI growth. Combining the estimated contributions of population aging, changes in health, the entry of women into paid work, and the change in the OASI retirement age, Daly, Lucking, and Schwabish (2013) still find that 43.8% of SSDI caseloads over the past three decades was accounted for by other factors most likely related to incentives implicitly or explicitly built into the SSDI program (see Figure 8.1). We turn to these factors now.

Program-​Related Growth Factors SSDI recipiency rate fluctuations line up with changes in Social Security Administration policies that make it easier (or harder) to receive SSDI benefits. This can be seen in Figure 8.2, updated from Burkhauser, Daly, and Ziebarth (2016) to include more recent years, which shows the time series of SSDI recipients per 100 working-​age persons (ages 16–​64) from 1970 to 2018, with recession years shaded in gray. After a period of rapid increases in recipiency rates in the early 1970s, these rates fell in the late 1970s and early 1980s, first because program gatekeepers were urged to interpret existing rules more strictly and then because Congress, in 1980, required the Social Security Administration to reevaluate all current recipients to see if they still met the medical standards. This

190

190  Work and the Social Safety Net 6

5

4

3

2

1

1970

1974

1978

1982

1986

1990

1994

1998

2002

2006

2010

2014

0 2018

Figure 8.2  United States SSDI recipients per 100, Ages 16–​64, 1970–​2018. Source: Social Security Administration and Census Bureau. (Burkhauser, Daly, and Ziebarth 2016.)

rule change, which was rigorously enforced by Social Security Administration at the start of the Reagan administration, resulted in a drop in the SSDI rolls despite a deep recession. By 1983, the widespread reevaluation of those already on SSDI was largely halted as the courts, and then Congress, put limits on the Social Security Administration’s authority to reevaluate beneficiaries. In 1984, responding in part to a backlash against restrictive cuts imposed in the Social Security Disability Amendments of 1980, policymakers loosened eligibility standards by expanding the impairments with which a person could medically qualify for the SSDI program. The 1984 legislation moved away from a strict medical listing determination of eligibility to one that also considered an applicant’s overall medical condition and ability to work. These changes meant that applicants could qualify for SSDI based on having multiple conditions even when no single condition would meet the SSDI eligibility threshold. In addition, the legislation allowed for symptoms of mental illness and pain to be counted when assessing SSDI eligibility, regardless of whether the person had a verifiable medical diagnosis. In practice, this expansion of eligibility to more difficult-​to-​measure impairments has placed increasing burdens on the Social Security Administration to evaluate less verifiable claims and make determinations about their impacts on an applicants’ ability to work. (See Burkhauser and Daly 2011 for a more detailed discussion.)

19 1

Lessons for US Disability Policy  191 Another area of program growth has been applicants who move onto the program based on vocational requirements. While these criteria have not changed over the history of the SSDI program, their use by program gatekeepers to determine benefit eligibility has risen significantly since 1991. Currently, they are used to determine the majority of new awards, especially among those with the more difficult-​to-​determine conditions of mental illness and musculoskeletal conditions—​the primary condition of more than 50% of all newly enrolled beneficiaries. (See Burkhauser and Daly 2011 for fuller discussion.) The effect of this growing share of harder-​to-​evaluate applicants was a substantial variation in the flow of applicants onto the rolls. This variation comes from both fluctuations in applicant inflow and variations in decision-​making among SSDI gatekeepers. For example, Maestas, Mullen, and Strand (2013) using Social Security Administration administrative records find that, at the initial DDS decision-​making point, 23% of new applicants in 2005 were marginal cases whose admittance into the program was determined by the luck of being assigned to an easier rather than a stricter DDS gatekeeper. Importantly, when they compare the subsequent work histories of those who entered the program in this way with a matched set of applicants who drew a stricter DDS gatekeeper, they find the latter group’s employment was on average 20 percentage points higher. This difference was even greater for those with less severe medical conditions. This research suggests that, increasingly, applicants admitted to the SSDI rolls on these looser criteria had greater work capacity than assumed for those receiving SSDI benefits. The differences in allowances are important, especially when one considers how application rates fluctuate with economic conditions. Plots of the SSDI application rate and the national unemployment rate show that, with the exception of the double-​dip recession in the 1980s, application rates are highly correlated with the business cycle. They rise during recessions and fall during periods of economic growth. Most research on the consequence of business cycles on application rates finds that economic conditions play a substantial role in SSDI application and award patterns over time (see Burkhauser and Daly 2011, 2012; Miller et al. 2019). This especially was the case during the Great Recession and its aftermath, with recipiency rates hitting a historic high of 4.3% in 2013. However while SSDI rolls increased in 2014, for the first time since 1983 when recipiency rates were 1.7%, recipiency rates fell in 2014 and have fallen slightly since then to 4.1% in 2018 (the latest year for which data are available). In sum, a large share of SSDI growth has been driven by factors other than an aging workforce, health declines, the increasing SSDI coverage of women, and changes in the OASI normal retirement age. Loosening of program rules in the 1980s has made it more difficult for gatekeepers to judge eligibility and increased the likelihood that applicants facing rising replacement rates or declining economic opportunities will apply for SSDI benefits. A growing number

192

192  Work and the Social Safety Net Index 450

1970 = 100

400 20–29

350 300

30–39

250 40–49 50–59 60–64

200 150 100 50

1980

1985

1990

1995

2000

2005

2010

0

Figure 8.3  Normalized growth in SSDI prevalence by age, 1980–​2011. Source: Social Security Administration and FRBSF Staff Calculations. Note: prevalence calculation uses insured population. (Daly, Lucking, and Schwabish 2013.)

of individuals being allowed onto the rolls could work in some capacity and would do so if they were not judged eligible for benefits. One indicator that these program-​related factors have an independent effect on program growth can be seen in Figure 8.3. While it is certainly the case that aging baby boomers have increased program growth as they have become a larger share of the workforce, Figure 8.3 shows that the growth in the prevalence of SSDI program receipt compared to 1970 was far higher at younger than at older ages. The most rapid growth was among those ages twenty-​five through thirty-​nine. Because these recipients are likely to stay on the rolls many more years than those who enter at older ages, their lifetime impact on program costs will be far higher. Not factoring in the important role policy changes as well as business cycle effects have had and will continue to have on program growth can importantly affect SSDI program growth projection. Figure 8.4 is updated from Daly, Lucking, and Schwabish (2013), but it is based on official Office of the Chief Actuary (OCACT), Social Security Administration, and historical and projected SSDI beneficiary populations (Social Security Administration, various years). It shows actual SSDI prevalence rates between 1980 and 2020 and OCACT projections of future prevalence rates at various times in the past, as well as their 2021 projection. In some cases, they overestimate program growth but in most cases, these projections substantially underestimated future SSDI program growth. The reason is that OCACT did not fully take into consideration the

19 3

Lessons for US Disability Policy  193 Beneficiaries per 100 age 20–64 insured population 6 2013 Projection 2015 Projection 2021 Projection 2005 Projection

1995 Projection

5 4

2000 Projection 3 1990 Projection

Actual

2 1 0

1980

1985

1990

1995

2000

2005

2010

2015

2020

2025

2030

Figure 8.4  Past SSA projections of SSDI prevalence rate, 1980–​2020. Source: Short-​and long-​range actuarial projections of the old-​age, survivors, and disability insurance program (various years), Bureau of Labor Statistics, Census Bureau, and FRBSF Staff Calculations. Updated from Daly, Lucking, and Schwabish (2013).

influence of program effects on past behavior that accounts for a significant percent of the program growth discussed above. Comparing their 2013 projection in Figure 8.4 with the actual prevalence rates between 2013 and 2018, they did not anticipate the 2013 peak in prevalence rates and the modest decline in those rates through 2020. What is clear is that despite the modest decline in prevalence rates since their historic 2013 peak, it is unlikely that SSDI prevalence rates in the future will ever return to 1990s levels unless fundamental changes are made in SSDI policy.

UNITED STATES DISABILITY GROWTH VERSUS OTHER OECD COUNTRIES The number of workers receiving disability-​based benefits increased substantially in most OECD countries over the past forty years. Population growth accounted for part of this increase, but disability caseloads as a share of the working-​age population (the disability recipiency rate) also rose substantially. Pattison and Waldron (2013) argue that population growth explains the bulk of the rise in SSDI recipiency in the United States. Duggan and Imberman (2009) and Burkhauser et al. (2014) remove the influence of population growth and consider the factors that explain the remaining rise in the program as discussed above. Since population growth alone would not put additional financial pressure on the system, knowing what these other factors are is critical to policymakers

194

194  Work and the Social Safety Net tasked with funding the system. Rather than focusing on SSDI recipiency rates, Liebman (2015) focuses on SSDI incidence rates and argues that most of its increase, controlling for other factors including unemployment rates, occurred in the 1980s. Since then, this controlled measure of incidence has leveled off but at a substantially higher rate than in the early 1970s. Figure 8.5, plots yearly disability recipiency rates for the United States and the five OECD countries from Chapter 7 from the 1970s. It shows the total number of people who received long-​term categorical disability income benefits as a share of the working-​age population in each country. The US disability recipiency rate only includes beneficiaries receiving SSDI. When SSI-​disabled adults and SSDI program beneficiaries are combined, the level of the US disability recipiency rate was higher, but the patterns over time are roughly the same. This point is demonstrated in Burkhauser et al. (2013, ­figure 2). This is the most critical number to policymakers since it measures the magnitude of the fiscal burden that these disability programs place on country finances. The fiscal burden of Percentage

12 10 8 6 4 2

0 1970 1974 1978 1982 1986 1990 1994 1998 2002 2006 2010 2014 2018 Australia Germany Great Britain Netherlands Sweden United States

Figure 8.5  Disability receipiency rates across countries, 1970–​2018. Source: Social Security Administration, US Census Bureau, Australian Government Department of Social Services, Australian Bureau of Statistics, Department of Work and Pensions, Office for National Statistics, Statistics Sweden and Swedish Social Insurance Agency yearbooks, Statistics Netherlands, the Institute of Employee Benefit Schemes, and the OECD’s Social Benefit Recipients Database (SOCR).

19 5

Lessons for US Disability Policy  195 disability programs comes from the fact that beneficiaries receive income support and generally do not contribute to the tax base since they are largely out of the labor force. We provide a more detailed description of the US data in the Appendix. In 1970 (the first year of US data), its 1.2% disability recipiency rate was considerably lower than any other country. By 2018 (the final year of US data), it was 4.1%. This substantial increase in the US disability recipiency rate between the first and final year of data, with the exception of Germany, is also seen in the other examined countries. As the figure highlights, however, the pattern of within country growth is along significantly different trajectories. Importantly, the disability recipiency rate of the four European countries peaked in the 2000s or earlier and has been decreasing since then. This can be seen more clearly in Table 8.1, which adds information from the United States regarding its initial recipiency rate and year, its peak rate and year, and its most recent rate and year to those already seen for the other five OECD countries in this table in the previous chapter. In contrast to those countries, the disability recipiency rate in the United States grew continuously since the early 1980s, peaked in 2013, and has only modestly fallen since then.

WHAT CAN THE UNITED STATES LEARN FROM THE REFORM EXPERIENCES OF OECD COUNTRIES? In the remainder of this chapter, we consider how the reform experiences in the Netherlands, Sweden, Great Britain, Germany, and Australia can inform US Table 8.1  Disability recipiency rates for six OECD countries Australia

Great Britain

Netherlands Sweden Germany

United States

Year

1970

1971

1973

1970

1970

1970

DSP recipiency rate

1.6

2.7

3.1

3.5

4.2

1.2

Year

2011

2003

2003

2007

1984

2013

DSP recipiency rate

5.4

6.7

7.4

10.2

5.8

4.3

Year

2020

2018

2018

2018

2018

2018

DSP recipiency rate

4.0

4.9

4.8

5.6

3.4

4.1

Initial year

Peak year

Final year

Sources: Australia: Department of Social Services, Australian Bureau of Statistics; Germany: Deutsche Rentenversicherung, Statistisches Bundesamt, OECD’s Social Benefit Recipients Database (SOCR); Great Britain: Department of Work and Pensions, Office for National Statistics, OECD’s SOCR; Netherlands: Statistics Netherlands, Institute of Employee Benefit Schemes, OECD’s SOCR; Sweden: Statistics Sweden, Swedish Social Insurance Agency yearbooks, OECD’s SOCR; United States: Census Bureau, Social Security Administration.

196

196  Work and the Social Safety Net policymakers tasked with ensuring SSDI’s fiscal sustainability. First and foremost, to effectively design policies to curb disability recipiency rate growth, policymakers must acknowledge why it is growing. A large body of research has shown that, in advanced industrialized economies with large and complicated safety nets, both exogenous factors (population demographics, population health, or economy-​wide shocks) and endogenous factors (changes in eligibility rules, benefit replacement rates, or access to other social programs) can influence disability recipiency rate growth. The disability reforms undertaken in the Netherlands, Sweden, Great Britain, Germany, and, to a lesser degree, Australia were possible once these countries acknowledged that changes in behavior rather than changes in health were driving program growth. By recognizing that their policy designs influenced behavior and program outcomes, policymakers opened the door to more fundamental disability policy changes. Importantly, this acknowledgment also allowed policymakers to rethink their national disability scheme’s goals and bring them more in line with modern notions of disability. This meant discarding terminology and criteria of incapacity and replacing them with a focus on ability. More than just semantic window dressing, the new terminology changed the expectations of people with impairments from dependence on disability benefits based on not working to work. Of course, for some individuals work was not possible, but this increasingly became the exception rather than the norm. As the previous chapter demonstrated, each of these OECD countries implemented reforms slightly differently, but all shared the goal of curbing unsustainable program growth by changing the culture and social expectations of and for people with disabilities and better aligning the incentives embedded in program design with these expectations. The data show the reforms have generally been successful. Although each country recognizes their reforms have not been completely successful, from the US perspective, these reforms demonstrate that policies matter and provide a relevant starting point for discussions about reforming the US system. Below we provide the four lessons most relevant to United States policymakers.

Lesson 1: Disability Does Not Mean Incapacity Disability is frequently thought of as an immutable, health-​based condition that limits functionality and prevents the performance of socially expected tasks (such as attending school or working). Modern notions, however, of the productive capacity of all people is at odds with this characterization. A more accurate concept of disability (and one embodied in the ADA in the United States and in the World Health Organization’s [WHO] International Classification of Functioning, Disability, and Health [ICF]; WHO 2001) considers disability to be the product of an interactive process between an individual’s health conditions

19 7

Lessons for US Disability Policy  197 and the social and physical environment. From this perspective, disability is neither immutable nor determined solely by health. Each of the OECD countries reviewed in the previous chapter embraced this idea: ensuring that their citizens with disabilities have opportunities to work in the paid labor market is now a government priority (see Organisation for Economic Cooperation and Development [OECD] 2003; Bernd et al. 2004 for a discussion of this transformation in disability policy goals). Key among the findings based on their successful efforts to reduce their disability recipiency rates is that a substantial share of people who were moving onto long-​term cash transfer disability programs were, with reasonable levels of government-​provided support, able to find or maintain employment. Of course, a subset of workers with disabilities had impairments so severe that work was not possible, but this was a smaller portion than previously accepted onto these programs. Those now coming onto the rolls are, by demonstration rather than assumption, unable to integrate effectively into the labor market even with appropriate incentives and support. Importantly, in Great Britain, the assumption is that with appropriate support the majority of people with disabilities can perform some type and amount of work. This design offers an important advantage over programs based on categorical determinations of disability that can expand in unintended ways when individuals are incentivized to change their behavior (i.e., attempt or not attempt to work) to become eligible. The lesson for US disability policy is that the population with disabilities is heterogeneous and many of its members can work. This view is at odds with the current US system in which the Social Security Administration requires applicants to demonstrate an inability to perform substantial gainful activity before receiving access to benefits or any other type of support, including work support. Embracing the ideas of many European countries about the work capacity of individuals with disabilities calls for restructuring the US system to bring forward the focus on employment and make long-​term cash benefits a last-​resort option once rehabilitation and accommodation have failed.

Lesson 2: Incentives Affect Behavior Once it is recognized that the social and cultural environment faced by individuals with disabilities partly determines the extent to which their impairment limits them, it is easy to see that the incentives embedded in policy design can affect outcomes. The important lesson from the reforms in our sample of countries is that it is important to correctly incentivize all actors in the process to attain the outcomes desired. In the Netherlands, this meant making employers bear more of the direct costs of the program and making employees comply with rehabilitation and retraining in order to maintain benefits. In Sweden, this meant standardizing the disability

198

198  Work and the Social Safety Net screening process and holding disability gatekeepers accountable for engaging applicants in work rehabilitation plans. Sweden also mandated that employers document how they had attempted to accommodate impaired workers. Finally, reforms focused on making workers comply with the work-​first approach by reducing or eliminating benefits to those workers who did not comply with the rehabilitation and accommodation plans. Great Britain similarly made employers responsible for paying sickness benefits for the first twenty-​eight weeks of a claim, standardized the disability screening process, and made it increasingly difficult to remain passively on long-​term disability benefits without engaging in work-​related activity. Reforms have also sought to address financial incentives related to imbalances in replacement rates between disability benefits and other noncategorical cash benefits. In Australia (since 2011) it has meant implementation of a new disability benefit assessment procedure under which eligibility for benefits typically requires evidence that claimants were unable to obtain employment through vocational rehabilitation or other employment services over a period of up to two years, during which time claimants only receive lower noncategorical social minimum benefits (New Start Allowance) and associated support services. The lesson for US policymakers is that program incentives affect how people with disabilities and their employers react to and fare after the onset of a health impairment. In the United States, SSDI is funded from a payroll tax, and the federal government is responsible for a great share of the costs associated with providing long-​term disability benefits to working-​age people with disabilities. Because they bear no direct responsibility for funding benefits paid to former employees, employers have no direct financial incentive to accommodate and rehabilitate employees who become impaired. Incentivizing employers to make greater investments in accommodation and rehabilitation by creating a scheme that makes employers internalize some or all of the costs of moving employees onto long-​term disability could curb SSDI growth by more effectively aligning incentives. Autor and Duggan (2010) and Burkhauser and Daly (2011) propose two different models for doing this in the United States. As the reforms in these countries show, incentives cannot be changed solely for employers. Incentives for people with disabilities and program gatekeepers also need to be aligned with program goals to realize sizeable and lasting reductions in disability recipiency rates.

Lesson 3: Early Intervention Reduces Flows A recurring theme in the experiences of the Netherlands, Sweden, Great Britain, and Germany is that reforms focused on early intervention can be very successful at reducing the flow of new beneficiaries onto the program and/​or boosting the

19 9

Lessons for US Disability Policy  199 flow of new beneficiaries off the program. In the Netherlands, for example, early intervention that coordinated employer action following the employees’ particular health shocks was critical to keeping impaired workers in the labor force. Such intervention significantly increased the return to work of impaired workers and reduced time on the sickness or disability program. Evidence from Sweden shows that waiting even one year following the onset of impairment significantly reduced the chance that rehabilitation will result in a return to work. Evidence from Great Britain suggests that bringing medical screening forward can help to control program growth. In contrast, none of these countries, including Australia, was successful at moving existing longer-​term beneficiaries back into the labor market, although the process of rescreening and seeking to activate existing longer-​ term beneficiaries in Great Britain has not yet been fully evaluated. Across all countries, once enrolled on disability benefits for more than a few months, only a small fraction of recipients returned to work. Even when strict time limits are put in place, as in the case of Sweden, movement off the disability system for longer-​duration beneficiaries is difficult. Even when it is done, most frequently it results in a shift to another public program rather than into employment. These experiences have several implications for US policy discussions. First, the fact that most current SSDI recipients do not work is not evidence that they would have been unable to work if given alternative policy treatments (e.g., timely accommodation and rehabilitation). Indeed, the marked difference in outcomes between those given early versus later employment-​oriented services in the Netherlands and Sweden shows that, in a system oriented toward long-​ term cash benefits rather than work, many of those with residual work capacity will never return to work. The experiences in these countries also call into question the viability of ongoing attempts to gain control of the growth in SSDI rolls in the United States by funding additional continuing disability reviews or enhancing post-​entry rehabilitation or job training programs like Ticket to Work. While such programs have merits, the experiences in Sweden tell us that these efforts will fall short of bringing growth in the rolls down to sustainable levels. Finally, the reforms and outcomes in these countries show the difficulties of focusing policy reforms on current beneficiaries—​ a practice Social Security Administration is forced to follow by rules requiring Social Security Administration-​collected funds to be focused on current program recipients. This suggests that eliminating this rule and allowing the Social Security Administration to focus its energies on workers with health-​based work limitations who are trying to decide whether to stay on the job or apply for benefits will likely do more to curb SSDI growth.

20

200  Work and the Social Safety Net

Lesson 4: Hurdles to Reform in the US Are Surmountable Despite this growing body of evidence that structural reforms to long-​term cash disability programs can curb program growth and potentially improve outcomes for those with health-​based impairments, the political coalition necessary to achieve fundamental disability policy reform has been slow to evolve in the United States. So, what are the hurdles? There are several, including a number of concerns raised by supporters of the current system (see Reno and Ekman 2012 for a summary of these arguments). One issue raised in response to proposals for fundamental SSDI reform is that these benefits, while not especially generous, are essential to keeping millions of disabled worker beneficiaries out of poverty. The evidence from Europe shows that this is a very static view which assumes that, in the absence of benefits, individuals with disabilities would remain out of the labor market and dependent on other forms of public or private assistance for support. European disability reforms over the past decade provide plausible evidence that increased employment will occur when pro-​work policies replace policies that have had the opposite effect. Their reform experience shows that a significant number of people with disabilities who would otherwise have moved onto long-​term cash benefits were able, with reasonable levels of support, to return to work (OECD 2010). While it is always the case that tightening the criteria for disability benefits runs the risk of denying benefits to those who will not be able to find work, on balance the European experience suggests that reasonable pro-​work policies will both substantially reduce disability recipiency rates and increase the employment of those who would otherwise have been on the long-​ term disability rolls. Nonetheless, it is the case that fundamental SSDI reform could worsen the already precarious economic position of some people with disabilities. This, however, should be put into the context of what has been happening to the overall well-​being of working-​age people with disabilities in the United States. Over the past twenty-​five years, the household incomes of Americans with disabilities have stagnated relative to other working-​age people. This is associated, in part, with the near continuous decline in their employment. Our reading of the European evidence is that when programs are designed to award cash transfers in lieu of work, employment falls. In contrast, when programs are designed to encourage work and award transfers only when work clearly is not possible, employment rises. Since work generally leads to increased income, especially when public policies make work pay, efforts to promote work among those with disabilities can produce positive outcomes. The most important lesson though, is that by not working, long-​term disability beneficiaries will, over time, slip further and further behind the rest of

20 1

Lessons for US Disability Policy  201 the working-​age population whose employment allows them to benefit from increased economic growth. A final concern is that programs like SSDI are especially important in economic downturns where individuals with limited work capacity are not only more likely to be laid off, but less likely to find a new job. Past experience of European countries, especially the Netherlands, which intentionally or unintentionally used this logic to turn their long-​term disability programs into more general unemployment programs, suggests that it can be a very expensive and ultimately ineffective policy decision. Indeed, many European nations struggled to regain control over their disability systems, which for many decades were used as long-​term unemployment insurance programs. A key message from the European experience is that explicitly divorcing long-​term “unemployability” insurance from disability insurance is critical to targeting resources toward both populations. Together, the experiences of other nations suggest that it is possible to balance the competing goals of providing social insurance against adverse health shocks during working age and maximizing the work effort of all working-​age adults with and without disabilities. Past disability policies in both the United States and other OECD countries have focused more on the former than the latter, resulting in rapid growth in disability transfer populations that outpaced growth in the economy. Efforts to shift to more pro-​work policies over the past decade in Europe suggest that fundamental disability reforms, if done well, can lower projected long-​term costs for taxpayers, make the job of disability administrators less difficult, and, importantly, improve the short-​and long-​run opportunities of people with disabilities.

APPENDIX DATA DESCRIPTION AND SOURCES Table 8.A1  Summary of disability insurance data availability across countries Australia

Great Britain

Netherlands Sweden

Germany

United States

Initial Year

1970

1971

1973

1970

1970

1970

Final Year

2020

2018

2018

2018

2018

2018

Missing Years

(–​)

(–​)

(–​)

1984

gaps until '00

1981

Age Range of Working Population

16–​64

16–​64

15–​65

16–​64

16–​64

16-​64

United States: Historical population estimates are from the Census Bureau’s Annual Estimates of Resident Population (http://​www.cen​sus.gov). SSDI caseloads and covered workers data are from the Annual Statistical Supplement to the Social Security Bulletin (http://​www.ssa.gov/​pol​icy/​docs/​statco​mps/​sup​plem​ent/​). See Chapter 7 for these definitions for other countries.

20

202  Work and the Social Safety Net

REFERENCES Autor, D. H., and M. G. Duggan. The Growth in the Social Security Disability Rolls: A Fiscal Crisis Unfolding. Journal of Economic Perspectives 20 (2006): 71–​96. Autor, D. H., and M. G. Duggan. Supporting Work: A Proposal for Modernizing the U.S. Disability Insurance System. Washington, DC: The Hamilton Project and Center for American Progress, 2010. Bernd M., C. Prinz, and M. Queisser. Transforming Disability Welfare Policies: Toward Work and Equal Opportunity. Burlington, VT: Ashgate Publishing Co., 2004. Burkhauser, R. V., and M. C. Daly. The Declining Work and Welfare of People with Disabilities: What Went Wrong and a Strategy for Change. Washington, DC: AEI Press, 2011. Burkhauser, Richard, V., and M. C. Daly. Social Security Disability Policy: Time for Fundamental Change. Journal of Policy Analysis and Management 31, no. 2 (2012): 454–​461. Burkhauser, Richard V., and Mary C. Daly. Making Work a Priority for Working-​ Age People with Disabilities. In The US Labor Market: Questions and Challenges for Public Policy, ed. Michael Strain. Washington, DC: American Enterprise Institute Press, 2016, 204–​220. Burkhauser, R. V., M. C. Daly, and B. T. Lucking. Is Australia One Recession Away from a Disability Blowout? Lessons from Other Organization for Economic Cooperation and Development Countries. Australian Economic Review 46, no. 3 (2013): 357–​368. Burkhauser, R. V., M. C. Daly, D. McVicar, and R. Wilkens. Disability Benefit Growth and Disability Reform in the US: Lessons from Other OECD Nations. IZA Journal of Labor Policy 3, no. 4 (2014): 1–​30. Burkhauser, R. V., M. C. Daly, and N. R. Ziebarth. Protecting Working-​Age People with Disabilities: Experiences of Four Industrialized Nations. Journal for Labour Market Research. 49, no. 4 (2016): 367–​386. Daly, M., Lucking B. T., and Schwabish J. The Future of Social Security Disability Insurance. Economic Letters. San Francisco: Federal Reserve Board of San Francisco, June 2013. Daly, M., Lucking B. T., and Schwabish J. Explaining the Rapid Growth in SSDI Rolls. San Francisco: Federal Reserve Bank of San Francisco, 2013.http://​ www.frbsf.org/​econo​mic-​resea​rch/​files/​exp​lain​ing-​dis​abil​ity-​gro​wth.pdf. Duggan, M., and S. Imberman. Why Are the Disability Rolls Skyrocketing? The Contribution of Population Characteristics, Economic Conditions, and Program Generosity. In Health at Older Ages: The Causes and Consequences of Declining Disability Among the Elderly, ed. David Cutler and David Wise. Chicago: University of Chicago Press, 2009:337–​379.

20 3

Lessons for US Disability Policy  203 Institute of Medicine. Improving the Social Security Disability Decision Process. Washington, DC: National Academies Press, 2007. Liebman, J. B. Understanding the Increase in Disability Insurance Benefits Receipt in the United States. Journal of Economic Perspectives 29, no. 2 (2015): 123–​150. Maestas, N., K. J. Mullen, and A. Strand. Does Disability Insurance Receipt Discourage Work? Using Examiner Assignment to Estimate Causal Effects of SSDI Receipt. American Economics Review 103, no. 5 (2013): 1797–​1829. Miller, T., D. Pattison, and S. G. Ayala. Trends in Social Security Disability Insurance. Social Security Briefing Paper No. 2019-​01. Washington, DC: Social Security Administration, 2019. Organisation for Economic Cooperation and Development (OECD). Transforming Disability into Ability: Policies to Promote Work and Income Security for Disabled People. Paris: OECD Publishing, 2003. Organisation for Economic Cooperation and Development (OECD). Sickness, Disability and Work: Breaking the Barriers: A Synthesis of Findings across OECD Countries. Paris: OECD Publishing, 2010. Pattison, D., and H. Waldron. Growth in New Disabled-​Worker Entitlements, 1970–​2008. Social Security Bulletin 73, no. 4 (2013): 25. Reno, V. P., and L. D. Ekman. Social Security is Part of the Solution, Not a Cause of Work Disability: Response to Burkhauser and Daly. Journal of Policy Analysis and Management 31 (2012): 471–​474. Social Security Administration. The 2015 Annual Report of the Board of Trustees of the Federal Old-​Age and Survivors Insurance and Federal Disability Insurance Trust Funds, Washington, DC: Social Security Administration, 2015. World Health Organization (WHO). 2001. International Classification of Disability, Health and Functioning. Geneva: WHO, 2001.

204

9 ACTIVATION IN PUBLIC EMPLOYMENT SERVICES IN EUROPE J. Timo Weishaupt, Henning Jørgensen, and Alexander Nunn

B

y the early 1990s, there was widespread concern about a lack of competitiveness in European economies, as expressed in the famous Delors White Paper (European Commission 1993). With long-​term, structural unemployment being identified as a particular problem, the labor market was central to these concerns (Organisation for Economic Cooperation and Development [OECD] 1994). In this context, and inspired by US President Bill Clinton’s effort to “end welfare as we know it” in the United States, Tony Blair addressed Parliament in his first speech as Prime Minister with the words that “people need a hand up, not a hand out.” This quote perfectly captures the emerging spirit of welfare reform that began in the early 1990s in the United States and Western Europe. The “old” system of helping the needy through so-​called passive benefits was considered not only ineffective and inefficient, but also lacking in a social dimension because benefit recipients were allegedly incentivized to remain dependent on the welfare state (Weishaupt 2013). To most policymakers, a new welfare state efficacy was needed that combined welfare rights with obligations, promoting self-​reliance and social inclusion through labor market attachment. This shift in policy agenda, however, was not limited to the Anglo-​Saxon world. It found whole-​hearted, often social democratic, supporters across the European continent (Weishaupt 2011a) as it mixed liberal economic concerns with a preference for the market and social democratic concerns of social investment and state intervention to correct for market failures (Eichhorst and Hemerijck 2010). However, this progressive interpretation was not widely accepted; some interpreted the new activation J. Timo Weishaupt, Henning Jørgensen, and Alexander Nunn, Activation in Public Employment Services in Europe In: Work and the Social Safety Net. Edited by: Douglas J. Besharov and Douglas M. Call, Oxford University Press. © Oxford University Press 2023. DOI: 10.1093/​oso/​9780190241599.003.0009

20 5

Activation in PES in Europe  205 paradigm as a broader shift toward a Schumpeterian post-​national workfare regime, where supply-​side state intervention is focused on pro-​market objectives and the achievement of labor market flexibility (Jessop 2005; Nunn 2014; Nunn and Beeckmans 2015). Others similarly described it as an incentives-​ strengthening approach, including “enforced emancipation” (Jørgensen 2009) or “disempowered empowerment” (Nunn and Morgan 2018) in order to activate “passive” benefit recipients. Regardless of its interpretation, simply changing the laws of benefit receipt and active labor market policy (ALMP) would not suffice to deliver practical, “street-​level” results. Rather, the public employment service (PES), the government’s most important delivery arm of labor market policy, needed an institutional overhaul to become an agent of change. Policy makers were eager to adapt their PESs not only to implement the activation paradigm, but also to modernize their operative and delivery functions, often in the name of New Public Management (NPM) (Schram et al. 2010; Nunn 2019; Weishaupt 2010b). Modernization seemed necessary to adapt to new labor market situations that arose from the transition toward knowledge-​and service-​based economies, the persistence of long-​term/​structural unemployment and rising numbers of social assistance recipients, and tightening public budgets in an age of “permanent austerity” (for the latter, see Pierson 2001). A wave of incremental and radical PES reforms followed in the 1990s and early 2000s that often shared several tendencies. • The introduction of new performance management systems through management by objectives (MBO) (Castonguay and Sol 2010; Mosley et al. 2000); • The individualization and contractualization of staff–​client relations (Sol 2010; Sol and Westerveld 2005; van Berkel and Valkenburg 2007) • The decentralization of administrative tasks (at times in the context of recentralization authority) (Knuth and Larsen 2010; van Berkel and Borghi 2008a) • Increased interagency cooperation and the creation of so-​called one-​stop shops, whereby various services ranging from job placement to benefit receipt are being integrated, often in the context of harmonizing different types of benefits and a recategorization of risk (Clasen and Clegg 2011) • Tightened and frequently time-​limited benefit eligibility (Immervoll 2010; Weishaupt 2013) • Privatization and marketization, including both competitive tenders when it comes to the delivery of services and the (partial) outsourcing of PES tasks to private providers (Bredgaard and Larsen 2008; Finn 2010). While governments in Europe shared the same vision and faced the same challenges, their reform trajectories were “variegated” due to differing

206

206  Work and the Social Safety Net institutional starting points, political ideologies, and the number and strength of the actors involved in designing the reform steps (Nunn 2019). Moreover, the reforms are a continuous process, leading to frequent adaptions and more radical reform steps after new governments are sworn into office. Given the perpetual reform of PES and the (varied) impact of the Great Recession, we wish to make three contributions to the literature. First, we offer an in-​depth review of prominent PES reform in three countries. Second, we critically evaluate the (unintended) consequences of reforms taken in light of the available evidence while also using the Great Recession as a “stress test.” Finally, we identify and discuss common and diverging trends in PES reforms. The country cases selected for review include three high-​profile crucial cases which have witnessed substantial PES reform and are seen as relatively strong cases of PES management, organization, and innovation (Nunn 2012, 2013). Denmark is the benchmark in the European Union and Organisation for Economic Cooperation and Development (OECD) for seemingly successful “flexicurity” policies (Weishaupt and Lack 2011), combining Anglo-​American flexibility with Scandinavian levels of employment rather than job security. The United Kingdom was, in turn, a leading implementer of “third way” active social policies during the Blair era and was the beacon for neo-​conservative liberalism after the election Prime Minister David Cameron. The United Kingdom is also interesting because the implementation of ALMPs coincided with the emergence of the UK labor market from a period of poor employment performance in the 1980s and early 1990s to being seen as a success story in the mid-​2000s. Despite this, both Denmark and the United Kingdom suffered from increased unemployment rates during the Great Recession and in its aftermath. Germany is chosen because it, too, reinvented itself from an employment laggard to a “miracle” performer (Krugman 2009). In contrast to the United Kingdom, however, the labor market’s performance improved during the Great Recession and its immediate aftermath. The evidence for this book chapter is mainly drawn from a careful review of existing reports and academic publications as well as a series of expert interviews conducted by the authors in recent years.

THE GERMAN CASE Reinventing German Labor Market Governance: Radical Departure, Smooth Ride? After four relatively calm years in government, the Social Democratic-​Green Party coalition government took the initiative to overhaul the German PES in the context of the so-​called Agenda 2010 and the “Hartz reforms” in 2002–​2003. These reforms were introduced as a “big bang,” simultaneously restructuring the German unemployment insurance and social assistance systems, introducing

20 7

Activation in PES in Europe  207 a new set of labor market instruments (promoting especially atypical work in the fast-​growing service sector), and embracing the new activation leitmotif for the long-​term unemployed (Weishaupt 2010a). The reforms were intended to reverse the decades-​long trend of increasing long-​term unemployment after each economic downturn since the 1970s and the associated “inactivity” of large sections of the population. This was not only a socially undesirable situation, but high levels of often persistent unemployment and inactivity also negatively affected German competitiveness as payroll contributions, especially those for unemployment insurance, reached the highest level in Germany’s postwar history at 6.5%. Moreover, the government was concerned with the growing number of social assistance recipients who received labor market integration services solely from municipal authorities. This had several disadvantages, including heterogeneity of service provision (an equity concern) and their exclusion from the unemployment statistics as they were not registered with the PES and thus remained under the public radar (a social justice concern). In line with the wider European trends at that time, the government’s reform agenda was driven by an ideology that prioritized NPM solutions to governance questions (Weishaupt 2010b) and the activation of benefit recipients through a rebalancing of rights and duties (Weishaupt 2010a, 2011a). While the former included the belief that public-​sector institutions would run more efficiently with an “entrepreneurial spirit,” the latter subscribed to the idea that Germany’s social welfare system was overly passive, discouraging self-​efficacy and promoting benefit dependency. Successor governments continued this path without major changes to the German labor market regime. The PES itself, however, sustained the reform process internally in the name of “effectiveness” (generating and improving quantifiable results) and “efficiency” (delivering more with less financial resources), as outlined below.

Changing Governance Structures Following the Bismarckian welfare trajectory, the German PES (Bundesagentur für Arbeit [BA]) is a self-​governing, quasi-​independent body created under public law (§367 of the Social Code Book III). The BA is run like a joint-​stock company and commands its own budget, primarily financed through mandatory payroll contributions paid in equal parts by employers and employees. The BA has a mandate to deliver integrated, multifaceted services ranging from the administration of unemployment insurance tasks to job counseling and placement and the realization of active labor market policy and lifelong learning measures. The Federal Ministry for Labor and Social Affairs (BMAS), in turn, issues qualitative “framework objectives,” which set the priorities for labor market policy. Germany is a federal country, and the organization of the BA is structured along three tiers of governance: national headquarters (located in Nuremburg), ten

208

208  Work and the Social Safety Net regional directorates (some Länder/​city states are merged into one region), and 156 employment agencies (plus some 600 additional business offices). The Hartz reforms induced inter alia a series of changes at the operational level. Besides introducing a new name—​Federal Employment Agency—​the new PES is headed by a three-​person executive management board, representing the BA and managing its daily operations, and a twenty-​one-​member (previously fifty-​one) Board of Governors (Verwaltungsrat), the BA’s main “monitoring, advisory, and legislative body,” responsible for the BA strategic decisions (see BA Satzung, Article 3, author’s translation). The management board consults the tripartite board of governors on all important issues, and decisions are typically made unanimously. The board of governors includes seven representatives from the trade unions, seven representatives from employer associations, and seven representatives from public authorities at the federal, regional, and local levels. The BA is run with a comprehensive MBO-​system that allows the management board to gain insights into the performance of all local offices and their staff. This is expected to maximize transparency and to allow for a systematic use of benchmarking of PES offices across Germany with similar labor market situations. As a result, managers of well-​ performing PES offices are rewarded financially while poor performers are required to learn from their peers (Nunn 2012, 2013). The regional directorates are also managed by three-​member boards, appointed by the management board. The regional directorates mainly function as “transmission belts” between the BA headquarters and the local PES offices, translating national targets into regional and local ones while also monitoring, communicating with, and giving advice to local PES offices. The regional directorates also help with the planning and implementation of the federal Länder governments’ labor market programs. The social partners are consulted in advisory boards, which were (re-​)institutionalized in 2012. At the local level, labor market services are offered by either employment agencies (Agenturen für Arbeit, local PES offices responsible for the provision of services to all unemployment insurance claimants) or Jobcenters (a separate office, responsible for long-​term unemployed job seekers and all other recipients of Germany’s minimum income scheme). The BA is directly responsible for all of the employment agencies, which operate under a high degree of local autonomy with respect to the use of their budgets and available instruments provided they fulfill their targets (Jakob 2006, 43; Weishaupt 2014). The social partners are represented in local governance committees and advise the local management boards in all questions regarding local labor markets and PES performance. The Jobcenters, in turn, are run either by collaborations (Gemeinsame Einrichtungen) between the PES (essentially providing staff and expertise with regard to employment policy) and municipalities (responsible for all other, mostly social aspects) or by municipalities alone (representing about 25% of all Jobcenters). The BMAS steers all Jobcenters, using the BA only as a service provider in the collaborations. Hence the BA’s tripartite board of governors

20 9

Activation in PES in Europe  209 has no advisory function and no rights to information in regard to Jobcenter clients (Adamy 2006). The government justified this decision by reference to the funding of the Jobcenters: general taxation rather than payroll contributions. The social partners, however, are represented in advisory councils, enjoying a statutory right to advise the Jobcenter managers. While the possibility for municipalities to “opt out” from cooperation with the PES was the result of a political compromise between the Social Democrats and the Christian Democrats—​with the latter favoring the municipal option—​the split in employment agencies and Jobcenters reflects the particularity of the new benefits regime that emerged out of the Hartz reforms. In the past, long-​term unemployed persons whose unemployment benefits expired—​after up to thirty-​two months, depending on age and contributory history—​would receive a means-​tested but income-​related unemployment assistance payment. This payment was paid out by the BA offices and the long-​ term unemployed job seeker would remain under the supervision of BA staff. Municipalities, in turn, were responsible for all other persons in need. The municipalities would issue means-​tested social assistance payments and often also launched their own ALMPs. As a result of the reforms, unemployment benefits would only be paid out for a maximum of twelve months (longer durations are still possible for older workers), while the unemployment assistance and social assistance payments were merged into a new minimum income scheme called “unemployment compensation II” (Arbeitslosengeld II). Under the Hartz reforms, the former unemployment assistance scheme—​which job seekers in need received after their unemployment benefits expired—​was merged with the tax-​funded, locally administered social assistance system in 2005. As a result, Germany has a two-​tiered benefit system: a payroll-​financed unemployment insurance system and a tax-​financed minimum income scheme (popularly known as Hartz IV).

Changing Routines: Improved Customer Orientation and More Choice (Through Vouchers)? The Hartz reforms also modernized the daily operations of the local PES offices. On the one hand, all employment offices were remodeled into agencies with new customer centers. The centers include open windows for walk-​in customers, while appointments for benefit recipients are made in advance to minimize waiting times. On the other hand, administrative tasks and placement work were formally separated to maximize efficiency of the new case managers, who were relieved of much of the administrative burden. Moreover, new profiling procedures were introduced and subsequently improved. More specifically, case managers, following central guidelines, rely on up to one-​hour-​long, in-​depth, computer-​assisted initial profiling interviews, which they apply to customers of employment agencies and Jobcenters alike.

210

210  Work and the Social Safety Net The profiling system seeks to identify both job seekers’ entry hurdles and capabilities, which in turn affects the intensity and type of future activation offers (Bundesagentur für Arbeit 2012, 11–​12). Since 2010, the profiling system differentiates between those job seekers who are likely to engage in gainful employment within twelve months (profile category “close to labor market integration”) and those who are unlikely to do so (profile category “complex situation”). Each of these categories is divided into three subcategories based on how “close” job seekers are to integration or how “complex” is their situation. Case managers have full discretion as to what instruments they consider appropriate to offer given an established profile and local labor market situation. The offers typically range from short-​term training measures (soft skill and vocational education preparatory courses) to further vocational education to wage subsidies (or, in the case of long-​term unemployed, so-​called short-​term reintegration offers). Referrals to vacancies, however, are not affected by profiling categories as PES officers always try to offer employers the “best” candidates in order to build trust and long-​term relationships. Improving contact with and services for employers has generally become a central concern for the BA. In response, all local PES offices created new units that cater solely to the needs of employers, which are—​based on customer satisfaction surveys—​well received by most (predominantly large) firms. While “creaming” is thus a routine and deliberate practice, weaker job seekers receive activation offers in order to avoid “parking” and long-​term unemployment. Particularly among the minimum income benefit recipients, this has resulted not only in improved services and choice, but also in tighter monitoring, low-​quality activation offers that “can’t be refused” and harsher sanctions. Similarly, for many long-​term unemployed job seekers, the only training options available are group-​based—​in contrast to individual—​ soft-​skills courses, while reintegration offers are often only based on six-​month-​ long employment opportunities, which are remunerated with one euro per hour in addition to benefits. As these “one-​euro jobs” offer little prospect for progression—​and indeed are perceived as degrading by many participants—​ many scholars interpret Germany’s welfare reforms as a step toward a “work-​ first regime” (Bruttel and Sol 2006; Eichhorst and Marx 2011, 78). The former BA Chairman Frank-​Jürgen Weise also publicly criticized these one-​euro jobs and—​unsuccessfully so—​lobbied for their replacement with wage subsidies to compensate employers for these workers’ deficiencies in productivity. Another aim of the reform was to maximize job seekers’ choices through the introduction of two types of vouchers: education (further vocational training) and placement. While further vocational training has always been delivered by partner organizations (Bildungsträger), the Hartz reforms changed the way in which contracts were awarded. Rather than PES offices referring job seekers to a specific program offered by a contracted further vocational training provider (Förderung beruflicher Weiterbildung [FbW]), eligible job seekers receive an education voucher with which they may purchase program slots with accredited

21 1

Activation in PES in Europe  211 providers. The introduction of a voucher system had two purposes: empower benefit recipients by giving them the opportunity to select their own providers and, relatedly, introduce a demand-​driven market for training programs. While the introduction of a demand-​based market successfully ended a discussion about alleged favoritism in the way contracts were awarded—​namely that the social partners involved in regional PES governance systems awarded contracts to those providers that were linked to employers’ associations and the trade unions—​it remains unclear if the change also improved customer choices for two reasons: first, in many locales there is only one provider (i.e., no choices possible); second, recipients, especially those with lower qualifications, tend to feel “overburdened” to choose a course or provider (Koch et al. 2011, 4). With the introduction of placement vouchers, job seekers may purchase the services of a private placement agency after an initial phase of unemployment. While insured job seekers have a right to receive vouchers when certain conditions are met, minimum-​income recipients receive vouchers only at the discretion of case managers. The effect of these vouchers on improved choice remains questionable as their use was limited (Bernhard and Kruppe 2010) and their placement success was highest for the easiest-​to-​place group: short-​term unemployed job seekers with good qualifications (Koch et al. 2011, 2). Private placement agencies receive partial payment six weeks after initial placement and again partial payment when jobs are retained for at least six months; a bonus is paid for placement of customers with complex profiles. Since 2013, private placement agencies accepting vouchers also need certifications from the BA. With respect to placement, in April 2007, the BA also signed a cooperation agreement with Germany’s fifteen largest temporary employment firms. In contrast to other countries, such as Austria, private placement companies are separate from temporary employment agencies. The former only acts as a job broker, while the latter issues a work contract (and then temporarily “lends” the worker to another firm). Hence, temp agencies cannot cash in vouchers. The hope was to improve service provision (increased capacity through the availability of more job offers) and offer “trampolines” into the labor market (while avoiding the scarring effects of long-​term unemployment). However, the cooperation with temp agencies has received widespread criticism as job seekers are placed quickly into unstable jobs. Based on an unpublished discussion paper by the head of the BA’s employee committee, placement of job seekers into fixed-​term employment increased substantially in the aftermath of the Great Recession: in some local PES offices, nearly 70% of all job placements were into temporary jobs (Einsiedler 2012, 4). The problem is that these placements have not been trampolines into the regular job market because only 8% of placed workers were in permanent contracts after two years (Crimmann et al. 2009, 88). Other studies similarly showed that, despite a much different political rhetoric, a seamless transition from temporary work to a permanent contract with the same employer occurred for only 7% of all temp workers (Baumgarten et al. 2012). In response, the BA

21

212  Work and the Social Safety Net chairman at that time reacted to this criticism by publicly stating that this was an undesirable outcome and that performance targets needed to be adjusted to minimize incentives to place job seekers into fixed-​term contracts (Spiegel Online 2013). In addition, the BA took concrete steps to promote closer cooperation between the BA and small and medium enterprises in order to acquire larger contingents of regular job offers (Interview, PES 2013) while also placing more focus on customer satisfaction (Interview, PES 2021).

INTRODUCING QUASI-​MARKETS: COMPETITIVE TENDERS FOR ACTIVATION MEASURES The introduction of education vouchers was only one element in the changes of awarding contracts to service providers. Local PES offices also purchase a variety of other courses from third-​party providers, including the services of assessment centers that provide specialized courses for persons with disabilities, lifelong learning measures, or courses for young or low-​skilled persons who are not yet prepared to attend vocational training courses. The local PES—​in consultation with the social partners—​estimates the number of places needed for such programs, and the regional “purchasing centers,” which are attached to the regional directorates, then select providers based on a process of open, regionally based competitive tenders. Selection criteria include both price and quality of the service offers. In order to guarantee high levels of quality standards, the service providers need to acquire certificates. These certificates are issued by independent entities, which themselves are certified by the BA. Providers of these training services—​just like providers of further vocational training—​are paid on an output basis, but performance indirectly plays a role as tenders are reissued periodically. With respect to further vocational training, the introduction of a competitive tender process and the issuance of vouchers dramatically changed the service provider “landscape” as many (often smaller) institutions went bankrupt during the initial years. While the German Confederation of Trade Unions (DGB) regretted this development, the BA argued that the new system is more transparent and much more cost effective (Interviews PES 2007; DGB 2007). Over time, the provider market stabilized as many smaller service providers engaged in “consortia” which enabled them to compete with the larger providers (also on a regional level). Some concern remains, however, that regionally based tenders may lead to the selection of providers that are not rooted in the local economy. Partially in response to this criticism, in 2012, the government introduced a voluntary voucher system for these kinds of services on which employment agencies and Jobcenters may rely instead of the tender process. Finally, some concern remains that the tender-​based system takes more time and is thus less able to quickly

21 3

Activation in PES in Europe  213 and flexibly react to changing situations, and—​perhaps more importantly—​that trainers and educators are increasingly employed “precariously.” In other words, they are self-​employed and hired on an ad hoc basis or receive fixed-​term, often low-​paid service contracts (Bundesregierung Deutschland 2010). To date, these controversies remain unresolved.

The Impact of the Global Financial and Economic Crisis In contrast to most other European economies, Germany coped with the crisis surprisingly well (Leschke and Watt 2010). While the government provided additional funds to hire more BA staff as part of their stimulus packages, the key ingredients of the German success was the concerted effort to preserve existing employment, especially via short-​time work, or Kurzarbeit, which was expanded (through more, easier, and longer access) and made more generous (as the BA also paid employers’ payroll contribution). Kurzarbeit allows companies to reduce the working hours of their employees (and thus reduce wage costs) while workers receive a compensation from the BA for the loss in income (and an opportunity to partake in training measures). In practice, the German strategy to protect existing jobs meant that unemployment grew only marginally in 2009—​ despite a drop in GDP of 5%. Most job losses were in the temp-​job sector, which, however, quickly recovered in 2010. As such, the Jobcenters in charge of long-​ term unemployment were not confronted with a significant influx of new benefit claimants. The scope (and acceptance) of the Kurzarbeit scheme was made possible by the institutionalized involvement of the social partners who swiftly found a common position, publicly endorsed the policy, and subsequently informed all relevant bodies and actors. As a result, the BA not only informed their staff in the regional and local offices, but the social partners also informed their Land associations and their members in the local Verwaltungsausschüsse about the new rules and opportunities of the short-​time work program (Interview, Confederation of German Employers’ Associations 2010). This “dual communication strategy” was crucial for the success of the Kurzarbeit scheme (Interview, PES 2013).

Summary and Future Challenges The German PES has undergone a phase of profound structural and procedural changes. The reform efforts were conducted with the explicit aim to maximize both effectiveness (outcomes) and efficiency (resources management). With respect to these aims, the PES can claim quite noteworthy results, including a significant reduction of long-​term unemployment and quicker job placements while gradually downsizing its budget and thus payroll contribution burdens for employers and employees alike. To what extent the introduction of quasi-​markets and vouchers have yielded the intended consequences of higher quality services,

214

214  Work and the Social Safety Net more choice, and lower costs remains unclear. While privatization of services remains marginal, private placement agencies increase the overall placement capacity. In the aftermath of the financial crisis, the new BA structures have proved themselves resilient and capable of delivering results. The continued reliance on and inclusion of the social partners in the PES’s governance structures turned out to be an asset, stressed by all involved parties and the BA chairman himself (speech given by the BA Chairman at SEEK-​C onference, April 25, 2013, hosted by the Mannheim Center for European Economic Research). The most recent experiences during the COVID-​19 pandemic have proved, once more, that the BA is able to react swiftly to a changing environment. Similar to, and in effect inspired by, the response in 2008–​2009, the government decided to expand ease access to short-​time work, which the BA transposed efficiently (Weishaupt 2021). This time, however, not only manufacturing firms were affected, but almost the entire economy, including the service sector. At its peak in April 2020, more than 6 million workers in some 600,000 firms received short-​time work payments administered by the BA. In addition, the pandemic placed the BA digitalization strategy on a fast track; it included offering “blended services” (combining online, video-​based and face-​to-​face services to clients) and developing more fine-​tuned forms of matching vacancies with job seeker competences (Interview, PES 2021).

THE UK CASE Two Phases of Reform: Toward More Workfare The delivery of ALMPs in the United Kingdom has gone through two distinct phases. First, up to around 2005–​2006 ALMPs and Jobcentre Plus (JCP; the UK PES) were tasked with dealing with some of the structural labor market problems left behind by neoliberal restructuring in the 1980s and ’90s. This period was marked by policies designed to overcome benefit/​poverty traps, incentivize work (“make work pay”), and activate the unemployed. Unemployment fell and the United Kingdom was positioned among the better performing Member States in the European Union. In the second period, after 2005–​2006, reforms to the benefit system and the organization of JCP attempted to increase an already high level of conditionality and privatization and incentivize autonomy and flexibility at the frontline of service delivery (Nunn 2008). This second phase of ALMP reform was impacted by the economic crisis (since 2008) and the election of a new coalition government of the Conservative and Liberal Democrat parties in 2010, which targeted welfare reform as one key way of reducing public spending. Despite this, the substantive trend in welfare reform during this period was one of continuity rather than change.

21 5

Activation in PES in Europe  215

Activation and Tightened Conditionality The first trend to note in relation to the delivery of employment services in the United Kingdom was a continuation and intensification of the “work-​ first” orientation of ALMPs and progressively tightening conditionality (Finn 2011). After the Freud Report of 2007 (Freud 2007), conditionality and discipline were progressively and markedly tightened.1 For example, in controversial reforms (see McHardy and Fife Gingerbread Community Researchers 2013; Save the Children UK 2012), conditionality was substantially in relation to lone parents, who faced some work-​related conditionality when their youngest child reached the age of one year and who were required to actively seek work when their youngest child was five years old (down from age sixteen in 2008) (see McGuinness and Harari 2013; Department for Work and Pensions 2012b). Similarly, conditionality and expectations were enhanced for those previously on sickness and inactive benefits, with many who had previously been able to claim inactive benefits reclassified as fit for work, a process labeled by some as “welfare retrenchment” (Grover and Piggott 2010; Piggott and Grover 2009; Bambra and Smith 2010). More generally there was controversy about the use of targets for the number of sanctions by JCP staff (Wintour 2013; Nunn and Devins 2012, 33), the introduction of a new hierarchy of sanction penalties (Department for Work and Pensions 2012b), and the use of Mandatory Work Activity and unpaid placements. This controversy led to the government distancing itself from these practices, but the punitive trajectory in welfare has remained, and, in some ways, expectations about job search have even been tightened, including expectations placed on people in work in return for access to in-​work benefits.

Governance and Steering These changes were supported by an evolving performance management system. While performance management regimes may seem dry and technical, they are often revealing of the political objectives behind policies that politicians often justify rhetorically as being in the universal interest (Nunn and Morgan 2018). In 2005–​2006, the old Job Entry Target (JET) was replaced by a new headline labor market target: the Job Outcome Target (JOT), which, in contrast to JET, allocated points to a local area (rather than an individual adviser) for all clients who entered work, regardless of whether an intervention had been instrumental in that transition or had even been provided at all. This was intended to remove any incentive to work with clients other than the hardest to help and also to reduce perverse incentives that were thought to exist prior to this time. Though difficult to verify, there was widespread suspicion that the requirement under JET to prove an intervention led to unnecessary activity to “claim” off-​flow 1. It is notable that Freud—​previously an investment banker—​was initially commissioned by the Labour government to provide a review of the future of welfare provision and was subsequently appointed as a Conservative welfare adviser.

216

216  Work and the Social Safety Net performance that would have happened without an intervention. However, despite positive early evaluation data (Johnson and Nunn 2005, 2006, 2007; Nunn 2007) JOT itself was subsequently criticized for being too slow to reflect performance (Nunn et al. 2010; Nunn et al. 2007). Prior to around 2008–​2009, JCP activities were heavily regulated and monitored according to a standard operating model (which ran some seventy pages). This set out what interventions (e.g., work-​focused interviews) were to be delivered to specific groups of clients within specified time periods. Performance against these expectations was monitored according to an “interventions delivery target” (IDT). Because JOT could not provide timely outcome information about the number of job seekers who had found work and could not analyze the relationship of flows from benefits to work as a result of a JCP intervention, managers instead resorted to managing performance via the IDT (Nunn and Devins 2012). The result was a widespread acknowledgment that the performance management system was once more creating perverse incentives to deliver expensive services to job seekers who frequently did not need them. This was compounded by research evidence that suggested that the performance advantage of private employment services (Hales et al. 2003) resulted from an increased freedom to innovate to meet the specific needs of individual job seekers (Department for Work and Pensions 2009b; Davies 2008). The result was an increasing emphasis on enabling flexibility and frontline autonomy in the delivery of employment services (Department for Work and Pensions 2009a). This was delivered in a variety of ways, as through the introduction of the Delegated Flexibility Pilots, later rebranded several times and rolled out nationally (Public and Commercial Services Union 2012). These changes promised flexibility to innovate at the local level through changes to the organization of offices, adviser (now “work coaches”) time, and even contracts with local service providers in an “earned autonomy” framework which rewards high levels of performance with increased flexibility to vary standard practices. It is too early to provide hard evidence on how these flexibilities are operating, but they are consistent with wider supportive changes. One of those wider changes was the introduction (in 2011) of yet another new performance management framework (PMF). The PMF reduced the previous six headline performance targets to two: moving people from benefits to work as soon as possible and reducing the value of fraud and error in the administration of the benefits system. The first substantially reinforced the “work-​first” message through JCP and can be seen as part of a wider move among PES in Europe to focus not just on incentivizing transitions from unemployment to work, but also the speed of these transitions (Nunn 2013). The PMF thus included a new indicator of off-​flow rates from benefits by particular milestones in welfare benefit claims (thirteen weeks, twenty-​six weeks, etc.). While it was promoted as reducing central targets and increasing autonomy for advisers, it was accompanied by a voluminous spreadsheet of performance expectations along the “customer

21 7

Activation in PES in Europe  217 journey” which include measures of activities and processes. The initial evaluation of the PMF suggested that it was partially successful in its own terms, but also that there were very few resources dedicated to training staff or managers on how to change their behavior, and there was some evidence of advisers being managed at the local level by process and activity targets (including sanctions). The evaluation concluded that there was a considerable distance to travel before the cultural emphasis on targets and processes in JCP was shifted to a more outcome-​oriented approach (Nunn and Devins 2012).

Internal Processes and Resources Since 2000, JCP has been under almost constant structural change, beginning with the merging of the previous Benefits Agency and Employment Service to create a single arms-​length body to administer working-​age welfare payments and ALMPs (National Audit Office 2008). New estates policies, performance management systems, the introduction of new IT systems, and a vast array of complex and successive welfare benefit reforms and ALMPs all put considerable strain on the organization (Casebourne 2012). For example, throughout and slightly preceding the onset of the economic crisis in 2008, JCP underwent a series of structural reforms, including staff reductions (prior to 2009–​2010), and several benefit delivery centres were closed. Cumulatively these changes resulted in service delivery problems which were widely acknowledged and contributed to a strained relationship between JCP, Parliament, and various stakeholder organizations (House of Commons Work and Pensions Committee 2007; Citizens’ Advice Bureaux 2007). Despite this, structural changes continued. In 2011, both JCP and the Pensions Service were reintegrated into “One DWP” so that the PES was no longer a separate organizational entity. This might seem at first a significant change and certainly goes against the tide of NPM-​style reforms in PES which tend to see the separation of the PES from the labor ministry. The precise motivations for the change were ambiguous, other than cost reductions, and they reflect a general antipathy in the Coalition government for “QUANGOs.” (QUANGOs are quasi-​ autonomous nongovernmental organizations and were initially introduced in the 1980s in the first wave of NPM reforms—​part of the “steering not rowing” reform agenda of Whitehall. Before coming to power in 2010, the current Conservative government promised a “bonfire of the QUANGOs” as part of a critique of the previous government’s public spending.) In the case of JCP, it was not entirely clear what impact this change had. Operational staff in Jobcentres were largely untouched by the change, and central staff in research, analysis, and management were co-​located—​often in the same rooms and at adjoining desks—​with DWP colleagues. There was also a high degree of interchange of staff. Perhaps the more important organizational change at the same time was the centralization of performance analysis staff to encourage greater sharing of good and bad practices and performance variation.

218

218  Work and the Social Safety Net A further ongoing structural theme was a gradual shift in the ways that employment services were provided by JCP, with “customers” increasingly encouraged to use “self-​service” channels when making a claim (by telephone and later online) and searching for vacancies. This was justified as focusing public resources on those who needed help the most.

Delivery and Privatization JCP is notable among European PES for being one of the most privatized systems. Private-​sector involvement has always had a significant role in the delivery of PES in the United Kingdom. Through a variety of schemes—​such as employment zones—​the degree of private-​sector involvement gradually increased as private providers increasingly took on the delivery of counseling and referral from JCP, usually for job seekers satisfying particular criteria, such as a particular duration of unemployment or personal characteristics that gained them “early entry.” The employment zones were noteworthy among these because they had elements of competition: payment by results for both outcomes (i.e., transitions from unemployment to work) and sustainability (i.e., staying in work for a defined period). Evaluations of these interventions suggested that they had some modest advantages (Hales et al. 2003), mainly because of their freedom from the standardization that governed JCP delivery (Department for Work and Pensions 2009b; Davies 2008). Critics worried about the introduction of creaming and parking (Bruttel 2005) and the disincentives that these schemes created for employers in relation to improving the quality of work, retention, and upskilling (Policy Research Institute 2006, 95). Furthermore, private provision was involved in around 60% of the Pathways to Work scheme which targeted those who had previously been on inactive benefits for reasons of ill health. Evidence suggested that this scheme underperformed against expectations and that private provision was less successful than that delivered by JCP (House of Commons Committee of Public Accounts 2010). Following the 2007 Freud Review (Freud 2007; Nunn and Johnson 2008; Nunn 2008), it became clear that future policy development would increasingly rely on private delivery. This occurred first in the Flexible New Deal (FND) and then subsequently in the Coalition Government’s replacement for FND: the Work Programme. While the Coalition government emphasized the differences between the two schemes, they were in fact very similar. Both adopted a prime contractor model, both offered significant payments for job outcomes and sustainability, and both also adopted a “black box” approach designed to foster innovation in frontline service delivery. The main differences were related to the level of obligations and pricing structure. FND was the fourth stage of the so-​called Job Seekers Regime, which involved unemployed benefit claimants working with JCP on a range of specified interventions until their benefit claim reached twelve months. At that point, job

21 9

Activation in PES in Europe  219 seekers were passed to private-​sector providers who received service delivery payments, outcome payments, and employment retention payments. Under both FND and the Work Programme a variety of categories of job seekers (younger job seekers, those making repeat claims, lone parents) were also able to access the program earlier than the twelve-​month point in their claim, bringing intensive activation further forward in the claim. A significant problem with FND was that it was commissioned on a competitive basis whereby prime contractors bid to run the program with a range of subcontractors at a local level. The commissioning process straddled the onset of the economic crisis and a substantial change in labor market dynamics, meaning that the expectations of the DWP and bidding contractors about inflows and outflows did not match, and, as such, the attempt to “make a market” substantively failed (House of Commons Work and Pensions Committee 2009). The Work Programme was intended to resolve some of these difficulties, but retained the prime contractor model. There were eighteen “primes” operating forty spatially defined Work Programme contracts. They were in subcontracting relationships with more than 800 suppliers of employment services in their supply chains. Fifteen of the eighteen prime providers were private-​sector organizations, with these being split between specialist welfare-​to-​work service providers (e.g., Working Links, A4e) and larger corporations which have substantial public-​sector contracts in a range of other sectors and countries (e.g., Serco, G4S). There was only one public-​sector (though this contract has now been terminated), one voluntary and community sector, and one mixed sector provider. Subcontractors were involved in the initial procurement process through “expressions of interest” to organizations bidding to be prime contractors (Lane et al. 2013). Subcontractors were organized into two tiers: Tier 1 contractors provided end-​to-​end services and are more likely to be larger private-​sector organizations with contracts determining a minimum number of referrals. Tier 2 subcontractors tended to be smaller and more specialized in nature and are used on more of a spot-​contracting basis for specific services at a point in the job seeker’s journey. Tier 2 subcontractors were more likely to be in the voluntary and community sector, though they witnessed much lower levels of referrals than Tier 1 providers, especially in the early years of the Work Programme, which were dominated by high levels of referrals of job-​ready and near job-​ready job seekers (McGuinness and Dar 2014). There was some controversy that smaller, specialized voluntary-​sector organizations were used as “bid candy” by prime contractors and that these organizations sometimes did not even recognize that they were part of the supply chain, reporting low levels of referrals or none at all. However, the evaluation found no evidence that this was the case and attributed low referral levels to the nature of the job seekers in the system during the period of high unemployment (Lane et al. 2013). It may be that the outcome-​ based contracting model and difficult labor market conditions at the outset of the Programme meant that primes and Tier 1 providers have preferred to keep service delivery in-​house to maximize their returns (McGuinness and Dar 2014).

20

220  Work and the Social Safety Net Job seekers were referred to the Work Programme by JCP at different milestones in their period of unemployment or inactivity. This ranged from the moment they claimed benefits (for people leaving prison, homeless people, or people leaving care) to twelve months (for Job Seeker’s Allowance [JSA] claimants older than twenty-​five). While there was a national system of rules about referral deadlines, there was also some local flexibility for JCP to make early referrals in some cases (McGuinness and Dar 2014, 3). The Work Programme is structured around a payment-​by-​results system with a varying tariff of payments for different groups of job seekers but with higher payments for outcomes rather than services (see Table 9.1). Differential payments were intended to reflect the difficulty and likelihood of successfully placing different groups of job seekers into work and help to overcome pressures to “cream and park” customers. They were also designed to reward providers for success in helping people find and sustain work. Table 9.1  Work Programme payment structure (summary maximum year 1 payments) in 2014 Job seeker group

Max year 1 attachment fee

Job seekers allowance 18–​24

£400

Job seekers allowance 25+​

£400

Job seekers allowance early Access/​Ex-​incapacity benefita

Max year 1 job outcome fee

Max year 1 sustainment fee

Total

£1,200

£2,210

£3,810

£1,200

£2,795

£4,395

£400

£1,200

£5,000

£6,600

Employment and support Allowance voluntary referralb

£400

£1,000

£2,300

£3,700

New employment and support Allowance claimants

£600

£1,200

£4,700

£6,500

£600 Employment and support Allowance who have progressed from incapacity benefitc

£3,500

£9,620

£3,720

Incapacity benefit/​income supportc

£400

£1,000

£2,300

£3,700

Job seekers allowance prison leavers

£300

£1,200

£4,000

£5,500

Source: Adapted from McGuinness and Dar 2014. a This group will have previously been claiming the Incapacity Benefit, but under new rules and Work Capability Assessments have been judged able to work. b ESA claimants are placed into different groups based on their ability to move into work. Some ESA claimants voluntarily ask for referral to the Work Programme, in advance of being required to do so. c The entire previous Incapacity Benefit caseload is being subjected to work capability assessments. Not all claimants had been assessed at the outset of the Work Programme. Following the assessment, they could either be placed in one of the ESA groups or onto job seeker’s allowance where they are expected to actively seek work.

Activation in PES in Europe  221 It was also less prescriptive about the services that providers had to offer and was therefore closer to the pure “black box” approach. This made it difficult to establish what services specifically were being provided by different providers and how these related to outcomes. However, the evidence suggested that the services offered to job seekers were very much in line with the type of “light touch” approach previously applied by JCP: adviser discussions about job aspirations, help with job search, and CV preparation (House of Commons Work and Pensions Committee 2013). Early evidence from the official evaluation suggested that there was very little substantive personalization in service delivery (Newton et al. 2012). Despite the effort that went into contract design, there were concerns that the Work Programme, like FND, had optimistic expectations (National Audit Office 2012). It is notable that there was controversy about fraud in the payments system, inappropriate unpaid mandatory work activity, and the inappropriate use of sanctions (Public and Commercial Services Union 2013). The evidence suggested that a combination of difficult market conditions and high work volumes meant that creaming and parking was a feature of the delivery of the program, that provider staff were disappointed at the quality of services being provided (Newton et al. 2012; House of Commons Work and Pensions Committee 2013), and that the program failed to meet its intended targets in the first two years (Centre for Economic and Social Inclusion 2012; House of Commons Committee of Public Accounts 2013; Deparment for Work and Pensions 2013b), as the National Audit Office suggested it might. Performance data in later years were more encouraging (Department for Work and Pensions 2014) but were affected by improving labor market demand and, as a consequence, a declining number of JSA claimant referrals and increasing caseloads of comparatively hard-​to-​help Employment and Support Allowance (ESA) claimants. An analysis of the evidence about the operation of the Work Programme suggested that it worked better for relatively mainstream job seekers and less well for harder to help groups with more specialized needs (House of Commons Work and Pensions Committee 2013). In part, changes to delivery reflected this problem by putting in place services for those job seekers who exit the Work Programme without having found employment. While the government repeatedly defended the Work Programme, these challenges ultimately led to its abandonment. While many aspects of the Programme were retained in a successor Work and Health Programme, it is much smaller in scale and the majority of delivery returned in-​house to JCP.

Coping with the Crisis The economic crisis impacted the delivery of employment policies in several ways. First, it led to increased demand as unemployment (and particularly youth unemployment) has risen. In the initial period, this increased work load was met by JCP with “borrowed” staff. Nevertheless, the increased adviser caseloads that

2

222  Work and the Social Safety Net resulted undermined the frontline flexibility initiatives outlined above. Second, it changed the characteristics of job seekers. More professional and “job-​ready customers” required employment services, thereby clouding the prior emphasis on activating those at some distance from the labor market and arguably contributing to the emerging evidence of creaming and parking (Nunn and Devins 2012, 30–​35). Finally, though not purely about the way in which services were delivered, it is notable that the most significant reforms associated with responding to the tight fiscal position of the UK government was the several waves of welfare cuts, such as changes to indexation, caps on maximum payments, and increased conditionality (e.g., Department of Work and Pensions 2012b). Perhaps the most significant challenge that arose from the crisis, as in the rest of Europe, was youth unemployment (at about 18–​19% over the course of 2012–​2013; Office for National Statistics Time Series Data: YBVQ, available at www.sta​tist​ics.gov.uk). In an immediate response, the New Labour Government introduced the Future Jobs Fund (FJF), which provided subsidies to organizations (mainly but not wholly in the public sector) to create jobs for young people who had been unemployed for six months. The Coalition government symbolically ended the FJF and replaced it with the Youth Contract (YC), a £1 billion scheme operating as part of the Work Programme, suggesting that the FJF was expensive and ineffective, though subsequent analysis (Deparment for Work and Pensions 2012a) suggested that there were net benefits from public investment in the FJF. The YC had fewer rules about the training and quality content of the jobs that were supported and was more clearly targeted at the private sector, which was offered generous subsidies for employing young job seekers. Like the Work Programme, many aspects of the YC were subsequently dropped. That said, elements of the Youth Contract have been retained in subsequent initiatives, such as the current Kick-​Start program that is part of the response to COVID-​19.

Summary and Future Challenges There are several important points to take from the preceding discussion. First, subsequent to the onset of the 2008 crisis and the change of government, the general tenor of reform has been continuity and acceleration rather than a departure from preexisting agendas. In some ways this is curious because the increasing focus on activation/​conditionality was initially conceived in the context of the need to increase labor supply in a tight labor market but has subsequently been justified against a mixture of policy messages, including active state intervention and one of discipline for an undeserving poor, which has proved divisive in terms of public and media support for ALMPs. Second, while there is continuity in objectives in welfare reform, there is a sustained trend of “restless change” in programs and interventions. For instance, the Work Programme succeeded the FND, which itself replaced earlier “New Deal” programs. The Work Programme itself has been replaced. This pattern of sustained reform through

22 3

Activation in PES in Europe  223 changing interventions is in many ways reflective of a “fail forward” approach, one in which policy and program problems lead to change but where aspects of their design pass forward to new initiatives (Nunn 2019). Third, continuous organizational change may have hindered JCP’s ability to cope with changing labor market conditions after 2009, and the organization remains under stress, with high adviser caseloads in both the public and private systems, hampering the ability of advisers to meet individual job seeker needs. Fourth, successive changes to the system of steering operational delivery have created an organizational culture in which it is difficult for the now demanded frontline innovation to occur, especially in the context of high demand for services, tight resources, and intense competition in the labor market for those needing the most help. In this context, it is perhaps surprising that there is not more debate about the prevailing “work-​first” ideology. Finally, the UK experience would suggest caution in moving forward with “black box” payment-​by-​results contracting out systems, at least until better evidence of their success or failure emerges.

THE DANISH CASE Governance Under Social Democratic Rule (1990s) and Centre-​Right Motivations for Reform During the 1990s, social democratic-​led governments developed a new policy mix of an expansive economic policy and ALMPs, one that subsequently spurred international interest in Danish “flexicurity” at the beginning of the twenty-​first century (Bekker 2012; Jørgensen and Madsen 2007; Wilthagen and Tros 2004). Combining a flexible labor market with income protection and employment promotion in a “golden triangle,” Denmark experienced a drop in unemployment from 12.4% in 1993 to 4% five years later. At the same time almost 300,000 new jobs were created and social cohesion remained well above European averages. Overall, the new set-​up looked like a big success for interventionist policies. The governance system behind the Danish model consisted of an operational and a decision-​making body: the PES (Arbejdsformidlingen [AF]), especially responsible for the members of the unemployment insurance funds (closely connected to the trade unions) and fourteen regional labor market boards with their own budgets and strong competences (Jørgensen 2002). The social partners—​called labor market parties in Denmark—​were in pivotal decision-​ making positions in these bodies; likewise, the AF employees on the ground took notice of the social partners’ interests. Two hundred seventy-​one municipalities were in charge of means-​tested social assistance payments and activation measures for the uninsured. This social democratic model of the 1990s was clearly altered after a center-​ right government came into office in late 2001. In 2003, the new government

24

224  Work and the Social Safety Net took a decisively more neoliberal approach, prioritizing increased labor supply through tougher activation measures while saving taxpayer money through new steering arrangements, marketization measures, and a shift of responsibility to municipalities. ALMP was altered into employment policy, with new priorities and a recalibration of instruments and implementation structures.

Enhancing the Labor Supply Through a Tightening of the Screw? Driven by ideological concerns, the newly elected right-​ wing government announced the goal to generate more “freedom” within the labor market and initiated a far-​reaching reform in 2002–​2003, called “more people to work.” Part and parcel of the reform was substituting longer (more expensive) training measures with a requirement for structured interviews with PES staff, intended to move job seekers more quickly into work. A series of sequential reforms followed from 2002 to 2010, with the aim of making the Danish system more job-​oriented and cost-​ efficient, effectively weakening security elements and introducing “work-​first” principles in activation (Baadsgaard et al. 2013; Betzelt and Bothfeld 2011; Dingeldey 2007; Weishaupt 2011a). The most important elements of these changes included, first, a substantial reduction in the role of training and education in offers to the unemployed. Education was clearly the most important element in activation during the 1990s, but, in 2011, it was reduced to less than 9% of the activation offers given (Baadsgaard et al. 2014). Not only was the scope of adult education and vocational training reduced by the new political priorities, but inexpensive activation offers also gained in prominence and the quality of activities was reduced. The frontline workers no longer made individual action plans with the unemployed persons, but “job plans” only. The unemployed were classified according to five (this was cut to three in 2010) groups of “matching” categories, according to skills and immediate employability. Subsequently, new methods and tools needed to describe and assess the unemployed and the sick were introduced. These included the work ability assessment method (arbejdsevnemetoden), the resource profile (ressourceprofilen), and the visitation toolbox with the dialog guide (visitationsværkstøjskassen with dialog-​guiden). The use of these methods and tools were compulsory and mandatory to the Jobcenters to standardize the meeting between citizens and the employment system. They described in detail how to assess and treat the unemployed person and how to (re)write the job plan. Cost-​benefit calculations were also part of the methodological basis of the approach. “Wicked problems” were reformulated as “tame ones” to be handled in accordance with centrally made manuals. As a result, social and private economic problems were no longer to be considered; job placement was considered “the solution” to all individual problems.

22 5

Activation in PES in Europe  225 Second, conditionality in the activation system was strengthened, affecting some of the work and practices within the employment system. Age-​related regulations were introduced, and conditionality also applied to job seekers older than fifty-​five, who were previously exempt. Youth, in turn, were required to participate in work or education offers; refusal resulted in a loss of social assistance. Third, special arrangements, meant for immigrants and refugees, were introduced. People on social assistance received lower support after six months, while immigrants received an extra low “start help”; access to these benefits required a minimum of 450 hours of work within the past two years. Similar conditionality measures affected people on sick allowances and people who had problems other than “just” the lack of a job. In essence, the work and access requirements of employees grew, but no systematic education within the system was arranged under these reforms. Fourth, control measures for staff and clients were strengthened. When embarking on NPM reforms during the 2000s, activities and operational codes changed such that the professional autonomy of Jobcenter staff was minimized. The frontline workers, of whom only about 40% had a relevant social worker education, received new instructions for how they should work with the unemployed (Baadsgaard et al. 2014). With a standardization of work and new tasks, they had less autonomy in case management. This, however, created tensions because central control and new professionalism was not easy to reconcile (Evetts 2009). Professional and ethical problems arose. The ethics of consequentiality—​ attached to peak performance of the individual Jobcenter and possibilities of supporting the economy of the local municipality—​clearly dominated those of professionalism, virtue, and discourse (Baadsgaard et al. 2014). This clash of differing ethics corresponded to recorded professional dilemmas about time, work identity, and execution of the employment efforts. The new tasks and the new regime implied a de-​professionalization of the semi-​professional employees while some administrative tasks were re-​professionalized. For job seekers, similar monitoring rules applied, and activation was redirected such that it was no longer something to “look forward to.” On the contrary, activation should—​as two economists formulate it—​be composed of “programmes that are truly cold, wet, hard, and have no skill-​enhancing components” (Rosholm and Svarer 2004, 35). Threats of sanctions were used to induce stronger discipline and enforce the acceptance of the Protestant work ethic. Real sanctions were also used strongly in practice, especially on people on social assistance. As a consequence, the very nature of the frontline work changed considerably within a decade (Baadsgaard et al. 2014). Fifth and finally, in June 2010, the government decided—​in the midst of the crisis—​to shorten the unemployment benefit period from four to two years and to double the period for regaining the right for support. This was pushed through, once more, in the name of increasing labor supply and improving Danish competitiveness. This decision provoked strong political and professional debates

26

226  Work and the Social Safety Net because it threatened the “security” part of Danish “flexicurity.” The reforms were also meant to bring savings and improvement to public budgets—​and they did.

Towards More Effectiveness and Efficiency Through Performance Management, Marketization, and Decentralization? Alongside the attempts to “improve” the supply of labor, a strong focus on quantitative output measures and performance metrics was introduced while omitting other important aspects such as accessibility, adequacy, quality, etc. This approach intensified with the process of municipalization described below. Beginning in 2003, the government took several steps toward contracting-​out the implementation of employment services. In theory, outsourcing promised more efficient and cost-​effective solutions along with a reduction in bureaucracy. This outsourcing of activation programs for both insured and uninsured people to private firms created a new market—​and new control functions within the employment system. The results have, however, not been according to the expectations; more creaming and parking of unemployed people and growing transaction costs have been recorded (Bredgaard and Larsen 2006, 2008; Larsen 2013). A re-​regulation of processes followed intense political discussion and critique from the media; the level of contracting-​out was subsequently reduced again. In 2011, the Minister of Employment reduced the number and character of privately organized courses and activation arrangement, giving priority to wage subsidies, company trainee places (virksomhedspraktik), and salary-​free internships in private firms. Activation became a shrinking market. Embedded in a larger regional governance reform, the fourteen PES regions were consolidated into four larger employment regions in 2007, while the number of municipalities was drastically reduced from 270 to 98 (Weishaupt 2011b). Part of the reform was the creation and integration of Jobcenters for both insured and non-​insured unemployed. Like the introduction of profiling, this was an idea copied from the Netherlands (Marsh and Sharman 2009). Initially, the PES and the municipalities were to steer activities together in seventy-​seven “joint” Jobcenters, while fourteen were handed over to the municipalities alone. The employer organizations and the trade unions jointly protested against this change, albeit without success. As a result, the labor market parties were no longer in pivotal positions in the steering bodies. Ninety-​one new local labor market “boards” were formed with private-​interest representation, but they lost former decision-​making competences and were left without economic muscle. The establishment of ninety-​one Jobcenters and all responsibility for activation placed by the local decision-​makers also meant abolishing regional labor market policy and regional corporatism (Jørgensen 2009). The number of Jobcenters was subsequently increased to ninety-​four.

22 7

Activation in PES in Europe  227 The next step of this strategy was taken in August 2009, when the seventy-​ seven “joint” Jobcenters were dismantled and responsibility for employment activities transferred to the municipalities. The state-​run PES was dissolved and full municipalization completed; local politicians and Jobcenter leaders became the new decision-​makers. This new design affected policy content, adjustment processes, and the daily functioning of the system. The employees of the former AF were subsequently employed by the municipality. The merger created, however, severe tensions between the state and the municipalities, while the center remains in a strong controlling position due to monitoring and new benchmarking system. Strong efforts were made to build an evidence-​based information, evaluation, and knowledge system, but with limited effectiveness. While responsibility for all actions was decentralized, control was centralized and contractualism introduced at all levels—​as in most continental employment systems (Sol and Westerveld 2005; van Berkel and Borghi 2008b). However, the autonomy of the municipalities was so strong that, in reality, different local employment strategies were created and not always in accordance with steering efforts from above. This was addressed through the establishment of an analytical “knowledge bank” and a strategy of targeted randomized controlled trials and regular systematic reviews of research evidence to rank and grade policy and practice knowledge and make it available to PES practitioners in the decentralized system. The municipalities have different strategies implemented—​some investment-​oriented, others looking more for cuts in the budget—​and the efforts of the state to control the municipalities directly from above created a lot of tensions, finally resulting in more freedom at the local level to create individual delivery strategies (Andersen and Larsen 2018). However, the state only accepted this because of a new economic steering arrangement, according to which the municipalities will have to pay over time for those not brought into employment. The knowledge bank has almost no relevance now. The economic steering system is the most important instrument used (Klindt et al. 2020).

The Global Financial Crisis and a Return to Social Democratic Rule In October 2011, a new Social Democratic-​led government—​consisting of three political parties, including a social-​liberal one—​came into power. They immediately canceled some of the most contested arrangements, such as “start help” and low social assistance based on less than 450 hours of work. A new reform of social assistance introducing work or education claims for youngsters and a reduced level of assistance for those not going into work or education was similarly decided on in 2013. Other initiatives were taken in 2012–​2015 to address unemployment, including stronger use of job rotation and training courses. The

28

228  Work and the Social Safety Net economic conditions in the wake of the global financial crisis have, however, made cuts in public budgets a follower of most reforms. More elements from “more people to work” slowly changed: the three match categories were abolished again; coordination between Jobcenters, social partners, and educational institutions was encouraged; and the Jobcenters were advised to get in contact with firms and reinstall job exchange activities, thus redirecting attention and activities to the demand side. New targeted educational measures were introduced in the first part of 2013, for people who were in danger of losing their unemployment benefit because they approached the end of the benefit period (benefits did not expire but the benefit level was reduced). From July 2012, a new and reduced unemployment benefit system was to be in effect. However, the change of government postponed the launch the scheme because, during 2013, around 35,000 people were in danger of losing their economic support, and among these many were losing the possibility of social assistance as well. In general, the rise in unemployment since 2008 had not been conquered effectively, job creation remained weak, and the LMP measures were, generally, not considered very successful. The January 2014 reform of the social assistance system was expected to bring young people—​put on low benefits—​ into education or into jobs. By 2014, the unemployment figure was still around 6%, while the Social Democrat-​led government reacted only slowly with the introduction of “acute” packages without visible effects. Strong conflicts over the employment policy and the political strategy dealing with the crisis subsequently emerged. The new governance structures with municipalities at the center seemed to be less capable than before in coping with the unemployment problems and other problems related to labor market imbalances. Many local actors had their own interpretation of problems and possible solutions, which weakened the horizontal coordination power of the system. The efforts of the local Jobcenters were contested. This seems partly due to the de-​corporatization and municipalization of Danish LMP. Steering arrangements were relatively new, and the Jobcenters behaved according to their own financial interests, strongly driven by the possibility of receiving refunds from the state. This again provoked strong critique of the municipalities due to the perception of the employees and the work done within the Jobcenters. In June 2014, the government finally succeeded in creating political support for a new employment policy for insured people. The reform reduced work-​first elements and reintroduced education and job exchange as central instruments, while the implementation structure has been supplemented with regional labor market boards—​again with labor market parties represented. In June 2015, a shift of government, with a liberal party taking office in a small minority arrangement, paved the way for a ceiling on social assistance and a reintroduction of a requirement of 225 hours of work within one year in order to have access to support. Furthermore, a reform of the unemployment benefit

22 9

Activation in PES in Europe  229 system was decided on following a commission report. The reform introduced incentives for taking up short-​time jobs, thereby supporting more flexible rules for regaining the right for unemployment benefits. The demands on street-​level bureaucrats to find new jobs and training opportunities grew once again—​but with different local strategies as a frame. The street and the state still meet with a classification of unemployed people as ready for job or ready for activation as the central criteria. In June 2019, finally, a new shift in government, with the Social Democratic party taking office again, has paved the way for more and newer elements of education and training activities. A human capital approach seems to be a preferred government priority, but the political situation at the parliamentary level supported only those changes made that were supported by the parties involved in the 2014 reform. Meanwhile, the COVID-​19 crisis called for comprehensive economic support from the state to firms and wage earners, and seventeen corporatist agreements with social partners were made in 2020 and the first part of 2021. Activation was suspended for many months and its future is unknown.

Conclusion and Discussion Years of municipalization, marketization, and other organizational reforms have taken their toll on the Danish system: fragmentation, disaggregation, narrow specialization, and less coordination. In the name of increasing the labor supply, Danish international status as a role model in ALMPs was slowly undermined. And this was done deliberately. Activation measures have been reoriented toward wage subsidies and company trainee placements. Importantly, most of the “work-​first” elements were introduced before the economic crisis in late 2008. They do not stem from the financial pressure of the crisis (Jørgensen 2013). Almost all operations within the employment system should be with the supply-​ side, the target groups. In 2015, the policy was redirected again. Politically, this was called a neoliberal reaction to “social democratic steering.” Work-​first elements, a new actor system, and processes did transform the policy, but operational reforms contributed to this as well. The shift of government in June 2019 again changed the preferred priorities, giving more support for education and training. But the local delivery strategies of the municipalities are not controllable from the center. The NPM-​inspired reforms of the administrative system during the 2000s seem to have created more than marginal professional and political discontent at the frontline level. The control efforts of the state (monitoring and performance management systems, standardization, and de-​professionalization) seemed to create more problems than they solved. The municipalities have gained more freedom during the past five or six years (Jørgensen and Schulze 2020). But frontline workers, however, have reported more professional dilemmas and have called for re-​skilling initiatives (Jørgensen et al. 2015).

230

230  Work and the Social Safety Net The new steering arrangements only marginalized the role of the labor market parties but perhaps also brought an end to Danish consensus culture. Regional LMP was abolished and new territorial principles, instead of functional ones, were introduced; strong coordination and control problems within the system are prominent. The shift of responsibility to the municipalities was not supportive to the security elements in Danish flexicurity, but the political will to reinstall a state-​run PES did not exist (Klindt and Jørgensen 2018). Tensions between the state and municipalities arose. Local decision-​makers did not always share centrally made problem definitions and implementation instructions, a problem exacerbated by the financial squeeze on municipalities arising from structural and inbuilt tensions between the local and central levels. Local strategies and work instructions differed. This resulted in problems for implementation agents. Similarly problematic were the efforts of marketization, which grew quickly during the first six or seven years after the introduction of reforms in 2003, but was in retreat thereafter. By 2015, a number of reforms materialized following a new shift of government, showing that partisan politics and power distributions matter. This pattern repeated itself again in 2019. The aim of the 2015 reforms was to give LMP positive motivating elements, including educational offers, that no longer built on threat effects—​for all kinds of unemployed people. Their purpose was to end or reduce programs that were “truly cold, wet, hard, and have no skill-​enhancing components” and the dominant administrative functions of social workers (Rosholm and Svarer 2004, 35). The employment reform of 2014 weakened “work-​first” elements, but the 2015 reforms again redirected policy profile and the policy-​making arrangements. By 2021, the government—​influenced by the experiments and lessons learned during the COVID-​19 crisis with interventionist policies—​might want to redirect activation into a new active labor market policy, leaving work-​first elements behind. New reforms are called for and planned. The changes made during the past ten years make it difficult to assess and classify the Danish employment system during these years of recurring transformation and contradictory steering signals. Activation will present more editions and practices over time.

OVERALL CONCLUSION In this concluding section, we tackle two questions outlined in the introduction. First, was there a convergence of PES governance structures across diverse cases during the Great Recession and its aftermath? Second, what role have (the different types of) PES played in minimizing the employment effects of the Great Recession? Both questions are answered sequentially. First, the preceding discussion suggests “difference within convergence” is an appropriate description of PES reform in the three cases and possibly across

23 1

Activation in PES in Europe  231 Europe. While there were clearly patterns of convergence of principles, rhetoric, and macro-​practices across PES, within these there remain considerable degrees of difference and divergence in relation to their implementation, demonstrating variegated and path-​dependent tendencies (Nunn 2019). This can be illustrated across the following four thematic areas.

New Public Management and Performance Management PES throughout Europe increasingly took up NPM principles and particularly performance management (Nunn 2012; Weishaupt 2010b, Nunn and Morgan 2018). More than this, there was a general shift toward outcome-​oriented performance management to align with the principles of activation and employment transitions and to accompany this with benchmarking between subnational units (e.g., regions, localities, offices). However, within this, there was a great deal of variation across Europe and within our three cases (Nunn 2013). This extended across the degree of centralization or local autonomy for target-​setting, the role and involvement of stakeholders and social partners, the number and nature of monitoring, and the way in which benchmarking is conducted. To illustrate the point among our cases, Denmark clearly went much further to move away from effective centralized performance management by allowing local municipalities to set their own targets and indicators even if centrally agreed goals were to bring more standardized results. In Germany, a middle road was taken, with centrally set indicators but negotiation over targets in the unemployment insurance system and contracting for local delivery in the minimum income (Hartz IV) system. In the United Kingdom, despite promises of greater local autonomy, performance management was a top-​down affair, with centrally set targets and performance expectations.

Privatization of Placement and Activation Services Our three cases are among the small group of countries (along with Australia, the Netherlands, and the United States) that have experimented with the privatization of various employment services. However, here again, there were substantial differences in how they went about this. In Denmark, a labor market reform from 2003 allowed private firms to carry out activation measures, thereby reducing the role of the PES and introducing economic incentives to steer these activities. The effects, however, were rather modest and produced unintended consequences. As a result, the degree of privatization was rolled back. Similarly, the reforms significantly reduced the role of further vocational training, prioritizing employer-​based internships and wage subsidies instead. Accordingly, the roles of municipalities (providing counseling and matching services) and

23

232  Work and the Social Safety Net employers were strengthened, while other private actors played a less prominent role. In Germany, privatization was undertaken via a voucher system, handing “consumer power” to job seekers to choose their preferred service and provider. The impact remained ambiguous. On the one hand, the placement vouchers were rarely used and had positive effects only for particular groups; on the other hand, the training and education vouchers were a continuation of third-​p arty–​b ased service delivery by other means. The voucher system was, however, mostly welcomed as it increased capacity (in the case of placement) or replaced a much more tedious, tender-​ based competition process (in the case of training/​education). In the United Kingdom, an even more ambitious scheme saw most employment services contracted out via regional contracts and an elaborate payment-​ by-​results system designed to promote stability in the “market” while fostering innovation in the delivery of services. The evidence, though, suggested that this was not meeting its own performance expectations and resulted in undesirable internal practices and outcomes, including creaming and parking, and it led to widespread public debate over the legitimacy of ALMPs and the subsequent abandonment of some aspects of the privatization program. Experience in all three countries suggests that privatization of placement services did not deliver on its promises and may have been counter-​productive.

Decentralization Again, decentralization was a watchword for PES across Europe (Mosley 2011) and was a key feature, at least in principle, in all our cases. In Denmark, decentralization has arguably been taken the furthest, with reforms in 2009 handing full control of the delivery of employment services to the municipalities. However, central monitoring remained crucial, and a set of national rules limited local autonomy and staff flexibility. In Germany, a middle road was taken, with substantial regional and local freedoms in the unemployment insurance and Hartz IV systems, respectively, but set within a system of nationally defined policy instruments. However, developments in both performance management and budgetary control suggested an increase in local autonomy in the German context as well (Weishaupt 2014). In the United Kingdom, however, decentralization was the least impressive implementation among our cases. Even in there however, decentralization was a notable theme and, though implementation was weak in the PES itself, substantial freedoms were offered to private providers in the contracted-​out “black box” system. In the PES, too, though, frontline advisers were freed of some of the previously rigid centralized controls. Decentralization then was also an example of persistent differences and to some extent further divergence within an overlying framework of convergence among PES about the desirability of local autonomy.

23 3

Activation in PES in Europe  233

Tightened Conditionality Tightened conditionality was evident across our three cases. This reflected a broadly neoliberal orientation in ALMPs, with the emphasis being on encouraging labor supply through measures to make intentional unemployment less attractive. However, within this general context, and in response to the Great Recession, measures were taken in Germany to extend some interventions to bridge the supply and demand side, such as subsidies for short-​time working and encouraging training to both retain employment and encourage human capital development. By contrast, both Denmark and the United Kingdom sustained a broadly “work-​first” message. Though tough conditionality was somewhat repealed in Denmark with the election of a social democratic government in 2011, in the United Kingdom there was cross-​party consensus (and indeed in public opinion as well) in support of a comparatively disciplinary approach to welfare conditionality. Part of the explanation for this different orientation and the resilience of social democratic or more liberal ideas may be the differential involvement of social partners. In the United Kingdom, trade unions were virtually excluded from JCP governance, while in Germany all social partners were structurally embedded in the Bismarckian unemployment insurance system. Even though ALMPs were no longer “owned” by social partners after governance reforms in Denmark, they still played a pivotal role in policy-​making at the central level while continuing to be important “service providers” in the system. Second, assessing the resilience of PES to the Great Recession is a contingent question to which a counter-​factual answer is not possible; in other words, it is hard to say what the impact of the recession on European labor markets would have been without ALMPs, PES, and welfare payments (acting as “automatic stabilizers”). However, we have shown that the different types of PES governance have produced varied effects, each with its own advantages and shortfalls. The Danish experience suggests that a strategy based on municipalization produced inherent tensions, especially during times of rising unemployment, as local and national priorities collided. Likewise, the Danish strategy of de-​corporatization turned out to be a disadvantage as the social partners’ expertise was missing, and it was harder to generate broad, consensual policy responses. The British reliance on private providers and work-​first activation also seemed problematic. Where do private placement firms or JCP put people in times of mass unemployment and no demand? The continuation and expansion of the previous government’s “job experience” programs can be, at best, only a temporary solution and risks to retain a “lost generation” remain relevant. The German experience, relying on a centrally and publicly governed PES with continued and institutionalized inclusion of the social partners held certain advantages. On the one hand, the social partners were there to co-​decide on, communicate, and carry out policy decisions. On the other hand, the relative strong central steering power exercised by the state and PES headquarters and the lack of private actors made swift

234

234  Work and the Social Safety Net reactions easier to transpose. While we cannot put hard numbers to the effects of the PES, the evidence suggests that a PES—​in contrast to a communal or private one—​is crucial in tough times when markets fail and states have to react swiftly, coordinately, and comprehensively.

REFERENCES Adamy,Wilhelm.GibtesnocheineSelbstverwaltunginderArbeitslosenversicherung? Vierteljahresschrift für Sozialrecht 24 (2006):175–​189. Andersen, N., and F. Larsen. Beskæftigelse for alle? Copenhagen: Frydendal, 2018. Baadsgaard, Kelvin, Henning Jørgensen, and Iben Nørup. Labour Market Policy Reforms and Professional Practice. De-​professionalization and Remaking of Professionalism. Paper presented at ILERA European Regional Congress, Amsterdam, June 2013. Baadsgaard, Kelvin, Henning Jørgensen, Iben Norup, and Peter Olesen Søren. Jobcentre og Klemte Kvalifikationer. Aalborg, Denmark: Aalborg University Press, 2014. Bambra, Clare, and Katherine E. Smith. No Longer Deserving? Sickness Benefit Reform and the Politics of (Ill) Health. Critical Public Health 20, no. 1 (2010): 71–​83. Baumgarten, Daniel, Michael Kvasnicka, Juliane Landmann, and Eric Thode. Durchlässiger Arbeitsmarkt durch Zeitarbeit?—​Rheinisch-​Westfälisches Institut für Wirtschaftsforschung. Studie im Auftrag der Bertelsmann Stiftung. Gütersloh, Germany: Bertelsmann Stiftung, 2012. Bekker, Sonja. Flexicurity: The Emergence of a European Concept. Cambridge: Intersentia, 2012. Bernhard, Sarah, and Thomas Kruppe. Vermittlungsgutscheine für Arbeitslose: Oft ausgegeben und selten eingelöst. IAB Kurzbericht 21 (2010): 1–​8. Betzelt, Sigrid, and Silke Bothfeld, eds. Activation and Labour Market Reforms in Europe. Basingstoke, UK: Palgrave Macmillan, 2011. Bredgaard, Thomas, and Flemming Larsen. Udlicitering af Beskæftigelsespolitikken. Copenhagen: DJØF Publishing, 2006. Bredgaard, Thomas, and Flemming Larsen. Quasi-​ Markets in Employment Policy: Do They Deliver on Promises? Social Policy and Society 7, no. 3 (2008): 341–​352. Bruttel, Oliver. Are Employment Zones Successful? Evidence From the First Four Years. Local Economy 20, no. 4 (2005): 389–​403. Bruttel, Oliver, and Els Sol.Work First as a European Model? Evidence from Germany and the Netherlands. Policy and Politics 34, no. 1 (2006): 69–​89. Bundesagentur für Arbeit, ed. Das Arbeitnehmerorientierte Integrationskonzept der Bundesagentur für Arbeit (SGB II und SGB III). Version 7.0—​Intern. Nürnberg, Germany: Bundesagentur für Arbeit, 2012.

23 5

Activation in PES in Europe  235 Casebourne, Jo. Innovation in Labour Market Programmes. NESTA, 2012. http://​ www.nesta.org.uk/​blog/​inn​ovat​ion-​lab​our-​mar​ket-​pro​gram​mes. Castonguay, Julie, and Els Sol. Benchmarking as NPM Instrument in Activating Social Security Policies. European Journal of Social Security 12 (2010): 324–​343. Centre for Economic and Social Inclusion. Work Programme Performance Statistics: Inclusion Analysis. London: Centre for Economic and Social Inclusion, 2012. Citizens Advice Bureau. Not Getting Through: CAB Evidence on the New System for Claiming Benefits from Jobcentre Plus. London: Citizens Advice Bureau, 2007. Clasen, Jochen, and Daniel Clegg, eds. Regulating the Risk of Unemployment: National Adaptations to Post-​ Industrial Labour Markets in Europe. New York: Oxford University Press, 2011. Crimmann, Andreas, Kerstin Ziegler, Peter Ellguth, Susanne Kohaut, and Florian Lehmer. Forschungsbericht zum Thema Arbeitnehmerüberlassung. Nuremburg, Germany: Instut für Arbeitsmarkt und Berufsforschung (IAB), 2009. Davies, Steve. Contracting Out Employment Services to the Third and Private Sectors: A Critique. Critical Social Policy 28, no. 2 (2008): 136–​164. Department for Work and Pensions (DWP). Building Britain’s Recovery: Achieving Full Employment. London: The Stationery Office, 2009a. Department for Work and Pensions (DWP. Realising Potential: Developing Personalised Conditionality and Support: A Discussion Paper on Next Steps in Implementing the Gregg Review. London: Department for Work and Pensions, 2009b. Department for Work and Pensions (DWP). Impacts and Costs and Benefits of the Future Jobs Fund. London: Deparment for Work and Pensions, 2012a. Department for Work and Pensions (DWP). Changes to Jobseeker’s Allowance Sanctions from 22 October 2012, 2012b. http://​web​arch​ive.natio​nala​rchi​ves. gov.uk/​201​3062​7060​116/​http:/​www.dwp.gov.uk/​advi​ser/​upda​tes/​jsa-​sanct​ ion-​chan​ges/​. Department for Work and Pensions (DWP). Work Programme Statistical Summary June 2013. London: Deparment for Work and Pensions, 2013b. Department for Work and Pensions (DWP). Work Programme Official Statistics to June 2014. London: Department for Works and Pensions, 2014. Dingeldey, Irene. Between Workfare and Enablement: The Different Paths to Transformation of the Welfare State: A Comparative Analysis of Activating Labour Market Policies. European Journal of Political Research 46, no. 6 (2007): 823–​851. Eichhorst, Werner, and Anton Hemerijck. Welfare and Employment: A European Dilemma? In United in Diversity? Comparing Social Models in Europe and America, ed. Jens Alber and Neil Gilbert. Oxford: Oxford University Press, 2010, 201–​236. Eichhorst, Werner, and Paul Marx. Reforming German Labour Market Institutions: A Dual Path to Flexibility. Journal of European Social Policy 21, no. 1 (2011): 73–​87.

236

236  Work and the Social Safety Net Einsiedler, Eberhard. Perspektive Qualität. Discussion Paper. Nuremberg, Germany: Federal Employment Agency, 2012. European Commission. Growth, Competitiveness, and Employment: The Challenges and Ways Forward Into the 21st Century. Brussels: European Communities Commission, 1993. Evetts, Julia. New Professionalism and New Public Management: Changes, Continuities and Consequences. Comparative Sociology 8, no. 2 (2009): 247–​266. Finn, Dan. Outsourcing Employment Programmes: Contract Design and Differential Prices. European Journal of Social Security 12 (2010): 289–​303. Finn, Dan. Welfare to Work After the Recession: From the New Deals to the Work Programme. In Social Policy Review 23: Analysis and Debate in Social Policy, 2011, ed. Chris Holden, Majella Kilkey, and Gaby Ramia. Bristol, UK: Policy Press, 2011, 127–​146. Freud, David. Reducing Dependency, Increasing Opportunity: Options for the Future of Welfare to Work: An Independent Report to the Department for Work and Pensions. Leeds, UK: Department for Work and Pensions Corporate Document Services, 2007. Grover, Chris, and Linda Piggott. From Incapacity Benefit to Employment and Support Allowance: Social Sorting, Sickness and Impairment, and Social Security. Policy Studies 31, no. 2 (2010): 265–​282. Gulliford, Jenny, and Beth Foley. The New Help to Work Scheme Could Use Some Work. London School of Economics and Political Science British Politics and Policy Blog, May 5, 2014. http://​blogs.lse.ac.uk/​politi​csan​dpol​icy/​help-​to-​ work/​. Hales, Jon, R. Taylor, W. Mandy, and M. Miller. Report on a Cohort Survey of Long-​Term Unemployed People in the Zones and a Matched Set of Comparison Areas. London: Department for Work and Pensions, 2003. House of Commons Committee of Public Accounts. Support to Incapacity Benefits Claimants through Pathways to Work: First Report of Session 2010–​11. London: The Stationery Office, 2010. House of Commons Committee of Public Accounts. Department for Work and Pensions: Work Programme Outcome Statistics: Thirty-​Third Report of Session 2012–​13. London: The Stationary Office, 2013. House of Commons Work and Pensions Committee. Work and Pensions: Sixth Report of Session 2006–​07. London: The Stationery Office, 2007. House of Commons Work and Pensions Committee. DWP’s Commissioning Strategy and the Flexible New Deal: Second Report of Session 2008–​ 09. London: The Stationery Office, 2009. House of Commons Work and Pensions Committee. Can the Work Programme Work for all User Groups?: First Report of Session 2013–​14. London: The Stationery Office, 2013.

23 7

Activation in PES in Europe  237 Immervoll, Herwig. Minimum-​ Income Benefits in OECD Countries: Policy Design, Effectivenss and Challenges. OECD Social, Employment and Migration Working Papers No. 100. Paris, France: OECD Publishing, 2010. Jakob, Johannes. Die Gesetzliche Arbeitslosenversicherung. In Handbuch für Selbstverwalter -​Die Arbeit in der Renten-​, Kranken-​, Pflege-​, Unfall-​, und Arbeitslosenversicherung, ed. Ursula Engelen-​Kefer. Frankfurt: Bund-​Verlag, 2006, 21–​49. Jessop, Bob. The Future of the State in an Era of Globalization. In Challenges of Globalization: New Trends in International Politics and Society, ed. Alfred Pflaller and Marika Lerch. Brunswick, NJ: Transaction Publishers, 2005, 13–​26. Johnson, Steve, and Alex Nunn. Evaluation of the Job Outcome Target Pilots: Findings from the Qualitative Study. Leeds, UK: Corporate Document Services, 2005. Johnson, Steve, and Alex Nunn. Evaluation of the Job Outcome Target Pilots: Synthesis Report. Leeds, UK: Corporate Document Services, 2006. Johnson, Steve, and Alex Nunn. Working with JOT 18 Months on: Qualitative Research in Former Option 1 Pilot Districts. Leeds, UK: Corporate Document Services, 2007. Jørgensen, H., and M. Schulze. Aktiveringsparadigmet som dominerende princip for arbejdsmarkedspolitik i Tyskland, Danmark og UK i komparation. In Aktiv arbejdsmarkedspolitik—​Etablering, udvikling og fremtid, ed. M. P. Klindt et al. Copenhagen: DJØF-​Forlaget, 2020, 279–​297. Jørgensen, Henning. Consensus, Cooperation and Conflict: The Policy Making Process in Denmark. Cheltenham, UK: Edward Elgar Publishing, 2002. Jørgensen, Henning. From a Beautiful Swan to an Ugly Duckling. The Renewal of Danish Activation Policy Since 2003. European Journal of Social Security 11, no. 4 (2009): 337–​368. Jørgensen, Henning. Le Recent Recalibrage du Marché du Travail et du Système de Protection Sociale au Danemark. Politiques Sociales et Familiales 112 (2013): 15–​27. Jørgensen, Henning, Kelvin Baadsgaard, and Iben Nørup. De-​Professionalization Through Managerialization in Labour Market Policy: Lessons from the Danish Experience. In Restructuring Welfare Governance: Marketization, Managerialism and Welfare State Professionalism, ed. Tanja Klenk and Emmanuele Pavolini. Cheltenham, UK: Edward Elgar Publishing, 2015, 163–​182. Jørgensen, Henning, and Per Kongshøj Madsen, eds. Flexicurity and Beyond: Finding a New Agenda for the European Social Model. Copenhagen: DJØF Publishing, 2007. Klindt, M. P., and H. Jørgensen. Revisiting Danish Flexicurity After a Decade of Reform. Does the Labour Market Still Work for Everyone? In Hybrid Public

238

238  Work and the Social Safety Net Policy Innovations. Contemporary Policy Beyond Ideology, ed. M. Fabian and R. J. Breunig. New York: Routledge, 2018, 134–​151. Aktiv arbejdsmarkedspolitik: Etablering, udvikling og fremtid. Copenhagen: DJØF-​Forlaget, 2020. Krugman, Paul. Free to Lose. New York Times, November 12, 2009. http://​www. nyti​mes.com/​2009/​11/​13/​opin​ion/​13krug​man.html?_​r=​0. Knuth, Matthias, and Flemming Larsen. Increasing Roles for Municipalities in Delivering Public Employment Services: The Cases of Germany and Denmark. European Journal of Social Security 12 (2010): 174–​200. Koch, Susanne, Christiane Spies, Gesine Stephan, and Joachim Wolff. Kurz vor der Reform: Arbeitsmarktinstrumente auf dem Prüfstand. IAB Kurzbericht 11 (2011): 1–​8. Lane, Pippa, Rowan Foster, Laura Gardiner, Lorraine Lanceley, and Ann Purvis. Work Programme Evaluation: Procurement, Supply Chains and Implementation of the Commissioning Model. Department for Work and Pensions Research Report No. 832. London: Department for Work and Pensions, 2013. Larsen, Flemming. The Role of Quasi-​Markets in Governance and Employment Policy: The Case of Denmark. Paper presented at the Social Policy and Administration conference, Melbourne, February 2013. Leschke, Janine, and Andrew Watt. How Do Institutions Affect the Labour Market Adjustment to the Economic Crisis in Different EU Countries? ETUI Working Paper No. 2010.04. Brussels: European Trade Union Institute, 2010. Marsh, David, and Jason C. Sharman. Policy Diffusion and Policy Transfer. Policy Studies 30, no. 3 (2009): 269–​288. McGuinness, Feargal, and A. Dar. The Work Programme. House of Commons Library Standard Note No. SN/​ EP/​ 6430 London: House of Commons Library, 2014. McGuinness, Feargal, and D. Harari. Work Experience Schemes. House of Commons Library Standard Note No. SN06249. London: House of Commons Library, 2013. McHardy, Fiona, and Fife Gingerbread Community Researchers. Surviving Poverty: The Impact of Lone Parenthood: Research Into the Experiences of Lone Parents in Rural Fife. Glasgow: The Poverty Alliance, 2013. Mosley, Hugh G. Decentralisation of Public Employment Services. The European Commission Mutual Learning Programme for Public Employment Services Analytical Paper. Brussels: European Commission, July 2011. Mosley, Hugh, Holger Schütz, and Nicole Breyer. Operational Objectives and Performance Indicators in European Public Employment Services. Berlin: Wissenschaftszentrum Berlin, 2000. National Audit Office. The Roll-​Out of Jobcentre Plus Office Network. London: National Audit Office, 2008.

23 9

Activation in PES in Europe  239 National Audit Office. Department for Work and Pensions: The Introduction of the Work Programme. London: National Audit Office, 2012. Newton, Becci, Nigel Meager, Christine Bertram, Anne Corden, Anitha George, Mumtaz Lalani, Hilary Metcalf, Heather Rolfe, Roy Sainsbury, and Katharine Weston. Work Programme Evaluation: Findings from the First Phase of Qualitative Research on Programme Delivery. Norwich, UK: Department for Work and Pensions, 2012. Nunn, Alex. Job Outcome Target National Evaluation. Leeds, UK: Corporate Document Services, 2007. Nunn, Alex. Restructuring the English Working Class for Global Competitiveness. Papers in the Politics of Global Competitiveness No. 9. Leeds, UK: Leeds Metropolitan University Policy Research Institute, 2008. Nunn, Alex. Performance Management in Public Employment Services. The European Commission Mutual Learning Programme for Public Employment Services Analytical Paper. Brussels: European Commission, July 2012. Nunn, Alex. Review of Performance Management in Public Employment Services. The European Commission Mutual Learning Programme for Public Employment Services Peer Review Comparative Paper. Brussels: European Commission, 2013. Nunn, Alex. The Contested and Contingent Outcomes of Thatcherism in the UK. Capital and Class 38, no. 2 (2014): 303–​321. https://​doi.org/​10.1177/​ 03098​1681​4530​126 Nunn, Alex. Neoliberalization, Fast Policy Transfer and the Management of Labor Market Services. Review of International Political Economy 27, no. 4 0 (2019): 949–​969. https://​doi.org/​10.1080/​09692​290.2019.1625​424 Nunn, Alex, and P. Beeckmans. The Political Economy of Competitiveness and Continuous Adjustment in EU Meta-​Governance. International Journal of Public Administration 38, no. 12 (2015): 926–​939. https://​doi.org/​10.1080/​ 01900​692.2015.1028​645 Nunn, Alex, Tim Bickerstaffe, and Ben Mitchell. International Review of Performance Management Systems in Public Employment Services. Norwich, UK: Her Majesty’s Stationery Office, 2010. Nunn, Alex, and Dave Devins. Process Evaluation of the Jobcentre Plus Performance Management Framework. London: Department for Work and Pensions, 2012. Nunn, Alex, and Steve Johnson. Labouring and Learning Towards Competitiveness: The Future of Local Labour Markets After Harker, Leitch and Freud. Local Economy, 23, no. 2 (2008): 122–​137. Nunn, Alex, Steve Johnson, Sarah Kelsey, and David Usher. Job Outcome Target National Evaluation. Leeds, UK: Her Majesty’s Stationery Office, 2007. Nunn, Alex, and J. Morgan. The Political Economy of Public Employment Services: Measurement and Disempowered Empowerment? Policy Studies, 2018. https://​doi.org/​10.1080/​01442​872.2018.1540​777

240

240  Work and the Social Safety Net Organisation for Economic Cooperation and Development (OECD). OECD Jobs Study. Paris: OECD Publishing, 1994. Pierson, Paul. Coping with Permant Austerity: Welfare State Restructuring in Affluent Democracies. In The New Politics of the Welfare State, ed. Paul Pierson. Oxford: Oxford University Press, 2001, 410–​514. Piggott, Linda, and Chris Grover. Retrenching Incapacity Benefit: Employment Support Allowance and Paid Work. Social Policy and Society 8, no. 2 (2009): 159–​170. Policy Research Institute. Phase 2 Evaluation of Multiple Provider Employment Zones: Qualitative Study. Leeds, UK: Her Majesty’s Stationery Office, 2006. Public and Commercial Services Union. DWP Briefing: Freedoms and Flexibilities Update. Leeds, UK: Public and Commercial Services Union, 2012. Public and Commercial Services Union. Evidence to the Work and Pensions Committee on the Role of Jobcentre Plus in the Reformed Welfare System. London: Public and Commercial Services Union, 2013. Rosholm, Michael, and Michael Svarer. Estimating the Threat Effect of Active Labour Market Programmes. IZA Discussion Paper No. 1300. Bonn: Institute for the Study of Labor (IZA), 2004. Save the Children. Welfare Reforms to Hit Poorer Working Women, Pushing 250,000 Children Deeper into Poverty, March 13, 2012. http://​www.save​thec​ hild​ren.org.uk/​2012-​03/​welf​are-​refo​rms-​hit-​poo​rer-​work​ing-​women-​push​ ing-​250​000-​child​ren-​dee​per-​pove​rty. Schram, Sanford F., Joe Soss, Linda Houser, and Richard C. Fording. The Third Level of US Welfare Reform: Governmentality Under Neoliberal Paternalism. Citizenship Studies 14, no. 6 (2010): 739–​754. Sol, Els. Vouchers, NPM and the Provision of Public Employment Services. European Journal of Social Security 12 (2010): 343–​357. Sol, Els, and Mies Westerveld, eds. Contractualism in Employment Services: A New Form of Welfare State Governance. The Hague: Kluwer Law International, 2005. Spiegel Online. Umstrittene Praxis der Arbeitsagentur: Schnell vermittelt, schnell wieder arbeitslos Spiegel Online, January 12, 2013. http://​www.spie​gel. de/​wir​tsch​aft/​sozia​les/​arb​eits​agen​tur-​chef-​weise-​will-​verm​ittl​ung-​in-​zei​tarb​ eit-​begren​zen-​a-​877​192.html. Van Berkel, Rik, and Vando Borghi. Introduction: The Governance of Activation. Social Policy and Society 7, no. 3 (2008a): 331–​340. Van Berkel, Rik, and Vando Borghi. Review Article: The Governance of Activation. Social Policy and Society 7, no. 3 (2008b): 393–​402. Van Berkel, Rik, and Ben Valkenburg. Making It Personal: Individualising Activation Services in the EU. Bristol, UK: Policy Press, 2007. Weishaupt, J. Timo. Germany After a Decade of Social Democrats in Government: The End of the Continental Model? German Politics 19, no. 2 (2010a): 105–​122.

24 1

Activation in PES in Europe  241 Weishaupt, J. Timo. A Silent Revolution? Management Ideas and the Reinvention of European Public Employment Services. Socio-​Economic Review 8, no. 3 (2010b): 461–​486. Weishaupt, J. Timo. From the Manpower Revolution to the Activation Paradigm. Explaining Institutional Continuity and Change in an Integrating Europe. Amsterdam: Amsterdam University Press, 2011a. Weishaupt, J. Timo. Social Partners and the Governance of Public Employment Services: Trends and Experiences from Western Europe. Labour Administration and Inspection Programme (LAB/​ ADMIN) Working Paper No. 17. Geneva: Labour Administration and Inspection Programme, International Labour Office, 2011b. Weishaupt, J. Timo. Origin and Genesis of Activation Policies in ‘Old’ Europe: Toward a Balanced Approach? In Minimum Income Protection in Flux, ed. Ive Marx and Kenneth Nelson. Basingstoke, UK: Palgrave Macmillan, 2013, 190–​216. Weishaupt, J. Timo. Central Steering and Local Autonomy in Public Employment Services. The European Commission Mutual Learning Programme for Public Employment Services Analytical Paper (October 2014). Brussels: European Commission, 2014. Weishaupt, J. Timo. German Labour Market Resilience in Times of Crisis: Revealing Coordination Mechanisms in the Social Market Economy. German Politics, 2021. doi:10.1080/​09644008.2021.1887852. 1-​20). Weishaupt, J. Timo, and Katja Lack. The European Employment Strategy: Assessing the Status Quo. German Policy Studies 7, no. 1 (2011): 9–​44. Wilthagen, Ton, and Frank Tros. The Concept of Flexicurity: A New Approach to Regulating Employment and Labour Markets. Transfer: European Review of Labour and Research 10, no. 2 (2004): 166–​187. Wintour, Patrick. May 19, 2013. Job seekers and Benefits Data Release Postponed by DWP. The Guardian, May 19, 2013 http://​www.theg​uard​ian.com/​soci​ety/​ 2013/​may/​19/​benef​i ts-​unemp​loym​ent.

24

10 WORKFORCE DEVELOPMENT SERVICES IN THE UNITED STATES Carolyn J. Heinrich

INTRODUCTION The COVID-​19 pandemic sparked unprecedented disruptions to labor markets around the world and forced many employers to reduce spending on workers. As labor demand has revived with economic production coming out of the pandemic, the skills gaps and worker shortages that have been intensifying over decades are likely to constrain productivity and economic growth. Yet as a recent report (Li et al. 2020) from the Centre for Vocational Education Research (CVER) in London conveys, employers are not investing more in training to prepare or upskill their workers for the labor market needs ahead. While employers have long been less likely to offer training to their less educated, lower-​skilled employees (Bassanini et al. 2007), Li et al. point out that training investments in more highly educated and skilled workers have been declining as well. Employer efforts to address skills gaps and shortages through recruitment rather than providing quality training are likely to come up short, with broader negative consequences for global economic growth in the absence of greater public investments and policy advances in workforce development. In the United States, the American Rescue Plan has provided $350 billion in emergency funding for eligible state, local, and other government entities to respond to the emergency and “bring back jobs.” Since the first $2 trillion in coronavirus relief funds were approved in March 2020, governors and state legislatures have been allocating some of the federal relief dollars toward short-​term training Carolyn J. Heinrich, Workforce Development Services in the United States In: Work and the Social Safety Net. Edited by: Douglas J. Besharov and Douglas M. Call, Oxford University Press. © Oxford University Press 2023. DOI: 10.1093/​oso/​9780190241599.003.0010

24 3

Workforce Development Services in the US  243 programs in high-​demand fields such as healthcare and information technology, as well as paying some employers to provide on-​the-​job training (OJT; Quinton 2020). Labor and workforce development experts admonish, however, that shorter-​term training programs are unlikely to be the remedy for chronic skills shortages or for training needs associated with significant labor market shifts brought about by the pandemic. As policymakers and employers consider how to address these critical workforce issues, it is worthwhile to reflect on the evolution of labor activation policies and training programs in the United States and consider the lessons and guidance they offer for the challenges we are facing globally today. Although the US Department of Labor (USDOL) was established in 1913, an active role for the department in labor activation and training did not begin until the 1930s, when President Franklin D. Roosevelt appointed Frances Perkins to his Cabinet to develop plans to alleviate unemployment and spur recovery from the Great Depression. Prior to the Manpower Development and Training Act (MDTA) of 1962, which officially established the federal public employment and training system in the United States, programs including the Civilian Conservation Corps, the Works Progress Administration (WPA), the Public Works Administration, and the National Youth Administration were viewed as temporary solutions to workforce challenges, with unemployment the primary concern. The Comprehensive Employment and Training Act (CETA) of 1973, which succeeded MDTA, extended the WPA approach in that it sought to provide work for the long-​term unemployed and those with low incomes, as well as summer jobs for low-​income youth. CETA also aimed to cede more control to state governments in administering employment and training programs, a trend that was advanced under the Job Training Partnership Act (JTPA) of 1982, reflecting the Reagan era of “New Federalism.” Compared to its predecessors, and consistent with the Reagan administration agenda to reduce the role of government, JTPA was distinguished by a more decentralized administrative structure that enlarged the role of the private sector in arranging for and delivering publicly funded employment and training services. JTPA also substantially diminished the public-​sector’s part in directly creating employment opportunities by eliminating the public service employment (PSE) and participant stipend components of CETA. While some states have recently used coronavirus relief funds to support OJT and textbooks and transportation for some trainees, PSE is not returning as a conventional path to supporting workers’ labor market entry or re-​entry. In addition, JTPA introduced a performance standards system to measure program outcomes across states and local service delivery areas, with the objectives of increasing local-​level accountability and encouraging more efficient program management. JTPA was also the first federal program in the United States to adopt an outcomes-​based performance management system that set national standards for program performance and attached incentives and consequences to the results reported by states. These reforms to US labor administration

24

244  Work and the Social Safety Net reflected the overarching goal of the Reagan administration to lessen the federal government’s fiscal responsibility and managerial role in addressing social problems. The Workforce Investment Act of 1998 (WIA) replaced JTPA beginning in July 2000, but it continued the governing philosophy that centralized authority should be limited so that state and local agencies can adapt employment and training programs to their own political and economic contexts. This has also contributed to considerable variation across states and local areas in how workforce development programs are organized and how and what services are delivered. Under JTPA, the non-​overlapping program jurisdictions were known as Service Delivery Areas, but some job-​training agencies were organized as public entities at the state, county, or municipal government level, while others were formed as private, not-​for-​profit, or for-​profit organizations. In the WIA program, states were required to establish a State Workforce Investment Board, composed of the Governor, members of the state legislature, and representatives of business, labor, educational entities, economic development agencies, and community-​based organizations. The local jurisdictions (Workforce Investment Areas) were directed and supervised by a board of representatives from business, labor, the community, and local elected officials. The boards played a central role in determining who was served, the types of services made available, and who should provide the services (within the limitations of the statute). One-​stop career centers, now more widely known as American Job Centers or AJCs, are one of the key mechanisms for local-​level planning and coordination that every local workforce investment area is required to operate. The AJCs are intended to coordinate and co-​locate more than a dozen federally funded education, workforce, and worker support programs to offer a basic menu of services that can help to meet the needs of a diverse set of individuals seeking assistance with training and/​or employment. The primary services provided include vocational training, OJT, basic or remedial education, job search assistance, work experience, and other services such as counseling and assessment, job-​readiness activities, and case management. The WIA (Gold Standard) experimental evaluation found considerable variation across the AJCs in how job seekers access these services, but also increasing collaboration among the co-​located partners to support client participation (Social Policy Research Associates 2016). The AJCs continue to operate under the Workforce Innovation and Opportunity Act (WIOA), which replaced WIA in 2014 and became effective in July 2015. Under WIOA, expectations for coordination have been further elevated, as partnerships and co-​location with other programs such as the Wagner-​Peyser Employment Services, Temporary Assistance for Needy Families (TANF), and US Department of Education programs became mandatory rather than encouraged. With new legislative and program priorities under WIOA and new pandemic-​ driven urgency to address long-​standing workforce development challenges, it is an appropriate time for a critical look not only at the public sector’s role in

24 5

Workforce Development Services in the US  245 workforce development and labor activation, but also at what the research base suggests about how effective US programs have been in supporting the overarching workforce development goals of helping job seekers access the employment, education, training, and support services they need to succeed in the labor market and for employers to compete effectively in the global economy.

ROLE OF THE PUBLIC SECTOR IN SUPPORTING HUMAN CAPITAL DEVELOPMENT Since their origins (as described above), employment and training efforts in the United States have relied on some combination of public-​and private-​sector resources, although private-​sector employers account for the lion’s share of workforce development activity and continue to dwarf public-​sector investments. US government spending on workforce development has averaged less than 0.5% of gross domestic product (GDP) in recent decades, shares that continue to be well below most Western European countries, as shown in Figure 10.1. Organisation for Economic Cooperation and Development (OECD) countries such as Denmark, Belgium, the Netherlands, and Finland have devoted up to ten times greater shares of GDP to labor market policy expenditures (Martin 2014). In terms of the incidence of employer-​sponsored training, the United States is

4.0 3.5 3.0 Denmark

2.5 2.0 1.5 1.0 0.5

United Kingdom United States

0.0 2000

2002

2004

2006

2008

2010

2012

2014

2016

2018

Figure 10.1  Public spending on labor market programs, 2000–​2018, for OECD countries. The graph depicts public spending measured as a percentage of GDP based on data on public spending on labor market programs that include public employment services (PES), training, hiring subsidies and direct job creations in the public sector, and unemployment benefits. The data are derived from country budgets and accounts and annual reports of bodies which implement the programs. Source: https://​data.oecd.org/​social​exp/​pub​lic-​spend​ing-​on-​lab​our-​mark​ets.htm for further details.

246

246  Work and the Social Safety Net in about the middle of the distribution (relative to other countries), although US employers have done less well in particular categories, such as occupational training for younger workers (Lerman 2016). These patterns raise questions about the adequacy of US workforce investment, as well as the extent to which public workforce investments should complement or undergird employer-​led training, or whether these investments should be targeted toward individuals or the types of workforce investments where private-​sector efforts are lacking. Economic theory on returns to training suggests that workers who acquire more training, if it in turn increases their individual productivity, should realize returns in the form of higher wages (Mincer 1974). Employers that provide training specific to their firm’s needs are likely to increase a worker’s wage to reduce turnover and are less likely to provide training when competition for employees is higher among firms (Rzepka and Tamm 2016), with the implication that returns to (firm-​specific) employer-​provided training are more likely to be privately realized. This suggests an unpersuasive case for public subsidization of this type of training. However, by the same line of reasoning, employers may underinvest in more general or portable types of training that would be more likely to generate external benefits, not only for other employers, but also for economic growth and efficiency (that improve societal well-​being), if training increases worker productivity. Using firm-​level data from Ireland that distinguished between general and specific training, Barrett and O’Connell (2001) found that general training has a statistically significant, positive effect on productivity growth that persists when controlling for a range of factors (e.g., firm size, initial level of human capital, corporate restructuring), but they did not find comparable effects for specific training. The American Society for Training and Development estimated that about three-​quarters of employer spending on training is for formal internal workplace learning (Rivera and Paradise 2006), and Lerman et al.’s (2004) analysis likewise found that employer training efforts disproportionately favor better-​educated and skilled workers. In addition, Bassanini et al. (2007) similarly found that in Europe (as in the United States), the provision of training by private firms increases with educational attainment and the skill intensity of occupations. In sum, privately funded training is more often likely to be narrowly targeted both in terms of who gets training (the higher skilled in more competitive markets) and in the type of training offered (i.e., firm-​specific, internally oriented). Both theoretical and empirical analyses (Gersbach and Schmutzler 2006; Holzer et al. 2011; Holzer 2013) suggest that as labor markets become more globally competitive and integrated, an even smaller segment of the workforce will have sufficiently high skills and productivity levels to induce additional investments by their employers. This, in turn, suggests a worsening inequality between higher-​and lesser-​skilled workers in access to private-​sector training opportunities and wage increases. Gersbach and Schmutzler attribute at least part of the decline in apprenticeships in Germany and some of the widespread

24 7

Workforce Development Services in the US  247 decline in the provision of general training to product market integration (associated with globalization) that reduces training investments made by firms. Citing his own work with colleagues (2011) and that of Acemoglu and Autor (2012), which points to “a growing complementarity over time between personal skills and firm wage premia, and strong labor market demand relative to supply for workers with these skills,” Holzer (2013, 6) questions whether the United States would be competing more effectively in the global labor market for “good jobs” if its public policies were more effectual in increasing human capital. There appears to be a growing consensus in labor market analyses that we are undersupplying workers with the required skills and credentials to satisfy labor demand for well-​ paying middle-​and high-​skill jobs, despite the apparently attractive labor market incentives for young and working-​age individuals to make these investments (Fouarge et al. 2013; Goldin and Katz 2008; Autor and Handel 2009). This raises another question about the role of the public sector in workforce development: If young, working-​age people are not responding to labor market incentives to pursue postsecondary education and training opportunities that would prepare them for well-​paying jobs that are in high demand, is there a role for the public sector to address this disconnect or the market failings that contribute to it (e.g., imperfect or asymmetric information, labor markets that are not perfectly competitive, externalities)? In the United States and in Europe, some suggest that we need to increase and improve opportunities for career and technical education before young people leave high school (Biavaschi et al. 2013; Rumberger 2011). Debate in the United States is ongoing about whether an overemphasis on college preparation in high schools has steered students away from technical course-​taking (or squeezed them out of high school course offerings), resulting in an inadequate pipeline of students trained for or on a trajectory to work in well-​paying, middle-​skill jobs. A growing body of research points to the importance of offering young people education and training opportunities that they see as relevant to their future job prospects and that provide this career context for learning, particularly for low-​income or disadvantaged youth who might otherwise drop out of high school (Center for Education Policy 2012; Holzer 2013; Lerman 2013). The latest findings from a multisite, experimental evaluation of the Year Up program—​which provides young adults (eighteen to twenty-​four years) with intensive technical training over six months, professional training (e.g., workplace culture, communications), and foundational supports (e.g., literacy, reasoning) over the course of a year—​revealed some of the largest impacts yet to be seen from a training program for young adults, with average annual earnings increases of nearly $8,000 (a 34% rise) compared to the control group (Fein et al. 2021). A distinctive feature of Year Up is a six-​month internship in information technology (IT) or financial services, often with Fortune 500 firms, with the employers paying about three-​fifths of the program costs. It is also notable that the Year Up program carefully screens applicants, and nearly all (>99%) have

248

248  Work and the Social Safety Net high school degrees when starting the program. New program models are also being tested for disconnected youth (i.e., those not working or in education or training), such as Project Rise (in New York, New Jersey, and Kansas City) that are aimed at increasing their educational attainment and employment opportunities through services that combine classroom education, internships, case management, and community projects. The Project Rise implementation evaluation (Manno et al. 2015) found that within twelve months of time in the program, 45% earned a credential or diploma, and approximately about 25% began in an unsubsidized job, although information on longer-​term outcomes or impacts is not available. Although the evidence base of proven youth programs is still relatively thin, we are not lacking for promising interventions (based on initial outcomes or impacts) that engage youth in career and vocational education that is targeted toward economically growing sectors (Heinrich and Holzer 2011). Funding for these programs, however, is not keeping pace with the level of program need among youth (Fein et al. 2021; Field 2011). Both public-​and private-​ sector investments in training will likely be constrained by tight budgets for some time to come, making it increasingly important that spending is well-​targeted in terms of how and for whom it can be most effective, as well as in consideration of where skills shortages lie. The existing evidence base on the effectiveness of workforce development programs, however, is limited in many ways. With the possible exception of the WIA Gold Standard impact evaluation and the experimental evaluation of Year Up, which has been gradually scaling up to become a national intervention, even the most comprehensive evaluations have been restricted in terms of the coverage and representativeness of the programs they have evaluated and the outcomes they have examined. Still, there are some consistent findings across rigorous research efforts that offer some basic guidance for workforce development policy, as well as research that illuminates where findings are mixed or suggest promising interventions that would benefit from further study (and/​or where better data are needed for evaluation).

THE EVIDENCE BASE ON TRAINING PROGRAM EFFECTIVENESS AND ITS LIMITATIONS The literature on employment and training program impacts is vast and spans more than four decades of research and evaluations. Fortunately, scholars have undertaken efforts to synthesize this literature, including meta-​analyses of active labor market policy (ALMP) evaluations (Card et al. 2018; Haelermans and Borghans 2012), training programs worldwide (Fares and Puerto 2009), and US government-​sponsored training programs and welfare-​to-​work programs (Greenberg et al. 2003, 2005), as well as other summaries of the empirical

24 9

Workforce Development Services in the US  249 evidence (Decker 2011; Fares and Puerto 2009; Brunello et al. 2007; Greenberg et al. 2006; Heckman et al. 1999). The meta-​analysis by David Card, Jochen Kluve, and Andrea Weber (2018) summarizes more than 200 studies of ALMPs (and more than 800 separate program impact estimates) from thirty-​eight countries between 1995 and 2014 and considers short-​term, medium-​term and long-​ term impact estimates, as well as the effectiveness of different program types. More than half of the studies came from just four countries (Denmark, France, Germany, and the United States), and although most were nonexperimental in design, they found (along with Greenberg et al. 2006), that experimental and nonexperimental evaluations of ALMPs and programs yielded similar results and conclusions about their effectiveness. Of course, that does not imply that these studies are without limitations regarding what generalizations we might make from them. Card et al. generally found null short-​run program impacts, but, particularly for programs that increased human capital accumulation, program impacts typically turned positive by about two to three years after completion. Table 10.1 presents a summary of the evidence base that focuses on studies from the past two decades, including the recent WIA Gold Standard impact evaluation and comprehensive reviews (e.g., syntheses and meta-​analyses) of the workforce development programs ALMPs. This summary is not intended to be all-​inclusive of the large and continually expanding body of research and individual studies on these programs, but rather to focus on some of the most compelling evidence and sources of cumulative knowledge and findings to date. The table provides basic information on the studies included, the types of programs and policies they examined, and findings on program and policy outcomes. Other findings and limitations of the studies are also indicated in the summary table. What may stand out most in the summary table are the limitations in the evidence base for drawing concrete conclusions about the effectiveness of workforce development programs/​ALMPs, including in terms of the measurement of outcomes, program costs and coverage, and longer-​term impacts. If numeric estimates of program impacts are reported, they are almost exclusively focused on average employment and/​or earnings or wages. Only a few studies monetize other impacts, such as government savings or reductions in welfare and crime, and there is little discussion or measurement of skills, credentials, or qualifications gained through training. Of 345 studies of training programs in 90 countries reviewed in Fares and Puerto’s (2009) meta-​analysis, only 16 attempted some accounting of costs and benefits, and obtaining accurate data on even direct program costs is a frequently acknowledged limitation in this body of research. Card et al. (2018) likewise reported that the data (from more than 200 studies) were inadequate for undertaking cost-​effectiveness or cost-​benefit analyses. The studies also vary in the length of time that they are able to follow program participants after receipt of services, and those studies that have followed outcomes over a longer period provide ample evidence that program impacts may change (grow or decay) over time. At the same time, one can make some

Use large cross-​country datasets available Scandinavian countries, France and New Zealand identified Documenting cross-​country variation for OECD countries to examine as the most training intensive countries (participation rates in training is difficult due to education and training in Europe, above 45%, more than thirty hours per employee). US idiosyncratic definitions of training theoretically and empirically: (i) OECD participation rates estimated at 41.4% and 17.9 in different surveys and country aggregate training data; (ii) Continuing hours per employee. 80% of vocational training courses paid data. Vocational Training Survey (CVTS); (iii) for or provided by employers, yet there are few studies on International the impact of training on productivity (due to lack of data Adult Literacy Survey (IALS); and (iv) on productivity). Rates of return estimates European Community Household Panel are even scarcer because data on cost are even more difficult (ECHP). to find than data on output. It is difficult to make a strong case for underprovision of workplace training; more research and information needed on externalities and costs, and more methodological checks on existing estimates.

Andersson et al. 2013

Bassanini et al. 2007

Outcomes by different types of programs and/​or country Other findings and limitations

Sample/​methods

Earnings differentials tend to be negative during first several Authors suggest that their findings Data on Workforce Investment Act quarters after WIA registration for training recipients. imply that job training efforts (WIA) participants (WIASRD data) are Earnings impacts become positive around the 6th should consider the jobs and linked to data on workers, employers firms for which workers are being and employment outcomes from the quarter and grow larger over the next several quarters, peaking at approximately $400–​500 per quarter; trained (e.g., akin to sectoral Longitudinal Employer Household estimated annual impacts for adults are $1,250–​1,700. approaches) if we are to increase Dynamics (LEHD) program for two Results are less favorable for dislocated workers (peak lower the effectiveness of training. states. Workers who received training are in one state and do not turn positive over twelve quarters in matched to workers who only received other (core or intensive) services at One-​ the other state). Stop Centers and inverse propensity Training appears to increase the probability of switching score weighting is used to estimate industries over time and is associated with some measures impacts. Objective was to measure of firm quality (i.e., may help workers gain employment in a wider range of impacts on worker higher-​paying firms and industries). outcomes with richer controls. Estimated impacts do not differ by gender.

Study

Table 10.1  Summary of findings from key studies on workforce development program/​active labor market policy effectiveness

250

Meta-​analysis of 207 studies (857 Subsidized public sector programs have least favorable Active labor market programs estimates) from 38 countries, 1995 outcomes over any time horizon; job search assistance has tend to have larger impacts in and 2014. Classified impact estimates positive shorter-​term impacts; classroom training more periods of slow growth and higher as significantly positive, significantly positive over medium-​and longer-​term impacts (short-​ unemployment. The estimated negative, or insignificantly different term impact estimates measure effects approximately one precision of program impact from zero. Multivariate regression year after program completion; medium-​term measured estimates varied widely across models with controls for program type approximately two years after completion; longer-​term studies. Very limited information and sample and study characteristics measured as three-​year impacts or longer). Country available on cost-​effectiveness in used to estimate effects on employment, differences are small after controlling for program type. the studies. Cost-​benefit analysis wages, unemployment duration, future Female participants and the long-​term unemployed tend to or calculation of social returns not unemployment. have larger program effects than other groups. feasible.

Card, Kluve, and Weber 2018

(continued )

Dual apprenticeship program Main outcome: probability of being in regular employment; German active labor market policies for accounts for half of all vocational also look at participation in higher education. Except for youth; administrative data for youth training entries each year (in job creation and preparatory training, programs improve (age twenty-​five or younger) entering secondary schooling). Preparatory unemployment in 2002 (n =​51,019) and probability of regular employment—​initial lock-​in phase, system for low education followed until 2008. Quasi-​experimental with impacts stabilizing at around two years after entry and methods applying inverse probability 5 to 20 percentage point increase in monthly employment attainment youths; low-​education weighting to seven programs: job search from third year on (varying by program and region). youth most vulnerable—​need and assessment, short-​term training more time to turn subsidized work Wage subsidies to regular employment most effective (max =​eight weeks), wage subsidies for (20 percentage point impact); long-​term training impacts experience into employment. By regular employment, job creation, long-​ around 10 percentage points (severe lock-​in effects); job sample design, majority in job term training (max =​approx one year), creation consistently negative effects. search or short-​term training. preparatory training (max =​one year). Probability to participate in unsubsidized education: positive increase in education probabilities of about 10 percentage points through longer-​term training, and professional qualifications increase by 20 percent. Preparatory programs do prepare youth for entering apprenticeships; no effects for employment programs.

Caliendo, Künn, and Schmidl 2011

Report increasing convergence toward comprehensive active labor market programs. Better evidence was not generated until early 1990s (63% of studies in sample 1990 or later). Little discussion of outcomes.

Fares and Puerto 2009

41% of 345 interventions found to have positive effects; 18% negative or no effects; 34% insufficient evidence. Only sixteen studies include cost-​benefit analyses. Interaction of in-​classroom +​workplace training increases positive impacts. Youth programs in LAC effective in increasing employment (by 5–​21%) and earnings (by 10–​35%), although overall, impacts of programs targeting youth have significantly lower impacts (30% lower) than those for adults. Training programs more effective in low-​and low-​middle income countries.

Meta-​analysis framework to review findings from 345 studies of training programs in ninety countries (controls for country characteristics), distinguishing in-​classroom training (37% of studies), workplace training (15%), classroom +​workplace (19%), classroom +​workplace +​supplemental services (29%); 61% were publicly financed training programs.

Decker 2011

Outcomes by different types of programs and/​or country Other findings and limitations

Sample/​methods

Review of studies on the implementation JTPA: 15% earnings increase for women, 8% increase for men, JTPA evaluation was experimental and impacts of Workforce Investment and net benefits per enrollee of $763/​quarter for women and but WIA evaluations were Act (WIA) programs, as well as pre-​and $781/​quarter for men. OJT/​JSA impacts higher for women nonexperimental. Potential for post-​1995 evidence (MDTA, CETA & and larger long-​run earnings effects (over $5,000 on average selection bias remains a concern JTPA). for women and men). with program impact estimates. WIA: larger estimated effects than JTPA on earnings. Heinrich Study samples are not nationally et al. (2008) estimates of $320–​692 per quarter for four years representative. after program entry and higher employment (5–​13% per quarter). Hollenbeck et al. (2005) earnings impacts higher starting at program exit ($773–​887 per quarter over eight quarters) and employment effects of 10.6% for women and 6.2% for men); impacts of training increase over time. JSA effects more immediate but short-​lived. Trade adjustment assistance and dislocated worker programs: a number of studies find small and/​or statistically insignificant effects. Differing estimation approaches suggest forgone earnings costs are high during program participation.

Study

Table 10.1 Continued

Meta-​analysis based on seventy-​one Main finding: average wage effect of on-​the-​job training is Too few studies measure the duration estimates of returns to on-​the-​job 2.6 percent, which is larger than the average return to of training, so the authors training from thirty-​eight studies education (reported by Ashenfelter et al., 1999). Using measured training as a dummy published between 1981 and 2010. Only estimation techniques that correct for selectivity bias, variable. Methodology and studies that computed the effect of on-​ the age until which an average training course is profitable data quality play a major role in the-​job training on wages were included. is fifty-​five years. determining the return to on-​the-​ Substantial heterogeneity in wage effects of training courses is job training. also found. Comparing the average number of hours spent on on-​the-​ job training with the average number of hours spent on schooling gives a wage increase of 30% for on-​the-​job training, compared with 8% for the return to schooling.

Haelermans and Borghans 2012

(continued)

Data from thirty-​one random assignment Mandatory programs: job search more effective; impacts Sample is from welfare-​to-​work evaluations of welfare-​to-​work positive for five to seven years but declining in magnitude programs and includes over 90% programs (twenty-​seven mandatory, after two to three years. More effective for less advantaged single parent families; study four voluntary). Measures of impacts on (without recent employment and longer-​term welfare also examined child outcomes earnings, percentin employment, welfare receipt). Net benefits are small (societal net benefits of about (emotional and behavioral)—​small, received, and percent receiving welfare $500 and $400 in govt savings per treatment member). mixed effects found. (up to twenty quarters after random For voluntary programs, more expensive programs produce assignment). larger impacts. Program participants earn about 10% more than the control group, but the effect fades (as does the employment effect); welfare receipt is reduced. Labor market controls suggests programs are more effective when demand for labor is greater.

Greenberg, Cebulla, and Bouchet 2005

Employment Retention and Advancement Examined course-​taking (types) and the completion of Data suggest that not all of the (ERA) program designed to encourage qualifications or credentials. ERA increased the likelihood training was motivated by the ERA human capital development. Personal of course-​taking and the probability of combining work financial incentives. adviser and financial support for and training, but there is no evidence yet of an effect of this training among low-​wage workers increased training on qualifications; it also did not affect and financial incentives (bonuses) for total time spent in training, but it did increase enrollment in completing training and working full courses relevant to specific occupations. time; targeted lone parents and long-​ Outcomes from training were only analyzed qualitatively term unemployed in the UK. in this report. Five-​year impact evaluation findings were Randomized controlled trial with outcomes expected in 2011, but no publication is evident yet. measured twelve months and twenty-​ four months after random assignment; sample sizes of approx. 2,293 and 1,248.

Hendra, Ray, Vegeris, Hevenstone, and Hudson 2011

In almost all states, Adult program impacts are positive—​ Costs incurred in the WIA program earnings benefits are smaller in the first four to six quarters were not available; using than after two to three years. Average increment in earnings for available data from published women is nearly $2,400 per year, about 26% of average earnings, sources, average per capita direct and for men it is nearly $1,700, about 15% of average earnings. expenditures were estimated to Program participation increases employment in a given quarter be in the range of $2,400–​$2,700, for women by about 7 percentage points and for men by about with higher costs for Dislocated 6 percentage points. Workers ($2,800–​$3,200). Increments in annual earnings for dislocated workers are much smaller than for the Adult program, just over $500 for women and less than $150 for men (10% on average) in most of the Eurozone countries and, in countries like Greece and Spain, it was over 20% in 2016 (OECD 2016b) before gradually declining over the rest of the decade to levels comparable to or better than the United States in some countries. The global COVID-​19 pandemic in 2020 and 2021 again roiled labor markets around the world. The severity and speed of the economic contraction it generated was greater than the Great Recession as numerous countries from China to the United States to the European Union imposed lockdowns and adopted various public health measures to slow the spread of the virus. The result of these measures was an unprecedented contraction in both labor supply and demand. In the United States, real GDP would decline 5.0% in the first quarter of 2020 and then 31.4% in the second quarter. Between February and April 2020, over 22 million jobs would be lost, erasing nearly all of the employment gains since February 2010. These record-​setting adverse shocks were followed by a 33.4% growth in the third quarter and rapid increases in employment brought on by both an easing of public health restrictions and trillions in fiscal (and monetary) stimulus under Presidents Trump and Biden. Nonetheless, there were still about 10 million fewer jobs at year’s end and still 8 million fewer in April 2021 than in February 2020, and the rate of recovery has slowed. On average, GDP fell 6.1% during 2020 in the European Union, but there was substantial variation in how individual countries faired economically during the pandemic. GDP in Spain and Italy declined 10.8% and 8.9%, respectively, in 2020 while it actually grew by 3.4% and 1.8% in Ireland and Turkey, respectively. Not surprisingly, unemployment also surged over the year in most EU countries but not to the extent seen in the United States, potentially making for another interesting case study of the role of active and passive labor market policies. While policymakers in the United States are no longer facing a labor market in severe crisis, it is one in which the public is increasingly pessimistic about its ability to deliver a rising standard of living. Policy action is constrained by partisan political gridlock that caused a shutdown of the federal government with the United States coming within hours of defaulting on its sovereign debt. In 2015, the deficit, in dollars and as a share of GDP, was actually lower than it was before the recession. However, in 2015, the Congressional Budget Office projected that, between 2015 and 2026, the deficit would rise from 2.5% of GDP to 4.9% while the ratio of publicly held debt to GDP would jump from 73.6% to

31 1

Labor Policy in Aftermath of Great Recession  311 85.6% and continue rising to 103% of GDP by 2040 (see CBO 2015b; CBO 2016). This looming crisis is the continuing result of a lack of political will to address long-​term fiscal overhang associated with the flat revenues (as a share of GDP) and rising entitlement expenditures for Social Security and medical care (primarily but not exclusively for the elderly). Since 2015, little to nothing has been done to address this looming crisis. In fact, the Trump tax cuts in 2016 and the three COVID-​19 spending bills in 2020 and 2021 have combined to increase the deficit to 10.3% of GDP in 2021 (the highest level since World War II). Deficits (as a share of GDP) are projected to come down some over the next few years but remain higher than normal. Indeed, the earlier projections were “too rosy,” as now CBO projects deficits of 5.7% of GDP and a debt-​to-​GDP ratio of 107 by 2031. While the Great Recession certainly exacerbated the problem of sovereign debt in the United States and Europe, it did not create it. The sustainability of debt levels in Greece, Portugal, Spain, Italy, and elsewhere in the European Union are a major source of uncertainty and risk for the EU as both a political and economic union, and its (and indeed the world’s) economy. The vote by the United Kingdom to exit the EU only heightened this economic uncertainty. Whether in the United States or Europe, policymakers will likely have to operate in a more austere fiscal environment where the public, in many countries, seems to be increasingly skeptical that their economies can create strong growth and, more importantly, enough good-​paying jobs. Is this the new normal, or can policy reverse or ameliorate these trends? If so, what types of policies are needed? These questions underlie both the evidence I discuss and the recommendations I propose in this chapter. Labor market activation policies have a role to play in this new environment because they offer the prospect of both increased and improved re-​employment with reduced social welfare expenditures. This dual benefit has made them an increasingly popular tool in Europe in labor market reform proposals. In this chapter, I look at that experience to see if there are lessons that can be drawn for the United States as the country rethinks labor policy in the years ahead.

ECONOMIC POLICY DURING THE GREAT RECESSION AND ITS AFTERMATH The rapid deterioration in the economy and labor market in 2008 brought conflicting calls for action that went beyond the steps taken by the Federal Reserve and Treasury to shore up the financial sector. Whether you came to view Quantitative Easing and the Troubled Asset Relief Fund as a taxpayer-​funded bail out of Wall Street or as prudent moves to stave off global collapse would color the political debate on both the left and the right with increasing force over

312

312  Work and the Social Safety Net the remainder of President Obama’s term in office. In Europe and in the United States, populous candidates for office took center stage, with many arguing that the Great Recession proved that the economy was “rigged” in favor of the elites and against workers. Public trust in Congress, big business, the media, the courts, and most of our institutions fell to near record lows (see Gallup 2016). The election of populous reality show/​businessman Donald Trump in 2016 and the Brexit vote in the United Kingdom would perhaps have both been unthinkable without the backlash set off by these actions. While this declining faith in our institutions is decades old, the Great Recession seems to have crystallized it. What role macroeconomic policy played in curing the Great Recession is open to debate. As the Great Recession unfolded, first President Bush then President Obama sought to further stimulate the economy with fiscal policy. The $152 billion Economic Stimulus Act of 2008, which consisted of two-​thirds tax cuts and one-​third spending increases, passed with the majority of Republican and Democrats voting in favor, while the $837 billion American Recovery and Reinvestment Act (ARRA), which consisted of roughly two-​thirds (63%) spending increases and one-​third tax cuts, passed with the majority of Democrats voting in favor along with three Republicans in the Senate and none in the House (see CBO 2015a). While ARRA was certainly larger in scale than the Economic Stimulus Act, so was the economic crisis at the time of its passage. Nonetheless, opponents decried it as wasteful or even counterproductive while some progressives worried it was too small to boost the economy (see Cato Institute 2009; Krugman 2009). The overall impact of ARRA on the economy’s recovery is a topic of hot debate. While there is no doubt that the optimistic projections of the President’s top economists did not come close to holding true, studies by Wilson (2012) as well as by CBO, Moody’s, and HIS Global Insights confirm the findings of Dan Wilson of the positive impact on employment and the positive net benefits to the economy. (In 2014, 82% believed ARRA reduced unemployment and 56% believed the benefits outweighed the cost; see IGM Economic Experts Panel 2012, 2014.) It is important to note that just because a set of policies may have net positive benefits during an economic downturn does not mean that they did not have important distortionary impacts nor that they should remain in place as the economy returns to full employment. Furthermore, even if overall ARRA was beneficial, it does not follow that each component generated net positive benefits. Finally, net positive benefits need not mean that there was no reduction in economic efficiency. In particular, some worry that the significant temporary expansions in the safety net associated with ARRA contributed to increasing the duration of unemployment and lowered the share of the population employed during the Great Recession by reducing the incentives to work (see Barro 2010). These effects, to the extent they existed, would have been temporary and need to be considered alongside the benefits from extending the social safety net in terms

31 3

Labor Policy in Aftermath of Great Recession  313 of improved consumption smoothing over the business cycle and improved growth. Nonetheless, there is concern that the interaction between the various phase-​ out rules for benefits under food stamps, the earned income tax credit (EITC), Medicaid, disability insurance (DI), the Affordable Care Act, and unemployment insurance (UI) may have generated high implicit marginal tax rates for moderate-​income benefit recipients. The combination of high implicit marginal tax rates with weak or nonexistent work tests (job search requirements) has been cited as a potential cause of the failure of the share of the population in employment to return to pre-​recession levels, increased long-​term unemployment, and the modest pace of economic growth (see Mulligan 2012). Below we review recent studies for evidence on these concerns both from Europe and the United States and look at various policy options other countries have explored to overcome or ameliorate them. Two of the most prominent changes in the social safety net during the Great Recession were designed to help the unemployed. Regular UI benefits were supplemented in many cases by those from the Extended Benefits (EB) and the Emergency Unemployment Compensation (EUC) programs that, with their various extensions, would result in some workers getting up to 99 weeks of benefits. The evidence seems clear that these extensions did contribute to increasing the durations of unemployment among beneficiaries. While estimates of the exact magnitude or importance of these effects vary across studies, they seem to indicate that the unemployment rate in 2010 increased from around 0.2 to 1.0 percentage points due to the UI expansions. Rothstein (2011), however, suggests that half or more of this effect is the result of greater labor force attachment rather than reduced re-​employment probabilities. Farber and Valletta (2015) and Farber, Rothstein, and Valletta (2015) looked at the impact of both the extensions of UI benefits during the Great Recession and their subsequent phase out in 2012–​2013. They, too, find that the increases in benefit duration during the Great Recession served to raise the unemployment rate and, conversely, their phase-​ out after 2012 served to lower the unemployment rate. In both cases, however, these impacts on the unemployment rate primarily came not from changes in worker incentives to find jobs but from changes in their decision whether to stay in or withdraw from the workforce. The declining attachment to the workforce during the phase-​out of extended benefits could help answer the riddle of why the share of the population employed remained lower than expected. That said, while this may be a part of the puzzle, it is a best a small piece as the estimated impact of the phase out on labor force attachment was quite small. The work of Card and Levine (2000) points out that it is important to distinguish between temporary and permanent changes in benefit duration. Through examining the New Jersey Extended Benefit program, they found that extending benefit availability had only a 1–​3 percentage point effect on the fraction of workers exhausting regular benefits. In the long-​term, however, the impact of

314

314  Work and the Social Safety Net extended benefits would be substantially higher as the fraction of workers exhausting their benefits would increase by 7 percentage points. Schmieder, von Wachter, and Bender (2012) look at the German experience with increasing duration of UI benefits and find that the negative impact on unemployment duration is lower during recessions. Using data from the United States, Kroft and Notowidigdo (2016) find that these moral hazard costs of extended benefits are lower when the unemployment rate is high, while the benefits from increased consumption smoothing do not vary with the unemployment rate. Schmieder, von Wachter, and Bender (2016) point out another channel through which UI benefits may have long-​term distortionary impacts: namely, wages. On one hand, UI extensions increase the value of outside options for workers (and, hence, their bargaining power or reservation wage) thereby leading to an increased equilibrium wage. Conversely, UI extensions reduce search effort and increase the duration of non-​employment, which in turn could generate stigma or skill decay effects leading to lower re-​employment wages. Thus, to fully assess the impact or costs of UI benefit expansions, one also needs to see how an increase in the duration of unemployment translates into decay in long-​term earning. Schmieder et al. estimate the impact of time out of work on re-​employment wages for middle-​aged workers in Germany and find that each extra month of non-​employment is associated with a 0.8% decline in earnings. This focus on the moral hazard effects of UI extensions is important but incomplete if one wanted to fully assess programs benefits and costs. First, one of the primary reasons for the creation of the program is to allow unemployed workers to smooth their consumption over periods of unemployment and, in so doing, prevent aggregate consumption and demand from falling (serving as an automatic stabilizer). Gruber (1997) has shown that the observed decline in consumption among the unemployed would be more than three times higher in the absence of UI. Second, work by Chetty (2008) suggests that 60% of the adverse effect of UI on the duration of unemployment spells is due to the welfare-​ enhancing removal of liquidity constraints, and only 40% is due to distortionary moral hazard incentives. Finally, work by Levine (1993) and Lalive et al. (2015) has shown that prior estimates of the impact of UI benefits on the duration of unemployment at the individual (micro) level overstate the economy-​wide (macro) impact. Macro estimates could theoretically be larger or smaller than micro estimates as they incorporate the spillover effect, or externality, generated by UI for non-​recipients and firms. Increasing the generosity of benefits may lower recipient search intensity, but, in the aggregate, this increases the likelihood that any search by nonrecipients will be successful (a positive market externality). Conversely, more generous benefits raise reservation wages, lowering the return to firms from creating a vacancy (a negative market externality). Levine (1993) used US data while Lalive et al. (2015) looked at the Austrian UI system’s experience with temporary extensions in benefits. Both find that the macro effect is substantially smaller than research

31 5

Labor Policy in Aftermath of Great Recession  315 on the micro effects would suggest. While UI expansions or extensions, especially during a downturn, may be less socially costly than previously thought, this does not mean that policymakers should ignore the moral hazard effects associated with UI or other social safety-​net benefits. The more lax program eligibility enforcement, the more likely these distortions are to occur. These effects are particularly important for populations with more tenuous labor force attachments and for those who may be eligible for multiple programs. The impact of the various social safety-​net programs on employment can often be difficult to untangle. For instance, work by Inderbitzen et al. (2016) shows that among older workers in Austria an increase in the duration of UI benefits lowered labor force participation in part by increasing the incidence of early retirement while also lowering participation in DI programs. The presence of program complementarity (i.e., UI and early retirement) and program substitution (UI and DI) seem to be important considerations for policymakers both in deciding the optimal level of benefits and in thinking through the incentive effects of various program design features. By 2015, the share of the civilian noninstitutionalized population aged twenty to sixty-​four receiving DI in the United States had more than doubled since 1990 (Figure 12.6), and the rising unemployment rate during the Great Recession was positively correlated with a surge in applicants (and a much more modest rise in awards). Extending the duration of UI benefits, if it were to forestall marginally disabled workers from going on DI, could yield substantial savings in part because, once on DI, workers rarely return to the work force. Mueller, Rothstein, and von Wachter (2016), however, find that the populations served by UI and DI are quite distinct, as most DI recipients had no labor force attachment in the year before collecting benefits. As a result, while changes in the unemployment rate may be an important determinant of participation in both programs, there seems to be little evidence in the United States of program substitution between UI and DI such that changes in one program’s rules lead to changes in enrollment in the other. Just as the Great Recession brought a surge in DI applicants, it also seems to have ushered in an expansion in the Supplemental Nutrition Assistance Program (SNAP; aka food stamps) participation. Between 2007 and 2013, the percent of the population receiving food stamps increased from about 8.5% to 15% (Figure 12.5). Ganog and Liebman (2013) find that changes in local unemployment rates explain about two-​thirds of this rise in enrollment. They find that changes in program eligibility rules associated with welfare reforms (time limits) and state policies (income and asset thresholds and temporary changes for childless adults) explain the rest of the increase in food stamp participation. Interestingly, despite the low unemployment rates found in 2018 and 2019, SNAP participation never returned to pre-​Great Recession levels. As seen in Figure 12.6, it is also striking how successful welfare reform and the strong economy of the mid-​1990s was at shrinking the rolls of Temporary Assistance for Needy Families (TANF)

60,000

Average SNAP Participants (000)

50,000 40,000 30,000 20,000 10,000

19 6 19 9 7 19 1 7 19 3 7 19 5 7 19 7 7 19 9 81 19 8 19 3 8 19 5 87 19 8 19 9 9 19 1 93 19 9 19 5 9 19 7 9 20 9 0 20 1 0 20 3 0 20 5 0 20 7 0 20 9 11 20 1 20 3 1 20 5 1 20 7 19

0

Participants (ooo)

Figure 12.5  Food stamp recipients. Source: USDA Food and Nutrition Service data, May 7, 2021.

Recipients of Government Assistance (millions) 10.000 9.000 8.000 7.000 6.000 5.000 4.000 3.000 2.000 1.000 0.000 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020

AFDC/TANF (millions) Workers on DI (million) Figure 12.6  Recipients of other assistance programs. Source: Congressional Research Service, RL32760, table A-​1 Trends in Cash Assistance Caseload, January 25, 2021 and Annual Statistical Report of Social Security Disability Program, 2019.

31 7

Labor Policy in Aftermath of Great Recession  317 and equally remarkably how there was an almost total absence of bounce-​back during the Great Recession despite significant increases in unemployment and poverty in the country. Was it the long-​standing activation efforts and lifetime benefit receipt caps that limited demand or the block grant nature of the funding mechanism which gave states a strong incentive not to let those rolls grow again, especially when those funds could be used elsewhere in strained budgets? It is possible that the dramatic increase in the duration of UI benefits during the Great Recession may have helped keep workers attached to the labor market and off the TANF rolls. If so, should we have seen some pick-​up in TANF as benefit durations were rolled back? While the answer awaits more rigorous analysis, the link between changes in the number of poor people potentially eligible for TANF and the number receiving UI benefits seems tenuous. First, to be eligible for UI benefits an individual must have an established consistent work history, which many poor people do not have. Indeed, in 2016, less than three in ten unemployed workers collect UI benefits, which is the lowest rate since 2000 (author’s calculations using data from the weeks of November 18, 2000, and May 21, 2016 [Bureau of Labor Statistics (BLS) 2016; Employment and Training Administration (ETA) 2016]). Even at the depth of the Great Recession in 2009, less than half (45%) of the unemployed were receiving UI benefits (Based on author’s calculations using the four-​week moving average of continuing UI claimants/​number of unemployed workers for the week of June 13, 2009, and month of June, respectively [BLS 2016; ETA 2016]). Second, adding discouraged workers and others who are marginally attached to the count of unemployed, the safety net, as represented by UI benefits, clearly left broad swaths of the non-​ employed population excluded. If the poor did not get UI benefits, where did they find support? As noted above, the 2000s were associated with steady increases in food stamps enrollment. In thinking about social insurance program incentives, we must be cognizant of both the economic efficiency cost as well as the benefits. There has been considerable concern over the high cost of the food stamp program, but recent work by Hoynes et al. (2016) present evidence that access to food stamps in utero and early childhood leads to better adult health and economic outcomes. Did food stamps become the new program of last resort? Understanding any program spillover effects needs to be considered as part of any reform efforts. When it comes to UI benefits, however, it is important to note that there is very little evidence of complementarity between it and other benefit programs. In other words, households that receive UI benefits do not appear to also supplement their income from other benefit programs, perhaps because most of these other benefit programs have family income and/​or asset limits that might preclude married or formerly high-​earning UI recipients from qualifying. A report from the US Department of Agriculture found that, even in the depth of the Great Recession in 2009, only about 14% of those households collecting SNAP benefits also collected UI benefits during the same year, and only a little over 4% collected

318

318  Work and the Social Safety Net SNAP and UI benefits concurrently (see Finifter and Prell 2013). To put this in context, despite the severity of the Great Recession at its worse point in 2009, only a little over 1% of all households in the United States collected benefits from both programs. While UI and SNAP participation did not seem to go together, it is striking that, despite the increase in poverty during the Great Recession, only about 7% of SNAP households were also getting TANF income (see Gray 2014). These programs that were designed to complement each other seem to have become substitutes. Nonetheless, it is certainly possible that changes in the rules and structure of one program may have caused the enrollment in other social welfare programs to change in response to economic shocks. Furthermore, either individually or collectively, the various programs in the social safety net may impose high marginal tax rates on work for certain workers resulting in moral hazard or potentially harmful disincentives to work. A key question then is what can be done to offer protection for those displaced or under economic distress while maintaining labor force attachment and work as the most effective means of supporting families? Can labor market activation policies offset some of these effects? Various states experimented with programs that required UI recipients to come in for mandatory eligibility reviews and report on job search behavior, combined with offers of re-​employment services. These simple actions seem sufficient to raise the opportunity cost of unemployment enough to make those who were either noncompliant with search requirements or the most readily employable drop off the UI rolls. This is not altogether surprising since numerous studies found that the work-​related sanctions in TANF had significant negative impacts on caseloads (see Danielson and Klerman 2008). In the case of TANF, Grogger and Karoly (2005) also found that work sanctions led to increased employment and earnings. The evidence on the impact of work-​related sanctions in UI programs is more mixed as to whether they lead to improved labor market outcomes. Furthermore, it is also not clear whether these effects extend to reducing aggregate unemployment or simply make those workers subject to the reviews find employment at the expense of other workers. Finally, Herbst (2008) found that the impact of work sanctions in TANF on employment were substantially greater in good times than in bad. Activation policies in many European countries have, since the 1980s and increasingly through the 1990s, played a central role in helping them to reduce unemployment. Given the extensive social welfare system and generous benefit support scheme, concerns over the role of moral hazard in generating high unemployment and dependency led many of these countries to look for new ways to improve economic performance while maintaining their social cohesion. In 2016, the United States was second only to Mexico among OECD countries in spending the smallest share of national income on labor market programs.6 6. OECD spending on labor market programs includes spending for employment services, job training, employment incentives, sheltered employment, direct job creation, startup incentives, UI or other out of work income maintenance, and early retirement.

31 9

Labor Policy in Aftermath of Great Recession  319 Aside from increasing the level of such support, shifting the composition of spending toward active labor market programs has been thought to help facilitate lower unemployment, improve economic outcomes, and ultimately lower expenditures on passive programs (benefits). During the years after the Great Recession, while the United States spent two to three times more on benefit payments (passive programs) than it did on active programs, the Nordic countries (Sweden, Denmark, Norway) and the United Kingdom actually spent more on activation efforts than on passive benefits (see Martin 2015, table 1). These countries were all more generous in their benefit levels than the United States, so this greater relative spending on active programs did not reflect low spending on passive programs. Across the OECD (excluding the United States), about 59% of labor market program expenditures went to passive programs, while in the United States about 70% went to passive programs. Spending on active labor market programs per unemployed worker as a share of per capita GDP was 5% in the U.S and 19% in the other OECD countries (see Nie and Sturby 2011). Martin (2015) provides an extensive review of studies on the European experience with active labor market policies (ALMPs) in OECD countries over the past thirty years. He, too, concludes that job search assistance and monitoring job seeker behavior are cost-​effective and that some training programs work if they are tied to local economic needs. He finds that targeted hiring subsidies can also work, but that they often come with significant displacement effects, and public employment programs rarely work. The positive impact of ALMPs in reducing unemployment led many European countries to extend those efforts to reach harder-​to-​employ populations such as disability recipients, early retirees and single parents on social assistance. His review suggests that activation policies worked to move single parents into the workforce (as also found in the United States), but that they have been far less successful in moving those on disability benefits into employment. Petrongolo (2009) finds that, in the United Kingdom, enhanced job search requirements succeeded in lowering the number of individuals on unemployment benefits but that they were not successful in moving them to better jobs (i.e., their post-​program earnings were actually lower even four years later) and had a negative short-​term impact on weeks worked. This increase in UI program rule stringency was also associated with an increase in the number of people on disability (incapacity) benefits. The effects tended to be bigger for youth (16–​24) than for mature workers (25–​64). This link between UI benefits and DI benefits was also found in Lammer, Bloemer, and Hochguertel (2013), in the Netherlands. They found that these enhanced job search rules increased the employment rate for workers but that there was also an increase in the number of workers on DI. Finally, a study by Blundell et al. (2004) looked at the effectiveness of a job search assistance program in the United Kingdom called the New Deal for Young People. This program focuses on eighteen-​to twenty-​four-​ year-​old youths who have been unemployed for six months or more and requires

320

320  Work and the Social Safety Net them to participate in an intensive job search assistance program. If this step does not result in employment, then recipients move to other interventions, including subsidized employment in the private or volunteer sectors, government employment, or intensive education and training. Looking at just the impact of the job search assistance component, the authors find that it raises the share of participants in employment by 20%. Michaelides et al. (2012) and Michaelides and Mueser (2016) argue that so few workers availed themselves of re-​employment services in prior experiments in the United States that while we know mandatory reviews and activity checks can reduce the moral hazard associated with UI benefits, we do not have evidence on whether these or other activation services can lead to increased employment and earnings. Their papers look at an intervention in Nevada that combined early mandatory eligibility reviews and reporting on job search activity with staff-​assisted job counseling services (e.g., work search review, re-​employment plan development, local labor market information) and voluntary access to re-​ employment services (e.g., resume and job search assistance, training). With Nevada suffering from double-​digit unemployment during this period, this program also provided evidence on the efficacy of activation schemes during bad times. They found that, even in a very weak labor market, workers given both suites of services had shorter average duration on UI, were less likely to need extended benefits, were re-​employed faster, and earned more. Random assignment experiments with mandating re-​employment and eligibility assessment services in Idaho, Florida, Illinois, Minnesota, and North Dakota during this time period have also yielded positive results in terms of lower duration on UI, increased employment, and increased earnings (see CLEAR 2016). Klerman, Minzner and Benson (2015) examined how differences in the structure of sanctions applied to those who failed to comply with state-​mandated participation in re-​employment and eligibility assessment service programs in Indiana, New York, Washington and Wisconsin. They found that compliance is highest when participants have flexibility in scheduling appointments, when benefits were withheld until the participant complied, and when the number of required meetings was minimized. Given these results, at the individual level with activation policies in the United States, the question arises: Does increased activity or spending on activation efforts lower unemployment in the labor market? Nie and Sturby (2011) provide some preliminary evidence that the answer would be yes. Increased spending on ALMPs by 1% would reduce the unemployment rate by 0.11 percentage points. In their analysis, a similar increase in passive labor market spending would increase unemployment by .01 percentage points. This work further suggests that there is little or no adverse effect from increases in the replacement ratio of UI benefits, but that increases in the duration of benefits by a year would increase the unemployment rate by 0.29 percentage points, creating significant adverse

32 1

Labor Policy in Aftermath of Great Recession  321 effects from increased duration.7 According to this study, among active policies, job search assistance and training are the most effective ways to reduce unemployment, while direct job creation is the least effective. These recent studies of labor activation policies in the United States, United Kingdom, Germany, and other countries may hold out promise. Another area where we appear to be following the growing European consensus is in work-​ sharing. These layoff prevention programs seek to provide incentives to reduce hours rather than cut employment. As noted above, these types of programs were particularly successful in helping minimize unemployment in Germany and Denmark during both the Great Recession and in response to COVID-​19. Under these programs, workers whose employers reduce their hours but keep the worker employed receive a pro rata share of their UI benefits. The employer benefits from reduced labor costs, and, by keeping the worker employed, its turnover costs from losing trained workers are cut. Workers also receive a higher income than if they had been placed on unemployment. By the mid-​2010s, about twenty-​four states had versions of work-​sharing programs, with six states and the District of Columbia adding them since the Great Recession. This expansion was partially driven by the passage of the Middle-​Class Relief and Job Creation Act of 2012, which contained provisions to streamline the process and provide federal subsidies to states that have or agree to set up these work-​sharing or short-​term work programs. While the number of employers participating in these programs in the United States remained modest, this rare bipartisan effort signaled a willingness to look at ways to restructure other programs to promote employment. The Department of Labor has commissioned an evaluation of policy issues, lessons learned, and administrative challenges associated with UI programs during the Great Recession.8 Preliminary evidence suggests that where these programs are used, short-​time compensation programs can help prevent layoffs. During the COVID-​19 pandemic, these work-​sharing or short-​time compensation type programs were used by twenty-​seven states and the District of Columbia, but the overall incidence remains modest (especially compared to Germany) because of lack of program knowledge by employers and frictions associated with participation. The European experience may have some positive lessons for policy experimentation, but it also has some negative ones in what not to do if one wants a well-​functioning labor market. While the paths of unemployment over the course of the Great Recession and its aftermath look similar in the United States, 7. The CARES Act, passed in March 2020, created a $600/​week supplemental benefit that for lower income workers dramatically increased replacement rates on average and, for some, more than replaced their pre-​layoff earnings. While subsequent legislation reduced this benefit to $300/​ week, it was extended until September 2021. While this benefit has been blamed by business for worker shortages, rigorous evaluation is still needed. 8. See “The Great Recession: Lessons Learned for the Unemployment Insurance System,” William Congdon and Wayne Vroman, forthcoming, September 2021.

32

322  Work and the Social Safety Net United Kingdom, Denmark, and, to a lesser extent, Germany, unemployment rose faster and remained significantly higher in Spain, Portugal, and for the European Union as a whole. Unemployment in Spain increased from 8.3% in 2007 to 26.1% in 2013 and remained higher than 20% in 2016. As seen in Figure 12.7, while about 25% of the unemployed in the United States were without a job for a year or more in 2013, 56.4% of the unemployed in Portugal and 49.7% of the unemployed in Spain were unemployed for a year or more (OECD 2012). Indeed, for the European Union as a whole, 46.4% of the unemployed were unemployed for a year or longer, and even in Germany, which had a lower unemployment rate than almost any other EU country, 44% of the unemployed were unemployed for a year or more. By European standards, at least, the US long-​ term unemployment problem was much less serious. This pattern of high incidence of long-​term unemployment existed despite extensive employment protection legislation (EPL) in many of these countries—​ legislation that in some cases is even enshrined in their Constitutions. OECD summaries of employment protection legislation (governing, easing of dismissal, layoff protections, and restrictions on temporary employment contracts) typically rated the United States as the least protective OECD country, while Portugal, France, Spain, and Luxembourg are rated as the most protective (OECD 2004). Bentolila et al. (2012) argue that the scale of dislocation in Spain was in fact exacerbated by these restrictions rather than abated by it. Spanish EPL in the 1980s brought extensive layoff protections and constraints on geographic and functional mobility for firms, generous UI benefits, and strong collective bargaining for workers in the primary or permanent sector. With this structure in place in the primary sector, Bentolila et al. (2012) argues that, since the 1990s, almost all hiring occurred outside these traditional employers in “temporary” or contract jobs.

Percent Unemployed for 12 months or longer in 2013 and 2019 60 50 40 30 20 10 0

Sweden

Australia Denmark

United States

United Germany Kingdom

Long Term Unemployment Rate 2013

Figure 12.7  Long-​term unemployment remains high. Source: OECD 6C Data: data.oecd.org.

EU

Spain

Portugal

Long Term Unemployment Rate 2019

32 3

Labor Policy in Aftermath of Great Recession  323 One might not be too concerned about the existence of a dual labor market structure as long as it did not serve to trap workers permanently or there was mobility across sectors that came about with experience (age) or skill. Indeed, the United States, Japan, and other countries have long had labor markets where those employed at smaller firms or in retail or service jobs had shorter mean employment durations and lower compensation. In Spain, “temporary” contracts provided opportunities for youth and others to gain access to secondary sector employment (90% of those younger than twenty-​one started on temporary contracts). However, a sizeable number of these workers get “stuck” in these temporary arrangements because employers have not been willing to transition them into permanent jobs (40% of these workers remain in “temporary” jobs at age forty-​five). As a result, efforts at EPL reform over the past two decades have attempted to both reduce employer severance costs and provide incentives to convert more employees to permanent contracts. Labor market reform efforts in Spain are not new; almost every other year between 1980 and 2016, some piece of EPL legislation was passed. What is striking about the reform efforts in Spain was their inconsistent nature. A quarter of the forty-​eight reform efforts represented major changes in the structure of the labor exchange. About 60% of the legislation increased flexibility, while 39% reduced it. Whatever lessons there are to be learned about the impact of regulatory uncertainty on economic behavior, Spain would seem to provide a good case study. Given the alternating stepping on the brake and then the gas pedal nature of Spanish EPL reform efforts, however, drawing inferences about the efficacy of any one of them should be done with a high degree of caution. Sarfati (2013) provides a survey of various labor and social policy changes that have attempted to deregulate the market for employed workers and increase activity among the non-​employed. She notes that, even before the Great Recession, the challenges of globalization and technological change combined with an aging population, changes in the structure of families, and high levels of sovereign debt gave impetus to these efforts. During the 1990s, labor market reform efforts were met with substantial resistance in many countries but were successful in others. Most of the reform efforts aimed to increase labor supply rather than change employer behavior. Along with these efforts often came reforms in the employment services designed to give more local (municipal) control over their activities and to tie funding to employment outcomes rather than to the level of services provided. In some countries, like Great Britain, the employment service has even been privatized. As a general rule, Sarfati argues that those countries that combined social insurance and wage reforms with macro stimulus elements and job training assistance were the most successful. In her view, the poster child for a successful activation strategy or reforms is Denmark. There they have achieved “flexicurity,” resulting in increased job mobility. This can be attributed to tailored training and monitoring of job search activities of those on benefits while preserving generous

324

324  Work and the Social Safety Net long-​term unemployment benefits. To some degree, Germany also followed this approach and maintained a relatively large primary sector through high levels of training and technological adoption. It also heavily subsidized short-​ time work to avoid layoffs in that sector. Strict EPL laws in Germany generated more labor hoarding in the primary sector and larger layoffs among temporary workers. During the Great Recession, firing costs and short-​time programs that encouraged hour reductions over employment cuts seem to have reduced overall employment losses and generated pro-​cyclical movements in productivity. Sarfati argues that while countries differ in the structure of their social insurance programs, most have a reciprocal structure where benefits are now tied to training or the acceptance of employment. Even in relatively weak job markets, tying benefits to the intensity of search effort put forth can have beneficial results, especially when search assistance is augmented by job training. The imposition of hiring and firing restrictions associated with EPL may have helped reduce job loss in the short-​term in Europe, but Sarfati notes that they had the unintended consequence of leading to longer mean durations of unemployment spells (especially for youth) and creating significant differences in average income and employment rates across insider and outsider groups. Women, ethnic minorities, and youth have more typically been found in the secondary sector in most of these economies. Older workers moved increasingly into a “third sector,” as judged by the dramatic increases in the number of workers receiving disability benefits. It would appear that many older workers and their employers used the disability system as a bridge in the transition from employment to retirement. Activation policies helped, but, absent attention to growth and labor market dynamics, they were not sufficient to deal with the substantial unemployment (and underemployment) that existed in the aftermath of the Great Recession.

CONCLUSION Cross-​country comparisons of policies of the type discussed in this chapter can be useful in providing initial evidence of what works and some of the unintended consequences of particular programs or policies. However, extrapolating these results from one country to another can be problematic without a clear understanding of each country’s underlying market structure and culture. Furthermore, social programs are often interrelated in their structures, so changing one program or regulation in isolation may yield very different results in Australia than in, say, Germany or the United States. Finally, as Blank and Freeman (1994) noted, just because individual employment policies or programs have particular micro incentive and behavioral impacts, it does not tell us the answer to whether the program’s benefits outweigh its costs nor what the macro (general equilibrium) consequences are of changing components of our labor/​ social insurance framework.

32 5

Labor Policy in Aftermath of Great Recession  325 Labor market and social safety-​net reform discussions in the United States start from very different reference points than those in Europe. As noted earlier, labor markets in the United States, at least by OECD standards, have had low levels of worker protection regulation, and the extent of institutionalized wage setting (by either unions or the government) has been fairly limited. In the United States, hiring and firing costs have been minimal, as reflected in various comparisons of OECD countries on employment protections in the years after the Great Recession. For example, on the OECD index of employment protection, the OECD average index was 2.04, with Portugal having a ranking of 3.18 and the United States a ranking of 0.26 (see OECD 2016a). Furthermore, only 4% of workers were at the minimum wage, and collective bargaining covered less than 10% of the workforce, making institutionalized wage setting a much more limited phenomenon. The OECD also identified a similar pattern in the generosity of social safety-​net benefits, where the United States’ spending as a percentage of GDP was dwarfed by that in the Nordic countries, Germany, and even the United Kingdom. In the aftermath of the Great Recession, the share of GDP that the United States spent on active labor market programs was about a fifth of what a typical OECD country spent (0.12% vs. 0.55%) and a fourth of what they spent on labor market programs as a whole (0.36% vs. 1.5%) (OECD 2013). This different institutional structure reflects, at least in part, the fact that, historically, the United States was characterized by a very dynamic labor market with high levels of both job destruction and job creation. Much of the latter came from new startup firms (or “gazelles”), but established firms mattered as well. Many layoffs were temporary, and most unemployment spells were short as workers moved between jobs and regions with relative ease. (This general pattern was less true for minorities and older workers, and inner cities and rural areas have had persistent pockets of long-​term unemployment.) While the evolution started before the Great Recession, it would appear that the dynamics of the US labor market and economy changed in the wake of the Great Recession. Over the past forty years we have seen stagnant real wage growth for the median worker and dramatic increases in inequality. Until recently, inequality has not generated substantial public concern because of the perception that the United States enjoys high rates of economic mobility within and across generations that prevents the creation of permanent insider–​outsider classes. Data from the Pew Foundation Economic Mobility Project (Isaacs et al. 2008) suggest that such confidence may be misplaced as, by the mid-​2010s, the United States lagged behind many European countries on various economic mobility measures. Less dynamism was also reflected in the fact that new firm births did not rebound from the pre-​recessionary level, and net job growth came mainly through reductions in the rate of job destruction (see Haltiwanger 2015). Voluntary job changes, especially among younger workers, have long been an important source of earnings growth. The lower rates of job churning that occurred in the aftermath of the Great Recession could well lead to fewer opportunities to

326

326  Work and the Social Safety Net find new employment (hence more long-​term unemployment) as well as slower earnings growth over their lifetimes. Davis and Haltiwanger (2014) have also found that those areas and populations that have experienced the biggest declines in labor market reallocation have also experienced the biggest declines in employment-​to-​population ratios since 2000. Adding to this, we see that geographic mobility declined steadily from the mid-​ 1980s to the mid-​2010s (US Census Bureau 2015). Despite less churning, median job tenure with an employer has been declining steadily (for males) since the late 1970s. While involuntary job loss has always been costly, the consequences of losing a job during the Great Recession, when the unemployment rate exceeded 8%, was estimated to be twice as high as when it is below 6% (see Davis and von Wachter 2011). A new economic reality requires a look at both our policies that impact the dynamics of our economy and the labor market and social safety-​net programs we have developed to help workers and families during dislocations. Restoring or improving labor dynamics is not the cure-​all for our labor market ills, but it is a potentially important piece of the equation to increase economic growth, improve earnings for workers, and reduce long-​term non-​employment. Speeding job transitions and supporting workers in ways that facilitate that process should be part of a rethinking and reform to our approach to the safety-​net and work support service. Equally important is a new look at the frayed structure of our social safety net that was designed for a different economic reality of brief transitions and steadily growing opportunity for the typical worker. As Martin (2015) points out, countries have been able to successfully combine generous benefit systems with activation schemes (e.g., Norway, Denmark, Switzerland) that had strict benefit conditionality and vigorous search and retraining efforts. Reforming UI to make sure it covers more of the unemployed (especially part-​time or temporary workers) for longer periods of time (with tiered benefits reflecting duration and automatic adjustment when the economy is bad) has been done both in Nordic countries and in countries with relatively less generous benefit systems (e.g., the United Kingdom and Japan). Key to these efforts is matching them with well-​designed activation efforts that can help move people back to employment more quickly. That means better monitoring, better search assistance, and better training programs all tied more closely to employers in local labor markets. Some, like Australia, chose to privatize the delivery of these employment services as part of its activation efforts, but most of the other countries who have done so continued to use the public employment service. In either case it would be critical to increase our investments in the capabilities of those delivering employment services as well as develop the right incentives and performance metrics commensurate with the enhanced role we expect them to play. Strict monitoring and performance standards for those delivering re-​employment services are needed if these efforts are to be successful. Cream-​skimming and a tendency to avoid the hardest to re-​employ is a challenge for public employment

32 7

Labor Policy in Aftermath of Great Recession  327 agencies, but it is likely to be even more acute in a privatized service delivery model unless performance contracts are carefully written. Martin (2015) notes that those countries that have kept the public employment service while enacting activation policies have tried to find ways to devolve more control to local entities that have better information and connections with employers. Block-​granting programs to the states is one way to achieve this, but not the only way. Furthermore, block grants can contain incentives that work counter to the goal of improved service, especially if funds are capped in dollar terms rather than per capita because this builds in the incentive to minimize the number of those getting assistance without a countervailing incentive to maximize re-​employment and earnings of the unemployed. That said, strong ties to local employers and the flexibility to adapt to local conditions is critical even in these days of globalization. While activation schemes can help limit the extent of moral hazard or the efficiency costs of having a broader social safety net, it should be kept in mind that they are not a cure-​all for the labor market. While they help increase employment or reduce unemployment on the margin, no one has found significant macro labor market gains from doing so. They work best with those who are most job ready and in the context of a growing economy. They have not proved successful in helping those on disability and are of only limited help for other long-​term non-​employed populations. Done right, they can help people get a job. Done wrong, they often end up chasing people from one benefit program to another or off benefits altogether, leaving them in an increasing state of poverty. Increases in earning are generally modest, but when they occur it is generally from well-​designed training programs. Too many of our training programs, whether provided by the government or private-​sector vendors, yield little to no appreciable earnings gain. Those that do work tend to be closely aligned with potential employers, provide soft and hard skills, and think holistically about the barriers the worker faces to getting and keeping a job. As the economy continues to rebound from the effects of the global pandemic there is concern about the pace of job creation and whether it will generate appreciable earnings growth for the median or typical worker. As was the case after the Great Recession, these superficial concerns rest, however, on two very different diagnoses of the problem. Some economists focused on the need to increase work incentives through scaling back food stamps, DI, and the ACA combined with supply-​side efforts to stimulate an increase in the economy (tax cuts, tort reform, labor, and environmental regulation overhauls, etc.) (see Taylor 2016 or Mulligan 2012). Others viewed the problem as rooted in a lack of demand due to increased inequality and austerity-​based government policies and call for increased investment in infrastructure projects, increased minimum wages, and tax and regulatory reforms to encourage companies to increase investment in innovation, repatriate earnings, and create jobs domestically (see Stiglitz 2016; Summers 2016). Both sides would agree that boosting the rate of growth in the

328

328  Work and the Social Safety Net Table 12.1  Economic indicators in selected OECD countries United States (%) Real GDP growth 2007–​09

Germany Australia (%) (%)

−3.1

−4.8

−5.8

−4.6

Real GDP growth 2014

2.5

2.2

2.2

1.6

2.9

Real GDP growth 2019

2.2

0.6

−0.3

2.8

1.4

10.2

9.2

6.3

7.9

8.4

2.90

4.50

4.33

3.36

Highest harmonized unemployment rate 2007–​16

4.49

Denmark United (%) Kingdom (%)

Harmonized unemployment rate 2017–​2020

4.02

Share of unemployed 12 months of longer in 2011

31.3

47.9

18.9

24.4

33.5

Share of unemployed 12 months or longer in 2014

23.0

44.3

21.8

25.2

35.7

Employment/​Population 2009

67.6

70.4

72.1

73.6

70.6

Employment/​Population 2015

68.7

74.0

72.2

72.1

73.4

Employment/​Population 2019

71.4

76.7

74.3

75.2

75.6

Passive labor market program spending/​GDP 2013

.24

1.01

.64

1.66

.32

Spending on active labor market programs/​GDP 2013

0.11

.67

.23

1.82

.23

Spending on incapacity (disability) programs/​GDP 2010

1.2

2.0

2.3

5.9

2.0

Source: OECD.Stat.

economy is a necessary condition for sustained improvement in the employment rate or earnings growth. Mixing in activation policies with these efforts could augment the impact on unemployment or non-​employment as well as earnings. The United States and Europe are still struggling with how to maintain our post-​World War II sense of social cohesion in which economic risks and rewards are broadly shared in a more global competitive environment where we must compete with nations paying far less and offering little in the way of a safety net. Denmark and others with their “flexicurity” have shown it is possible to be both competitive and have a robust social safety net. Finding an American version of flexicurity likely will involve fixing the holes to better cover the population, better integration and coordination between programs, and some simplification or even consolidation of the myriad of programs that make up our current safety net. It will require more targeted job training with stronger ties to employers, enhanced job search requirements and better assistance, investing in and reforming of the employment service with increased emphasis on performance, and increased temporary job creation and incentives for sharing. Perhaps

32 9

Labor Policy in Aftermath of Great Recession  329 an angry public fed up with what many perceive as inaction by the institutions designed to serve them will finally provide the impetus for policymakers to start moving forward on this agenda.

REFERENCES Barro, Robert. The Folly of Subsidizing Unemployment. The Wall Street Journal, August 30, 2010. http://​www.wsj.com/​artic​les/​SB10001424052748​7039​5970​ 4575​4544​3145​7720​188. Bentolila, Samuel, Juan Dolado, and Juan Jimeno. Reforming an Insider-​Outsider Labor Market: The Spanish Experience. Journal of European Labor Studies 1, no. 4 (2012): 1–​29 Blank, Rebecca, and Richard Freeman. Evaluating the Connection Between Social Protection and Economic Flexibility. In Social Protection Versus Economic Flexibility: Is There a Trade-​Off? ed. Rebecca Blank. Chicago: University of Chicago Press, 1994:21–​41. Blundell, Richard, Monica Costa Dias, Costas Meghir, and John Van Reenen. Evaluating the Employment Impact of a Mandatory Job Search Program. Journal of European Economic Association 2, no. 4 (2004): 569–​606. Bureau of Labor Statistics (BLS), U.S. Department of Labor. A–​1. Employment Status of the Civilian Non-​institutional Population 16 Years and Over, 1981 to Date, 2016. http://​www.bls.gov/​web/​emp​sit/​cpsee​a01.htm. Card, David, and Phillip Levine. Extended Benefits and the Duration of UI Spells: Evidence from the New Jersey Extended Benefit Program. Journal of Public Economics 78 (2000): 107–​138. Cato Institute. With All Due Respect Mr. President, That Is Not True, 2009. https://​obj​ect.cato.org/​sites/​cato.org/​files/​pubs/​pdf/​cato_​st​ imu​lus.pdf. CBO (Congressional Budget Office). Estimated Impact of the American Recovery and Reinvestment Act on Employment and Economic Output in 2014. Washington, DC: Congressional Budget Office, 2015a. https://​www.cbo.gov/​ publ​icat​ion/​49958. CBO (Congressional Budget Office). The 2015 Long-​Term Budget Outlook (June). Washington, DC: Congressional Budget Office, 2015b. http://​www.cbo.gov/​ sites/​defa​ult/​files/​114th-​congr​ess-​2015-​2016/​repo​rts/​50250-​LongTe​rmBu​ dget​Outl​ook-​3.pdf. CBO (Congressional Budget Office). Updated Budget Projections: 2016 to 2026 (March). Washington, DC: Congressional Budget Office, 2016. https://​www. cbo.gov/​sites/​defa​ult/​files/​114th-​congr​ess-​2015-​2016/​repo​rts/​51384-​March​ Base​line​_​One​Col.pdf. CBPP (Center on Budget and Policy Priorities). Chart Book: The Legacy of the Great Recession, 2016. http://​www.cbpp.org/​resea​rch/​econ​omy/​chart-​book-​ the-​leg​acy-​of-​the-​great-​recess​ion.

30

330  Work and the Social Safety Net Chetty, Raj. Moral Hazard Versus Liquidity and Optimal Unemployment Insurance. Journal of Political Economy 116, no. 2 (2008): 173–​234. CLEAR (Clearinghouse for Labor Evaluation and Research), U.S. Department of Labor. The Reemployment and Eligibility Assessment, 2016. http://​clear.dol. gov/​study-​summ​ary-​relati​onsh​ips/​reemp​loym​ent-​and-​elig​ibil​ity-​ass​essm​ent. Danielson, Caroline, and Jacob Klerman. Did Welfare Reform Cause the Caseload Decline? Social Science Review 82, no. 4 (2008): 703–​730. Davis, Steven J., and John Haltiwanger. Labor Market Fluidity and Economic Performance. NBER Working Paper No. 20479. Cambridge, MA: National Bureau of Economic Research (NBER), 2014. Davis, Steven J., and Till von Wachter. Recessions and the Cost of Job Loss. Brookings Papers on Economic Activity 43, no. 2 (2011): 1–​72. Employment and Training Administration (ETA), U.S. Department of Labor. Unemployment Insurance Weekly Claims Data, Report r539cy, 2016. http://​ workfo​rces​ecur​ity.dol​eta.gov/​unemp​loy/​wkcla​ims/​rep​ort.asp. Faberman, R. Jason, Andreas Mueller, Aysegul Sahim, and Giorgio Topa. The Shadow Margins of Labor Market Slack, National Bureau of Economic Research Working Paper 26852. Cambridge, MA: Nantional Bureau of Economic Research (NBER), March 2020. Farber, Henry, and Robert Valleta. Do Extended Unemployment Benefits Lengthen Unemployment Spells? Evidence from Recent Cycles in the US Labor Market. Journal of Human Resources 50, no. 4 (2015): 873–​909. Farber, Henry, Jesse Rothstein, and Robert Valleta. The Effect of Extended Unemployment Insurance Benefits: Evidence from the 2012–​2013 Phase-​Out. American Economic Review: Papers and Proceedings 105, no. 5 (2015): 171–​176. Finifter, David H., and Mark A. Prell. Participation in the Supplemental Nutrition Assistance Program (SNAP) and Unemployment Insurance: How Tight Are the Strands of the Recessionary Safety Net? Economic Research Report No. 157. Washington, DC: U.S. Department of Agriculture, Economic Research Service, 2013. Gallup. Confidence in Institutions, 2016. http://​www.gal​lup.com/​poll/​1597/​con​ fide​nce-​insti​tuti​ons.aspx. Ganog, Peter, and Jeffrey Liebman. The Decline, Rebound, and Further Rise in SNAP Enrollment: Disentangling Business Cycle Fluctuation and Policy Changes. NBER Working Paper No. 19363. Cambridge, MA: National Bureau of Economic Research (NBER), 2013. Glick, Reuven, and Kevin Lansing. Global Household Leverage, House Prices and Consumption. FRBSF Economic Letters 2010-​01 (January 2010): 1–​5. Gray, Kelsey Farson. Characteristics of Supplemental Nutrition Assistance Program Households: Fiscal Year 2013. Nutrition Assistance Program Report Series No. SNAP-​14-​CHAR (December). Alexandria, VA: United States Department of Agriculture, Food and Nutrition Service, Office of Policy Support, 2014.

33 1

Labor Policy in Aftermath of Great Recession  331 Grogger, Jeffrey, and Lynn Karoly. Employment, Labor Supply and Earnings. In Welfare Reform: Effects of a Decade of Change, 134–​154. Cambridge, MA: Harvard University Press, 2005. Gruber, Jonathan. The Consumption Smoothing Benefits of Unemployment Insurance. American Economic Review 87 (March 1997): 192–​205. Haltiwanger, John. Top Ten Signs of Declining Business Dynamics and Entrepreneurship in the U.S. Unpublished paper presented at the Kauffman Foundation New Entrepreneurial Growth Conference, June 2015. Herbst, Chris. Do Social Policy Reforms Have a Different Impacts on Employment and Welfare Use as Economic Conditions Change? Journal of Policy Analysis and Management 27, no. 4 (2008): 867–​894. Horowitz, Juliana, Ruth Igielnik, and Rakesh Kochhar. Trends in Income and Wealth Inequality. Washington, DC: Pew Research Center, January 9, 2020. Hoynes, Hillary, Diane Whitmore Schanzenbach, and Douglas Almond. Long Run Impacts of Childhood Access to the Safety Net. American Economic Review 106, no. 4 (2016): 903–​934. IGM Economic Experts Panel. Economic Stimulus, 2012. http://​www.igm​chic​ ago.org/​igm-​econo​mic-​expe​rts-​panel/​poll-​resu​lts?Surve​yID=​SV_​cw5O​9LNJ​ L1oz​4Xi. IGM Economic Experts Panel. Economic Stimulus (Revisited), 2014. http://​ www.igm​chic​ago.org/​igm-​econo​mic-​expe​rts-​panel/​poll-​resu​lts?Surve​yID=​ SV_​5bfA​Rfql​uG9V​YrP. Inderbitzin, Lucas, Stefan Staubli, and Josef Zweimüller. Extended Unemployment Benefits and Early Retirement: Program Complementarity and Program Substitution. American Economic Review: Economic Policy 8, no. 1 (2016): 253–​288. Isaacs, Julia, Isabel Sawhill, and Ron Haskins. Getting Ahead or Losing Ground: Economic Mobility in America. The Brookings Institution and Pew Foundation Report. Washington, DC: The Brookings Institution, 2008. Kapon, Samuel, and Joseph Tracy. A Mis-​Leading Labor Market Indicator. Liberty Street Economics, February 3, 2014. http://​lib​erty​stre​etec​onom​ics.new​ york​fed.org/​2014/​02/​a-​mis-​lead​ing-​labor-​mar​ket-​indica​tor.html. Klerman, Jacob A., Amy Minzner, and Valerie Benson. Understanding Program Non-​ Compliance in Mandatory Social Programs. Abt Associates Report (November). Boston: Abt Associates, 2015. Kroft, Kory, and Matthew Notowidigdo. Should Unemployment Insurance Vary with the Unemployment Rate? Theory and Evidence. The Review of Economic Studies 83, no. 2 (2016): 1–​33. Krugman, Paul. Nobel Laureate Paul Krugman: Too Little Stimulus in Stimulus Plan, 2009. http://​knowle​dge.whar​ton.upenn.edu/​arti​cle/​nobel-​laure​ate-​ paul-​krug​man-​too-​lit​tle-​stimu​lus-​in-​stimu​lus-​plan/​.

32

332  Work and the Social Safety Net Lalive, Rafael, Camille Landais, and Josef Zweimuller. Market Externalities of Large Unemployment Insurance Extension Programs. American Economic Review 105, no. 12 (2015): 3564–​3596. Lammers, Marloes, Hans Bloemen, and Stefan Hochguertel. Job Search Requirements for Older Unemployed: Transitions to Employment, Early Retirement and Disability Benefits. European Economic Review 58 (2013): 31–​57. Levine, Phillip. Spillover Effects Between the Insured and Uninsured Unemployed. Industrial and Labor Relations Review 47, no. 1 (1993): 73–​86. Martin, John. Activation and Active Labour Market Policies in OECD Countries: Stylized Facts and Evidence on their Effectiveness. Labour Policy Journal 4, no. 4 (2015): 1–​29. Michaelides, Marios, and Peter Mueser. Are Reemployment Services Effective? Evidence from the Great Recession. University of Cyprus Working Papers in Economics, 1–​43. Nicosia, Cyprus: University of Cyprus Department of Economics, 2016. Michaelides, Marios, Ellen Poe-​ Yanagata, Jacob Benus, and Dharmendra Turancharetti. Impact of the Reemployment Eligibility Assessment (REA) Initiative in Nevada (January). Columbia, MD: IMPAQ International, 2012. https://​wdr.dol​eta.gov/​resea​rch/​Ful​lTex​t_​Do​cume​nts/​ETAOP_​2012_​ 08_​REA_​Neva​da_​F​ollo​w_​up​_​Rep​ort.pdf. Mueller, Andreas, Jesse Rothstein, and Till von Wachter. Unemployment Insurance and Disability Insurance in the Great Recession. Journal of Labor Economics 34, no. 1 pt 2 (2016): s445–​s475. Mulligan, Casey B. The Redistribution Recession: How Labor Market Distortions Contracted the Economy. New York: Oxford University Press, 2012. Nie, Jun, and Ethan Sturby. Would Active Labor Market Policies Help Combat High US Unemployment? Federal Reserve Bank of Kansas City Economic Review 3rd Q (2011): 1–​69. Organisation for Economic Cooperation and Development (OECD). Employment Protection Regulation. OECD Employment Outlook. Paris: OECD Publishing, 2004. Organisation for Economic Cooperation and Development (OECD). OECD Employment Outlook 2012. Paris: OECD Publishing, 2012. Organisation for Economic Cooperation and Development (OECD). Public Expenditure and Participant Stocks on LMP: Public Expenditure of LMP by Main Categories (% GDP), 2013. http://​stats.oecd.org/​. Organisation for Economic Cooperation and Development (OECD). OECD Indicators of Employment Protection, 2016a. http://​www.oecd.org/​emp​loym​ ent/​emp/​oecdindic​ator​sofe​mplo​ymen​tpro​tect​ion.htm. Organisation for Economic Cooperation and Development (OECD. Short-​Term Labour Market Statistics: Harmonized Unemployment Rates (HURs), 2016b. http://​stats.oecd.org/​.

33 3

Labor Policy in Aftermath of Great Recession  333 Petrongolo, Barbara. The Long-​ Term Effects of Job Search Requirements: Evidence from the UK JSA Reform. Journal of Public Economics 93, no. 11 (2009): 1234–​1253. Rothstein, Jesse. Unemployment Insurance and Job Search in the Great Recession. Brookings Papers on Economic Activity 43, no. 2 (2011): 143–​196. Sarfati, Hedva. Coping with the Unemployment Crisis in Europe. International Labour Review 152, no. 1 (2013): 145–​156. Schmieder, Johannes, Till von Wachter, and Stefan Bender. The Effects of Extended Unemployment Insurance Over the Business Cycle: Evidence from Regression Discontinuity Estimates over 20 years. Quarterly Journal of Economics 127, no. 2 (2012): 701–​752. Schmieder, Johannes, Till von Wachter, and Stefan Bender. The Effect of Unemployment Benefits and Nonemployment Durations on Wages. American Economic Review 106, no. 3 (2016): 739–​777. Stiglitz, Joseph. How to Restore Equitable and Sustainable Economic Growth in the United States. American Economic Review: Papers and Proceedings 106, no. 5 (2016): 43–​47. Summers, Lawrence. The Age of Secular Stagnation: What It Is and What to Do About It. Foreign Affairs March/​April (2016): 2–​9. Taylor, John. Can We Restart the Recovery All Over Again? American Economic Review: Papers and Proceedings 106, no. 5 (2016): 48–​51. US Census Bureau. U.S. Mover Rate Remains Stable at About 12 Percent Since 2008, Census Bureau Reports, 2015. http://​www.cen​sus.gov/​newsr​oom/​press-​ relea​ses/​2015/​cb15-​47.html. Wilson, Daniel J. Fiscal Spending Jobs Multipliers: Evidence from the 2009 American Recovery and Reinvestment Act. American Economic Journal: Economic Policy 4, no. 3 (2012): 251–​282.

FOR FURTHER READING Aicep Portugal Global. Portugal Labour System. http://​www.cho​osep​ortu​gal. com.pt/​files/​1170_​Portuga​lLab​ourS​yste​m_​4d​a32b​36a5​668.pdf, January 2011. Albrecht, James, and Susan Vroman. Dual Labor Markets, Efficiency Wages and Search. Journal of Labor Economics 10, no. 4 (1992): 438–​461. Anderson, Patricia M. Continuing Eligibility. In Unemployment Insurance in the United States: Analysis of Policy Issues, ed. Christopher J. O’Leary and Stephen A. Wandner. Kalamazoo, MI: W. E. Upjohn Institute, 1977, 125–​162. Barlevy, Gadi. Evaluating the Role of Labor Market Mismatch in Rising Unemployment. Economic Perspectives 35, 3rd Q (2011): 82–​96. Chief Evaluation Office, U.S. Department of Labor. Current Study: Short-​Time Compensation Program Demonstration and Evaluation (ETA), 2016. https://​ www.dol.gov/​asp/​eva​luat​ion/​cur​rent​stud​ies/​30.htm.

34

334  Work and the Social Safety Net Congdon, William J., and Wayne Vroman. Covering More Workers with Unemployment Insurance: Lessons from the Great Recession. Washington, DC: Urban Institute, Issue Brief, Chief Evaluation Office, US Department of Labor, March 2021. Dickens, William, Lorenz Goette, Erica Groshen, Steinar Holden, Julian Messina, Mark Schweitzer, Jarkko Turunen, and Melanie E. Ward. How Wages Change: Micro Evidence from the International Wage Flexibility Project. Journal of Economic Perspectives 21, no. 2 (2007): 195–​214. Eberts, Randall W., and Edward B. Montgomery. Employment Creation and Destruction: An Analytical Review. Federal Reserve Bank of Cleveland, Economic Review 30 no. 3 (3rd Q 1994): 14–​26. Federal Reserve Bank of Minneapolis. The Recession and Recovery in Perspective, 2016. www.min​neap​olis​fed.org/​publi​cati​ons/​spec​ial-​stud​ies/​rip/​recess​ion-​ in-​pers​pect​ive. Feldstein, Martin. American Economic Policy in the 1980s. Chicago: University of Chicago Press, 1993. Feldstein, Martin. Employment Policy of the Middle Reagan Years: What Didn’t Happened and Why It Didn’t Happen. NBER Working Paper No. 5917. Cambridge, MA: National Bureau of Economic Research (NBER), 1997. Fujita, Shigeru. Economic Effects of the Unemployment Insurance Benefit. Business Review, 4th Q (2010): 20–​27. Fujita, Shigeru. Effects of Extended Unemployment Insurance Benefits: Evidence from the Monthly CPS. Federal Reserve Bank of Philadelphia Working Paper No. 10–​35/​R (January). Philadelphia, PA: Federal Reserve Bank of Philadelphia, 2011. Grubb, David. Assessing the Impact of Recent Unemployment Insurance Extensions in the United States. Working Paper. Paris: OECD, May 25, 2011. International Monetary Fund (IMF). Housing Recoveries: Cluster Report on Denmark, Ireland, Kingdom of the Netherlands-​The Netherlands, and Spain. IMF Multi-​Country Report 15 (1). Washington, DC: International Monetary Fund (IMF), 2014. Messenger, Jon C., and Naj Ghosheh, eds. Work Sharing During the Great Recession: New Developments and Beyond. Cheltenham, UK: Edward Elgar, 2013. Montgomery, Edward. Patterns in Regional Labor Market Adjustment: The United States Versus Japan. In Social Protection Versus Economic Flexibility: Is There a Trade-​Off?, ed. Rebecca Blank. Chicago: University of Chicago Press, 1994, 95–​118. Reinhart, Carmen, and Kenneth Rogoff. This Time Is Different: Eight Centuries of Financial Folly. Princeton, NJ: Princeton University Press, 2009. Valletta, Rob, and Katherine Kuang. Extended Unemployment and UI Benefits. FRBSF Economic Letters 12 (2010): 1–​4.

IN DEX

For the benefit of digital users, indexed terms that span two pages (e.g., 52–​53) may, on occasion, appear on only one of those pages. Page numbers followed by b, f, and t indicate boxes, figures, and tables, respectively. Numbers followed by n. indicate footnotes. Accelerating Connections to Employment (ACE), 268–​69 action plans, 30–​31, 101 activation. see labor activation active labor market policies (ALMPs) complementary, 19–​22 after COVID-​19 pandemic, 34–​35 effectiveness, 31–​32, 292–​95 effectiveness and limitations, 248–​63, 250t European experience, 319 evaluation of, 34–​35 evidence base, 270–​73 during Great Recession, 11 during high unemployment, 9 intensive, 49 internal processes and resources, 217 optimal timing, 38–​39 with private actors, 106–​7 public spending on, 292–​93, 294 Rehn-​Meidner model, 285–​86 requirements for participation in, 29–​30, 35 resources for, 34–​35 social assistance programs, 91–​92 spending on, 31–​32, 318–​19, 325

for stronger and fairer labor markets, 26–​27, 29–​30, 31–​33, 34–​35 timing, 38–​39, 53–​54 training programs, 248–​63 (see also training) work-​first approach, 215 (see also work-​first approach) for youth, 250t, 259–​60 Adamecz-​Völygi, Anna, 12, 38–​60 added worker effects, 65–​66 Affordable Care Act (US), 4, 313 Agenda 2010, 206–​7 age requirements in public employment services, 225 retirement age, 187, 188f, 191–​92, 315 aging population, 188f, 188, 323 Aid to Dependent Children (US), 2–​3, 123–​26 Aid to Families with Dependent Children (AFDC) end, 129 key characteristics, 126b labor activation, 2–​3, 14, 123–​26 participation rates, 125, 130–​31, 131f, 316f Alabama, 64 ALMPs. see active labor market policies American Families Plan (US), 4

335

336 Index American Job Centers (AJCs), 244, 264–​65, 270 American Recovery and Reinvestment Act (ARRA, Obama stimulus law), 64, 76, 137–​38, 312–​13 American Rescue Plan (ARP), 4, 142, 242–​ 43, 266–​67 Americans with Disabilities Act (ADA), 187, 196–​97 Anglo-​Saxon traditions, 150, 204–​5 apprenticeships, 246–​47, 250t, 259–​60, 265–​66 Arizona, 64 Assistance in Material Need (Czech Republic), 100t, 108 austerity, 205, 310 Australia active labor market policies (ALMPs), 29–​30 apprenticeships, 265 COVID-​19 crisis, 25–​26 Disability Support Pension (DSP), 16, 150, 151–​53, 152f, 154, 155f, 173–​76, 177, 178t, 194f, 195–​96, 195t, 198, 199, 201t, 295, 296 economic indicators, 328t Employment Pathway Account, 299 employment-​to-​population ratio, 328t Great Recession, 308–​9, 309f gross domestic product (GDP), 328t labor activation, 21, 287–​88, 292–​93, 296 long-​term unemployment, 322f New Start Allowance, 198 pay-​for-​performance contracts, 298–​99 privatized employment services, 326–​27 spending on labor market programs, 328t Star Rating system, 298–​99 unemployment insurance (UI), 71 unemployment rates, 6–​7, 291, 308–​9, 309f, 322f, 328t Austria apprenticeships, 259–​60, 265 job coaching program, 53 private placement agencies, 211–​12 profiling tools, 31 public employment services (PES), 33 short-​time working schemes (STWs), 33–​34 unemployment insurance (UI), 28–​29, 68–​ 69, 314–​15 unemployment rates, 6–​7, 291 autonomy, earned, 216 Bane, Mary Jo, 124 Belgium short-​time working schemes (STWs), 33–​34 spending on labor market programs, 245–​46 unemployment benefits, 29, 69, 289–​90 benchlearning, 19 Benefits Agency (UK), 104, 217

Berlin Job-​Offensive project, 47 Besharov, Douglas, 1–​24, 131–​32 Biden, Joe, 4 Bipartisan Budget Act (US), 184, 186–​87 black box approach, 218, 221, 222–​23, 232 Blair, Tony, 204–​5 blended services, 214 block grants, 127–​29, 327 Brexit, 311–​12 broad labor market policies, 10–​11 Burkhauser, Richard V., 16–​17, 183–​203 business cycle, 291–​92 California, 185 Call, Douglas M., 1–​24 Cameron, David, 206 Canada, 28–​29 Card, David, 248–​49 Career Academies, 260–​61, 262, 266 career progression support, 20, 297–​98 CARES (Coronavirus Aid, Relief, and Economic Security) Act, 64–​65, 142, 321n.7 caseload reduction credits, 15 case management, 287 case managers, 97, 101 caseworker meetings, 12–​13, 42–​48, 49–​50, 57 cash assistance programs, 2–​5, 13–​16, 92, 165–​66 Central Europe, 288, 289–​90. See also specific countries Centres for Work and Income (CWI), 98 child care subsidies, 4 child poverty rates, 134–​35, 135f, 136 Child Tax Credit (CTC), 1, 2, 3–​5, 15–​16, 123, 134, 143–​44, 266–​67 choice, 209–​12 Civilian Conservation Corps, 243 Clinton, Bill, 2–​3, 14–​15, 125–​27, 128, 129, 204–​5 coaches and coaching, 53, 216 communication, dual, 213 company trainee places (virksomhedspraktik), 226 competitiveness, 76–​77 competitive tenders, 212–​14 complementarities, 19–​22, 286–​87, 289–​92, 315 Comprehensive Employment and Training Act (CETA), 243 conditional benefits, 287, 289–​90 Congressional Budget Office (CBO), 312–​13 Connecticut, 268–​69 conservative-​corporativist welfare, 98–​99 Consolidated Omnibus Budget Reconciliation Act of 1985 (COBRA), 62–​63

Index 337 contracts, 205, 227 pay-​for-​performance, 298–​99 subcontractors, 219 temporary, 323 Contract with America, 127 Coronavirus Aid, Relief, and Economic Security (CARES) Act, 64–​65, 142, 321n.7 corporate welfare states, 150 cost efficiency, 54t, 56–​57 counseling. see job search counseling COVID-​19 pandemic active labor market policies (ALMPs) after, 34–​35 employment support, 29 impact on labor markets, 9–​11, 25–​27, 242, 284–​85, 300, 310 job retention schemes, 10, 34 lessons learned, 230 Pandemic Emergency Assistance Fund (US), 142 Pandemic Emergency Unemployment Compensation (PEUC), 65 Pandemic Unemployment Assistance (PUA), 65 Pandemic Unemployment Compensation (PUC), 65 public employment services (PES), 214, 222, 229, 230 Temporary Assistance for Needy Families (TANF) during, 141–​43 unemployment insurance, 66–​67, 142n.14 unemployment rates, 73, 284–​85 unemployment spending, 34–​35 work-​sharing, 321 creaming, 18, 210 Csillag, Márton, 12, 38–​60 customer focus of public employment services (PES), 209–​ 12, 216–​17, 218 of social assistance programs, 113, 117 Czech Republic Assistance in Material Need program, 100t, 108 Existence Minimum program, 108 minimum income scheme, 102t, 108–​9, 114t short-​time working schemes (STWs), 33–​34 social assistance, 93, 95–​96, 97, 98–​99, 100t, 102t, 106–​7, 113 Social Card, 108–​9 Daly, Mary, 16–​17, 183–​203 deadweight effects, 33 Deal, Nathan, 128 decentralization of public employment services (PES), 18, 205, 226–​27, 232

of social assistance programs, 115 de-​corporatization, 233–​34 Deficit Reduction Act (DRA), 137 Delegated Flexibility Pilots, 216 delivery of services, 96–​98, 113 Delors White Paper, 204–​5 Democratic Party, 1, 4–​5 Denmark active labor market programs (ALMPs), 29–​ 30, 31, 34–​35, 41, 229 Act on Active Social Policy, 93 Arbejdsformidlingen (AF), 223 caseworker meetings, 45–​46, 50 COVID-​19 crisis, 229, 230 dual apprenticeships, 259–​60 economic indicators, 328t employment protection legislation (EPL), 291 Employment Reform of 2014, 41 employment-​to-​population ratio, 328t flexicurity, 225–​26, 323–​24, 328–​29 Great Recession, 206, 227–​29, 308–​9, 309f, 321–​22 gross domestic product (GDP), 328t Jobcenters, 224, 226–​27 job search programs, 45–​46, 50 Kontanthjælp, 100t labor activation, 18, 288, 294, 326 labor market parties, 223 long-​term unemployment, 321–​22, 322f minimum income program, 102t, 109, 114t More in Work Act, 96 profiling tools, 31 public employment services (PES), 17–​18, 41, 206, 223–​30, 231, 233–​34, 300–​1 sick allowances, 225 social assistance, 7–​8, 93, 97, 98–​99, 100t, 102t, 104–​5, 106–​7, 109, 112–​ 13, 114t Social Democratic-​led government, 223–​ 24, 227–​29 spending on labor market programs, 245f, 245–​46, 318–​19, 328t training programs, 18, 248–​49 unemployment benefits, 7–​8, 65, 71, 73, 225–​26, 289 unemployment rates, 6–​7, 206, 223, 228, 308–​9, 309f, 321–​22, 322f, 328t welfare-​to-​work type reforms, 29 work-​first approach, 109, 112–​13, 224, 230, 233 work-​sharing, 321 digitalization, 214. See also e-​services disability, 154, 185, 196–​97

338 Index disability insurance (DI), 16–​17, 313 “back to work” bonus payments, 167–​68 cash transfers, 165–​66 “credits only” payments, 165–​66 early intervention, 198–​99 extending activation benefits to recipients of, 295–​96 Gatekeeper protocol, 160–​61 Individual Reintegration Plans, 162 OECD, 16, 150–​82 OECD reforms, 17, 195–​201 participation rates, 17, 194f, 194–​95, 295, 315, 316f, 319–​20 private occupational disability insurance (PODI), 172, 173f private work disability insurance (PWDI), 172, 173f reform hurdles, 200–​1 temporary, 185 US, 16–​17, 183–​203 Workers Compensation (US), 185 Disability Insurance (US), 80 Disability Premium (UK), 167 Disability Support Pension (DSP), 16, 150, 173–​ 76, 195–​96, 198, 199 recipiency rates, 151–​53, 152f, 154, 155f, 174–​75, 176, 177, 178t, 194f, 195t, 201t, 295, 296 disempowered empowerment, 204–​5 Dislocated Worker programs, 250t, 264–​65 displacement effects, 33 District of Columbia, 321 Dole, Robert, 128 dual apprenticeships, 259–​60, 265 dual communication strategy, 213 dual labor markets, 323 early entry, 218 early labor activation cost efficiency, 54t, 56–​57 during COVID-​19 crisis, 27 effects of, 45, 52–​57, 54t, 198–​99 in EU unemployment insurance programs, 38–​60 recommendations, 56, 287 walls to: effects of, 48–​49 early retirement, 315 early skills training, 55–​56 earned autonomy, 216 Earned Income Tax Credit (EITC), 2–​3, 15, 133–​34, 297, 313 earnings supplement programs, 139–​40, 140f Eastern Europe, 99n.2, 289–​90. See also specific countries economic mobility, 325–​26

economic policy, 311–​24 Economic Stimulus Act (US), 312 education European programs, 103f, 114t, 116 secondary, 265–​66 vocational, 19, 55, 265–​66 (see also vocational training) vouchers for, 210–​11 Ellwood, David, 124 emancipation, enforced, 204–​5 Emergency Unemployment Compensation (EUC), 64, 75, 79, 313 employability, 287 Employee Insurance Agency (UVW), 98 employment criteria for, 29–​30 in-​work benefits, 30, 297–​98 job opportunities, 287 low, 80 one-​euro jobs, 210 searching for (see job search) strengthening importance of, 94–​96 sustainable jobs, 298 temporary jobs, 323 work activities, 129 work availability requirements, 29–​30 work eligibility, 129 Employment and Support Alliance (UK), 111–​12 Employment and Support Allowance (ESA), 168–​69 Employment Pathway Account (Australia), 299 employment policy, 223–​24 employment protection, 290–​91 employment protection legislation (EPL), 290–​ 91, 322–​24 employment rehabilitation, 27 employment retention and advancement programs, 250t, 297 Employment Service (UK), 103–​4, 217 employment services cash aid, 2–​5 e-​services, 299–​301 private providers of, 18, 299, 326–​27 public (see public employment services (PES)) re-​employment services (see re-​employment services) for stronger and fairer labor markets, 25–​37 workforce development targeting, 263–​65 employment-​to-​population ratios, 326 Employment Zones, 95, 103–​4 empowerment, 204–​5, 210–​11 Enhanced Transitional Jobs Demonstration (ETJD), 258–​59

Index 339 entrepreneurial spirit, 207 e-​services, 10, 21, 299–​301 Estonia, 29, 300 ethnic minorities, 324 Euro crisis, 284 Europe. See also specific countries active labor market policies (ALMPs), 271–​ 72, 319 competitiveness, 76–​77 COVID-​13 pandemic, 284–​85 disability recipiency rates, 195 employment protection legislation (EPL), 324 employment rate, 80 Great Recession, 308–​9, 321–​22 Great Recession aftermath, 107–​12 gross domestic product (GDP), 310 labor activation, 5–​22, 7b, 65, 91–​121, 102t, 288, 318 minimum income programs, 13–​16, 91, 92, 98–​99, 100t, 102–​3, 102t, 103f, 114t public employment services (PES), 17–​18, 39–​42, 204–​41 social assistance programs, 1–​2, 13–​14, 91–​ 121, 100t unemployment insurance (UI), 69, 70, 73, 100t unemployment rates, 284–​85 welfare reform, 204–​5 workforce development, 18–​19 work-​sharing, 321–​22 European Central Bank, 111, 291n.10 European Commission, 28–​29, 291n.10 European Employment Guidelines, 286–​87 European Network of Public Employment Services, 272 European Public Employment Services, 10 European Union (EU). See also specific countries Brexit, 311–​12 COVID-​19 pandemic, 310 crisis packages, 111 employment-​to-​population ratio, 6–​7 Great Recession, 72–​73, 321–​22 gross domestic product (GDP), 310 labor activation, 5–​6, 283 labor force participation, 5–​7 long-​term unemployment, 321–​22, 322f unemployment insurance (UI), 28–​29, 38–​ 60, 80, 290 unemployment rates, 39–​40, 72–​73, 284, 321–​22 Youth Guarantee, 40b Eurozone, 308–​10 evidence base, 270–​73

Existence Minimum (Czech Republic), 108 Extended Benefits (EB), 64, 75, 313–​14 FAIR project, 47n.21 Family Support Act (US), 14, 124 financial work incentives, 30, 94–​95, 100–​1, 103, 103f, 105, 106–​7, 113, 114t, 116 Finland gross domestic product (GDP), 308–​9 labor activation, 20, 288, 292–​94 public employment service (PES), 300 spending on labor market programs, 31, 245–​46 unemployment benefits, 28–​29, 68, 289–​90 Flanders, 51–​52, 289–​90 Flexible New Deal (FND), 95, 100t, 103–​4, 111–​ 12, 218–​19, 221, 222–​23 flexicurity, 206, 223, 225–​26, 323–​24, 328–​29 Florida, 64, 320 food stamps. see Supplemental Nutrition Assistance Program (SNAP) Ford, Harold, 127 Förderung beruflicher Weiterbildung (FbW), 210–​11 former USSR, 99n.2 Fortune 500, 247–​48 France COVID-​19 crisis, 25–​26, 294 employment protection legislation (EPL), 291, 322 labor activation, 288, 294 Monthly Personalized Support (SMP), 97 public employment services (PES), 41–​42, 44–​45, 97 Revenu Minimum d’Insertion (RMI), 94, 109 Revenu Solidarité Active (RSA, minimum income system), 94, 97, 98–​99, 100t, 102t, 106, 109, 114t short-​time work schemes (STWs), 33–​ 34, 284–​85 social assistance programs, 93, 95–​96, 98–​99, 100t, 102t, 106–​7, 113, 114t training programs, 248–​49, 250t, 265 unemployment benefits, 29, 44–​45, 69, 289–​90 Youth RSA, 109 Freud Review, 215, 218 furlough/​wage subsidy schemes, 284–​85 future challenges, 20–​21, 213–​14, 222–​23, 295–​301 Future Jobs Fund (FJF), 222 Garlow, Stephanie, 133 Gatekeeper protocol, 160–​61 geographic mobility, 29–​30, 326

340 Index Georgia, 64, 268–​69 Germany 1:70 project, 47 active labor market policies (ALMPs), 29–​30, 250t, 259–​60, 271 Agenda 2010, 206–​7 apprenticeships, 246–​47, 259–​60, 265–​66 Arbeitsgemeinschaft (ARGE), 98, 104 Arbeitslosengeld II (ALG-​II), 94, 95, 98–​99, 100t, 209 Berlin Job-​Offensive project, 47 Bundesagentur für Arbeit (BfA or BA), 95, 98, 104, 207–​9 caseworker meetings, 47 Confederation of Trade Unions (DGB), 212–​13 COVID-​19 pandemic, 214 disability benefit programs, 16, 150, 169–​73, 177, 195–​96, 198–​99 disability benefit recipients, 151–​53, 152f, 155f, 170, 171–​72, 177–​78, 178t, 194f, 195, 195t, 201t economic indicators, 328t employment protection legislation (EPL), 291, 323–​24 employment-​to-​population ratio, 328t FAIR project, 47n.21 Federal Ministry for Labor and Social Affairs (BMAS), 207–​9 Great Recession, 206, 213, 233, 308–​9, 309f, 321–​22 gross domestic product (GDP), 308–​9, 328t Hartz reforms, 7–​8, 65, 109–​10, 206–​7, 208–​ 9, 210–​11, 231, 232, 289–​90 housing prices, 308–​9 integration agreements (IAs), 52 Jobcenters, 208–​9 Kurzarbeit (short-​time work), 213 labor activation, 288, 321, 323–​24 labor force participation, 7–​8 long-​term unemployment, 321–​22, 322f Merkel government, 8 Mini-​and Midi-​Jobs, 103 minimum income system, 102–​3, 109–​ 10, 114t Old-​Age Pension (OAP), 169–​70 payroll taxes, 169–​70, 206–​7 private occupational disability insurance (PODI), 172, 173f private work disability insurance (PWDI), 172, 173f public employment services (PES), 17–​18, 33, 41, 98, 206–​14, 231–​34 quasi-​markets, 212–​14 Schroder government, 8

short-​time work schemes (STWs), 33–​ 34, 284–​85 social assistance, 7–​8, 93–​94, 98–​99, 102–​4, 102t, 106–​7, 109–​10, 113, 114t, 210 spending on labor market programs, 328t spending on social assistance, 325 Statutory Accident Insurance (SAI), 169 training programs, 18, 55, 248–​49 Unemployment Benefits-​II (UB-​II), 98 unemployment insurance (UI), 7–​8, 29, 51–​ 52, 69, 71, 73, 94, 100t, 102t, 114t, 206–​7, 233, 289–​90, 313–​14 unemployment rates, 206–​7, 291, 308–​9, 309f, 321–​22, 328t Work Disability Pension (WDP), 169–​71, 172 work-​sharing, 321 youth ALMPs, 250t, 259–​60 Gilbert, Neil, 6 Gingrich, Newt, 2–​3, 128–​29 globalization, 323 Government Outcomes Lab (GOLab), 272–​73 Great Britain. See also United Kingdom Additional Pension, 166, 167 disability benefit programs, 16, 150, 165–​69, 177, 195–​96, 197–​99 disability benefit recipients, 151–​53, 152f, 155f, 167–​68, 177, 178t, 194f, 195t, 201t Disability Premium, 167 Employment and Support Allowance (ESA), 168–​69 Great Recession, 308–​9, 309f Incapacity Benefit (IB), 167, 168 Income Support (IS), 167, 168 Invalidity Benefit (IVB), 165–​67 Personal Capability Assessments, 167–​68 Restart reforms, 166 Sickness Benefit, 165–​66 Statutory Sick Pay (SSP), 166 Supplementary Benefit, 165–​66, 167 Unemployment Benefit, 165–​66 unemployment rate, 308–​9, 309f Universal Credit, 169 Work Capability Assessment, 168–​69 Great Recession aftermath, 107–​12, 311–​24 disability caseloads during, 16 Keynesian interpretation of, 76 labor activation during, 8–​9, 11, 14, 20–​21, 65, 301–​2 labor policy lessons, 306–​34 long-​term unemployment, 284 public employment services (PES), 34–​35, 40, 213, 221–​22, 227–​29 TANF response, 15, 137–​41 unemployment insurance (UI), 72–​75

Index 341 unemployment rates, 72–​78, 284, 308–​9, 309f, 321–​22, 326 Greece, 284–​85, 288, 310 group information sessions, 51–​52 Grubb, David, 286n.5 Guaranteed Minimum Income (Portugal), 93 Harmonized Unemployment Rate, 72–​73 Hartz reforms, 7–​8, 65, 109–​10, 206–​7, 208–​ 9, 210–​11 Hartz IV, 209, 231, 232 Haskins, Ron, 133–​34n.7 Hawaii, 185 health insurance, 62–​63 Health Professions Opportunity Grants (HPOG), 267–​69 Heinrich, Carolyn J., 18–​19, 242–​82 HIS Global Insights, 312–​13 housing prices, 308–​9 human capital or human resource development approach, 91, 109, 112–​13, 245–​48 Hungary, 44 Idaho, 64, 320 Illinois, 320 immigrants, 225 Incapacity Benefit (IB), 167 incentives, 7b, 197–​98. See also financial work incentives incentives-​strengthening approach, 204–​5 Income Support (IS), 167 index of employment protection, 325 Indiana, 320 individual focus, 103f, 114t, 205 individual job counseling, 54t, See also job search counseling information sharing, 12, 51–​52, 54t information technology (IT), 217, 247–​48 innovation, 267–​70 Institute for the Study of Labor in Germany (IZA), 271–​72 integration agreements (IAs), 48, 52 intensive ALMP care, 49 interagency cooperation, 205 International Classification of Disability, Health, and Functioning (ICF), 154, 196–​97 International Monetary Fund (IMF), 111, 291n.10 International Network for Data on Impact and Government Outcomes (INDIGO), 272–​73 internships, 226, 247–​48 interventions delivery target (IDT), 215–​16 Intreo (Ireland), 293–​94

Invalidity Benefit (IVB), 165–​67 Investment in Sustainable Work for Young People Act (WIJ), 110–​11 in-​work benefits, 30, 297–​98 Ireland employment protection legislation (EPL), 291 EU debt crisis, 291n.10 furlough/​wage subsidy schemes, 284–​85 gross domestic product (GDP), 310 housing prices, 308–​9 Intreo, 293–​94 JobPath, 299 labor activation, 20, 29, 293–​94, 298 Pathways to Work, 293–​94 short-​time working schemes (STWs), 33–​34 unemployment benefits, 70–​71, 289 unemployment rates, 284, 293–​94 Italy COVID-​19 crisis, 25–​26 gross domestic product (GDP), 310 labor activation, 20, 288, 294 National Employment Agency (ANPAL), 294 short-​time work schemes (STWs), 33–​ 34, 284–​85 unemployment benefits, 73, 289–​90 Japan COVID-​19 crisis, 25–​26 labor activation, 292–​93, 323, 326 short-​time working schemes (STWs), 33–​34 spending on ALMPs, 31 unemployment benefits, 69, 289–​90, 326 unemployment post-​2008, 291 Jewish Vocational Services (Boston), 268–​69 job brokering, 33, 211–​12 Jobcenters (Denmark), 224, 226–​27 Jobcenters (Germany), 208–​9 JobCentre Plus (JCP), 95, 104, 214–​15, 216–​22, 233, 293 job coaching, 53, 216 job creation, 27 Job Entry Target (JET), 215–​16 job opportunities, 287 Job Opportunities and Basic Skills (JOBS) program, 124–​25, 127–​28 Job Outcome Target (JOT), 215–​16 JobPath (Ireland), 299 job plans, 224 job readiness assistance, 7b job retention schemes, 10, 34 job search monitoring, 70, 300–​1 on-​the-​job, 70 UI requirements for, 7b, 29–​30

342 Index job search assistance, 19, 71 caseworker meetings, 12, 42–​48 information sharing, 12, 51–​52, 54t low-​intensity, 70–​72, 79 mandatory maintenance, 78–​79 nudging, 12, 51–​52, 54t with technology, 45 job search counseling, 30–​31 effectiveness, 12, 42–​47, 49–​50 effectiveness and cost efficiency, 54t, 57 personalized, 300–​1 video-​based, 10 job search training, 57. See also training job seeker motivation, 28, 287 (Income-​Related) Jobseekers Allowance (JSA), 98–​99, 100t, 102–​3, 111–​12, 220, 220t, 221, 250t Job Seekers Regime (UK), 218–​19 Jobs Study (OECD), 286 job training, 7b, See also training Job Training Partnership Act (JTPA), 243–​44, 250t, 257–​59, 263 Jørgensen, Henning, 10, 17–​18, 204–​41 Kansas, 64 Keynesian economics, 76 Kick-​Start program (UK), 222 Klerman, Jacob, 10, 13, 61–​90 Kluve, Jochen, 248–​49 knowledge banks, 227 Kontanthjælp (Denmark), 100t Kurzarbeit (short-​time work), 213 labor activation active labor market policies (ALMPs) (see active labor market policies [ALMPs]) agenda for enhanced US policies, 22 benefit conditionality, 289–​90 broad labor market policies, 10–​11 challenges for the future, 20–​21 combined with unemployment benefits, 326 competitive tenders, 212–​14 complementarities, 286–​87 cost efficiency, 54t, 56–​57, 327 after COVID, 9–​11 deadweight effects, 33 definition of, 1–​2, 91–​92, 287 delivery of services, 96–​98, 113 displacement effects, 33 early interventions, 12–​13, 38–​60, 287 effectiveness, 207, 226–​27 efficiency, 207, 226–​27 in Europe, 5–​22, 7b, 91–​121, 102t, 204–​41 extending to recipients of other welfare benefits, 295–​96

after Great Recession, 21–​22 in high unemployment, 8–​9 human capital or human resource development approach, 91, 109, 112–​ 13, 245–​48 integration with social benefits, 20, 103f, 104, 114t, 116 key factors for effective strategies, 287 key goals, 27 late, 52–​56 lessons from the Great Recession aftermath, 306–​34 lock-​in effects, 32 optimal timing, 38–​39 origins of, 5–​8 passive labor market policies (PLMPs), 19–​20, 22 privatization of, 102t, 231–​32 in public employment services (PES), 17–​ 18, 204–​41 for recently unemployed persons, 12 reflections for the future, 283–​305 reform waves, 14, 93, 98–​113, 103f Rehn-​Meidner model, 285–​86 roots, 285–​88 selection effects, 33 service delivery, 96–​98, 113 short-​time working schemes (STWs), 33–​34 in social assistance programs, 91–​121, 102t, 122–​49 spending on, 22, 31–​32, 318–​19, 325 strong approaches, 6–​7 for stronger and fairer labor markets, 25–​37 substitution effects, 33 timing of interventions, 38–​39, 52–​56 tradeoffs/​complementarities, 289–​92 US experiences, 2–​5, 122–​49 work-​first approach, 18 (see also work-​first approach) workforce development–​related, 17–​18 labor force men’s participation, 6–​8 women’s participation, 6–​7, 131, 132f, 133, 188f, 189, 324 labor market parties, 223 labor markets COVID-​19 crisis, 25–​26 dualism, 290–​91 dynamic, 325 quasi-​markets, 212–​14, 299 late labor activation, 52–​56 legislation, employment protection (EPL), 290–​91 Lerman, Robert, 265

Index 343 Local Support Commissions (Comissões Locais de Acompanhamento, CLAs), 105–​6 Locations for Work and Income (LWI), 98 lock-​in effects, 32, 55–​56 Lødemel, Ivar, 13–​14, 91–​121 long-​term disability programs, 164–​65. See also disability insurance (DI) long-​term unemployment, 321–​22, 322f low-​skilled youth, 26–​27 Luxembourg, 322 Madison Strategies Group, 268–​69 maintenance of effort (MOE), 128 make work, 67–​68 management by objectives (MBO), 205 managerial style, 287 Mandatory Work Activity (UK), 111–​12 Manpower Development and Training Act (MDTA), 243 marginal effective tax rates (METRs), 297 marketization, 205, 226–​27 Martin, John, 9, 10, 20–​21, 283–​305 Maryland, 268–​69 Massachusetts, 64 McVicar, Duncan, 16, 150–​82 Medicaid, 124, 128–​29, 313 men’s labor force participation, 6–​8 Mexico, 25–​26, 318–​19 Michigan, 64 Middle-​Class Relief and Job Creation Act (US), 321 Mini-​and Midi-​Jobs (Germany), 103 minimum income programs, 13–​16 European, 92, 98–​99, 100t, 101–​3, 102t, 103f, 114t models of administration, 102–​3, 102t, 103f, 114t Minnesota, 320 Missouri, 64, 74–​75 Moffitt, Robert, 133 Montana, 64 Montgomery, Edward, 9, 21–​22, 306–​34 Monthly Personalized Support (SMP), 97 Moody’s, 312–​13 moral hazards, 66, 67, 327 More in Work Act (Denmark), 96 Moreira, Amílcar, 13–​14, 91–​121 Moynihan, Daniel, 130 National Employment Agency (ANPAL), 294 National Job Corps (US), 250t, 261–​62, 263 National Youth Administration (US), 243 NEETs (not in employment, education, or training), 26, 40b negative financial incentives, 100–​1, 116

neo-​classical economics, 76–​77 Netherlands Act on Public Employment Services (PES), 97 ALMP evidence base, 271 caseworker meetings, 44, 52 Centres for Work and Income (CWI), 98 disability benefit programs, 7–​8, 16, 150, 154–​62, 177, 195–​96, 197–​99, 201, 296, 319–​20 disability benefit recipients, 151–​53, 152f, 155f, 160–​61, 162, 177, 178t, 194f, 195t, 201t, 295–​96 Employee Insurance Agency (UVW), 98 Investment in Sustainable Work for Young People Act (WIJ), 110–​11 labor activation, 29–​30, 65, 92, 93, 102t, 106–​7, 110–​11, 112–​13, 288, 298, 299, 300 Locations for Work and Income (LWI), 98 long-​term disability programs, 201 short-​time working schemes (STWs), 33–​34 social assistance, 7–​8, 29, 95–​96, 98–​99, 100t, 102t, 104–​5, 112–​13, 114t spending on labor market programs, 245–​46 Structure Implementation Work and Income Act (SUWI), 97 unemployment benefits, 29, 52, 71, 73, 289–​ 90, 300, 319–​20 Wet arbeidsongeschiktheidsverzekering zelfstandigen (WAZ), 154–​60 Wet arbeidsongeschiktheidsvoorziening jonggehandicapten (Wajong), 154–​60 Wet op de arbeidsong eschiktheidsverzekering (WAO)/​Wet werk en inkomen naar arbeidsvermogen (WIA), 154–​61 Work and Social Assistance Act (WWB), 95–​ 96, 98–​99, 100t, 110–​11 Nevada, 29, 292, 320 New Deal (US), 123–​24 New Deal for Long-​Term Unemployed (ND25+​), 65, 93 New Deal for Young People (NDYP), 93, 95, 319–​20 New Federalism, 243 New Jersey, 185, 313–​14 New Labour party, 111–​12 New Public Management (NPM), 17–​18, 93, 96–​97, 205, 231 New Start Allowance (Australia), 198 New York, 29, 185, 320 New York Times, 130 New Zealand, 6–​7, 29–​30, 31, 250t, 265

344 Index Nordic countries. See also specific countries active labor market policies (ALMPs), 34–​ 35, 271 disability benefits recipients, 295 labor activation, 288, 291–​92, 294, 326 public employment services (PES), 33 spending on labor market programs, 318–​19 spending on social assistance, 325 unemployment benefits, 289–​90, 326 North Carolina, 64 North Dakota, 320 Northern Ireland, 43–​44, 71, 165 Norway disability benefits recipients, 296 labor activation, 93, 102t, 104, 106–​7, 110, 113, 288, 292–​93, 294, 296, 326 Qualification Programme (QP), 97, 100t, 104, 110 social assistance, 97, 98–​99, 100t, 102t, 114t spending on labor market programs, 318–​19 unemployment insurance (UI), 68, 69, 71 unemployment rates, 6–​7 Welfare and Employment Agency (NAV), 104 not employed but actively searching for work, 67 not in employment, education, or training (NEETs), 26, 40b nudging, 12, 51–​52, 54t Nunn, Alexander, 10, 17–​18, 204–​41 Obama, Barack, 76–​77 Obama stimulus law. see American Recovery and Reinvestment Act (ARRA) occupational training, 265. See also training Old-​Age and Survivors (OASI) Trust Fund, 184, 186 Old-​Age Pension (OAP), 169–​70 older workers, 225, 287, 324 1:70 project, 47 one-​euro jobs, 210 one-​stop shops, 205 on-​the-​job job search, 70 on-​the-​job training (OJT), 55, 242–​43, 250t, 258–​59, 260, 263–​64, 265 Organization for Economic Cooperation and Development (OECD). See also specific countries Benefits and Wages, 289 COVID-​19 crisis, 25–​27 disability benefits, 16, 17, 150–​82, 183–​203 disability benefits recipiency rates, 151–​53, 152f, 178t, 183–​84, 194–​95, 194f, 195t economic indicators, 328t

employment protection legislation (EPL), 290–​91, 322 Great Recession, 284 Harmonized Unemployment Rate, 72–​73 index of employment protection, 325 Jobs Study, 286 key factors for effective activation strategies, 287 labor activation, 6–​8, 25–​27, 28–​30, 33–​34, 283, 306, 325 lessons for US disability policy, 17, 195–​201 public employment services (PES), 26–​27, 33 short-​time working schemes (STWs), 33–​34 spending on labor market programs, 245–​46, 245f, 318–​19, 325 unemployment benefits, 28–​30, 290 unemployment rates, 6–​7, 25, 284, 308–​9 outsourcing, 7b Oxford University, 272–​73 Pandemic Emergency Assistance Fund (US), 142 Pandemic Emergency Unemployment Compensation (PEUC), 65 Pandemic Unemployment Assistance (PUA), 65 Pandemic Unemployment Compensation (PUC), 65 parental leave, 4 parking, 18 partner organizations (Bildungsträger), 210–​11 passive labor market policies (PLMPs), 19–​20, 22, 204–​5 Pathways to Work (Ireland), 293–​94 Pathways to Work (UK), 155f, 167–​68, 218 Pavetti, LaDonna, 125 Pay for Performance strategies, 272–​73 Pay for Success (PFS) efforts, 272–​73 payroll taxes, 169–​70, 198 Pelosi, Nancy, 130 performance management, 226–​27, 231, 243–​44 performance management framework (PMF), 216–​17 Perkins, Frances, 243 Per Scholas (the Bronx, New York City), 268–​69 personal advisors, 44–​45, 101 Personal Capability Assessments, 167–​68 personalization, 113 personalized counseling, 300–​1. See also job search counseling Personal Responsibility and Work Opportunity Reconciliation Act (US), 129, 130–​36 Poland, 69 political conflict, 3–​5 political economy, 296, 306

Index 345 Poor Laws (UK), 113 population, aging, 188, 188f, 323 Portugal Comissões Locais de Acompanhamento (CLAs, Local Support Commissions), 105–​6 employment protection legislation (EPL), 322 Great Recession, 321–​22 Guaranteed Minimum Income, 93 labor activation, 288, 292, 325 long-​term unemployment, 321–​22, 322f Rendimento Social de Inserção (RSI, Social Insertion Income), 98–​99, 100t, 105, 111 RMG (Guaranteed Minimum Income), 111 social assistance, 98–​99, 100t, 102t, 105–​6, 111, 113, 114t Social Security, 105–​6 unemployment insurance (UI), 28–​29, 69 unemployment rate, 284, 292, 321–​22, 322f positive financial incentives, 100–​1, 103, 106–​ 7, 116 post-​Communist Europe, 99n.2, 100t, See also specific countries poverty, 134–​35, 135f, 136 private occupational disability insurance (PODI), 172, 173f private providers employment services, 18, 205, 218–​21, 231–​ 32, 299, 326–​27 placement agencies, 211–​12 re-​employment services, 21, 298–​99 social assistance programs, 102t, 103–​4, 103f, 106–​7, 113, 114t, 117 Star Rating system, 298–​99 workforce development services for youth and young adults, 265–​67 private work disability insurance (PWDI), 172 profiling, 31, 79, 209–​10, 224 Project Quest, 269–​70 Project Rise, 247–​48 public employment services (PES), 11, 17–​ 19, 32–​33 Act on Public Employment Services (PES), 97 Bundesagentur für Arbeit (BfA or BA), 95, 98, 104, 207–​9 career progression, 20 centre-​right motivations for reform, 223–​24 choice, 209–​12 creaming, 210 customer orientation, 209–​12 decentralization of, 18, 205, 226–​27, 232 delivery of, 218–​21 early activation, 39–​42

e-​services, 299–​301 in Europe, 39–​42, 204–​41 future challenges, 213–​14, 222–​23 governance and steering, 215–​17 governance structures, 207–​9 governance under social democratic rule, 223–​24 after Great Recession, 34–​35, 40 information and nudging, 51 job brokeriing, 33 key factors for effective activation, 287–​88 labor activation, 17–​18, 204–​41 marketization, 226–​27 New Public Management (NPM), 231 performance management, 226–​27, 231 privatization of, 205, 206–​7, 218–​21, 231–​ 32, 298 profiling, 209–​10 reflections for the future, 293 reforms, 205–​6 requirements, 29 spending on, 26–​27 temporary jobs, 211–​12 timing effects, 52–​53 training programs, 258–​59 vouchers, 209–​12 work-​first approach, 216–​17, 233 Youth Guarantee, 40b public opinion, 296 public sector support, 245–​48, 267–​73 public service employment (PSE), 243 Public Works Administration (US), 243 purchasing centers, 212 Qualification Programme (QP), 97, 100t, 104, 110 Quantitative Easing, 311–​12 quasiautonomous nongovernmental organizations (QUANGOs), 217 quasi-​markets, 212–​14, 299 Reagan, Ronald, 124 Reagan era, 243 re-​employment services during COVID-​19, 27 lessons, 320 private providers, 21, 298–​99 Re-​employment and Eligibility Assessment (REA), 63, 71–​72, 78–​79 Re-​employment Services and Eligibility Assessment (RESEA) Program, 63, 78–​79 reflections for the future, 292, 298 refugees, 225 Rehn, Gosta, 285–​86 Rehn-​Meidner model, 285–​86

346 Index Rendimento Social de Inserção (RSI, Social Insertion Income), 98–​99, 100t, 105, 111 Republican Party, 4, 127 resource profile (ressourceprofilen), 224 retirement age, 187, 188f, 191–​92, 315 Revenu Minimum d’Insertion (RMI), 94, 109 Revenu Solidarité Active (RSA), 94, 97, 98–​99, 100t, 102t, 106, 109, 114t Rhode island, 185 RMG (Guaranteed Minimum Income), 111 Romania, 28–​29 Roma population, 108–​9 Roosevelt, Franklin D., 123–​24, 243 sanctions in public employment services (PES), 215, 225 in social assistance programs, 99, 103f, 113, 114t, 115–​16 Sanders, Bernie, 1–​2 Scandinavian countries, 70–​71, 150, 265. See also specific countries Scarpetta, Stefano, 6, 10–​11, 25–​37 secondary schooling, 265–​66 sectoral training programs, 267–​70 SEIS, 269–​70 selection effects, 33 self-​service, 218 Service Delivery Areas, 244 Services and Sanctions programs, 29 short-​term reintegration offers, 210 short-​term skills training, 55 short-​time work (Kurzarbeit), 213 short-​time working schemes (STWs), 33–​ 34, 284–​85 sickness benefits. See also disability insurance (DI) allowances, 225 extending activation to recipients of, 295–​96 Invalidity Benefit (IVB), 165–​67 in OECD countries, 163, 164–​67 skills training, 55–​56, 67–​68. See also training Slovakia, 28–​29, 33–​34 Slovenia, 28–​29, 33, 69 social assistance programs, 13–​16 aims and objectives, 115 conservative-​corporativist type, 98–​99 delivery of services, 96–​98, 113 efficiency costs, 327 eligibility for, 13 European, 1–​2, 91–​121, 100t, 102t, 114t extending activation to recipients of, 295–​96 individual focus, 103f, 114t integration with labor activation services, 20, 103f, 104, 114t, 116

models of benefit administration, 102t, 103f, 115 moral hazards, 327 post-​NPM, 96–​97, 113 service delivery, 96–​98, 113 with single contact points, 101 spending on, 325 state-​centered models, 101 synchronization, 7b US programs, 5, 14–​16, 122–​49 Social Card (Czech Republic), 108–​9 Social Democratic steering, 223–​24, 227–​29 Social Impact Bonds (SIBs), 272–​73 Social Security (Portugal), 105–​6 Social Security (US), 17 Social Security Administration (US), 184, 186, 197, 199 Social Security Disability Amendments (US), 190 Social Security Disability Insurance (SSDI), 17, 183–​85, 186–​87, 191, 197, 198, 200–​1 Administrative Law Judge (ALJ), 186 caseloads, 184, 187–​93, 201t Disability Determination Stage (DDS), 186 growth, 185, 187, 188–​93, 188f, 192f growth projections, 192–​93, 193f prevalence rates, 192–​93, 193f recipients, 187, 189–​90, 190f, 191, 199 South Carolina, 64 Southern Europe, 33, 100t, 288, 289–​90, 291. See also specific countries South Korea, 25–​26 Spain COVID-​19 crisis, 25–​26 employment protection legislation (EPL), 322–​23 Great Recession, 308–​9, 309f, 321–​22 gross domestic product (GDP), 310 labor activation, 20, 288, 294–​95, 323 long-​term unemployment, 321–​22, 322f unemployment insurance (UI), 69, 70–​71 unemployment rate, 284, 308–​10, 309f SSDI Trust Fund, 184, 186–​87 St. Nicks Alliance, 268–​69 staff-​client relations, 205 Star Rating system (Australia), 298–​99 state-​centered models, 101 State Workforce Investment Boards (US), 244 Statutory Accident Insurance (SAI), 169 Statutory Sick Pay (SSP), 166 Structure Implementation Work and Income Act (SUWI), 97 subcontractors, 219 Subsidized and Transitional Employment Demonstration (STED), 258–​59

Index 347 substitution effects, 33 suitable job offers, 290 Supplemental Nutrition Assistance Program (SNAP, food stamps), 15, 80, 113, 122–​23, 128–​29, 139–​40, 143 during Great Recession, 138, 313 participation rate, 141, 315–​18, 316f Supplemental Poverty Measure (SPM), 135, 136 Supplemental Security Income (SSI), 80, 129, 184 Supplementary Benefit (UK), 165–​66, 167 sustainable jobs, 298 Sweden active labor market programs (ALMPs), 29–​ 30, 31, 53–​55 disability benefit programs, 16, 150, 163–​65, 177, 195–​96, 197–​99 disability benefit recipients, 151–​53, 152f, 155f, 164, 165, 177, 178t, 194f, 195t, 201t labor activation, 285–​86, 288, 294 long-​term unemployment, 322f profiling tools, 31 sickness benefits, 163, 164–​65 spending on labor market programs, 31, 318–​19 unemployment insurance (UI), 28–​29, 46, 48, 50, 68, 71 Switzerland active labor market policies (ALMPs), 29–​ 30, 34–​35 apprenticeships, 259–​60, 265 disability benefits recipients, 296 labor activation, 288, 292–​93, 296, 326 public employment services (PES), 300–​1 unemployment benefits, 71, 289–​90 unemployment rates, 6–​7, 291 TANF. See Temporary Assistance for Needy Families TANF Emergency Contingency Fund, 15, 137–​38 tax credits, 30, 111–​12 Child Tax Credit (CTC), 1, 2, 3–​5, 15–​16, 123, 134, 143–​44, 266–​67 Earned Income Tax Credit (EITC), 2–​3, 15, 133–​34, 297, 313 Working Families’ Tax Credit (UK), 95 Working Tax Credit (UK), 95, 103 taxes, payroll, 169–​70, 198 Taxpayer Relief Act (1997), 3 tax rates, marginal effective (METRs), 297 technology, 45, 217, 247–​48. See also e-​services teleworking, 25–​26

Temporary Assistance for Needy Families (TANF), 2–​3, 5, 14–​16, 122–​23, 127–​30, 135–​39, 140–​41, 142–​43, 144, 244 during COVID-​19 pandemic, 141–​43 Emergency Contingency Fund (ECF), 15, 137–​38 during and after the Great Recession, 137–​41 maintenance of effort (MOE) funding, 128, 140–​41 participation rates, 15, 130–​33, 131f, 138, 141, 142–​43, 315–​18, 316f TANF Emergency Contingency Fund, 15, 137–​38 work participation rates, 139–​40, 139f, 140f temporary jobs, 211–​12, 323 temporary layoffs, 284n.3 Texas, 265–​66, 268–​69 third way policy, 206 Ticket to Work, 199 time-​limited benefits, 205 timing of interventions, 38–​39, 52–​56 Title I, 265–​66 token payments, 139–​40 Towards Employment, 268–​69 trade adjustment assistance and dislocated worker programs, 250t tradeoffs, 289–​92 train-​first approach, 291–​92 training, 7b, 27, 30–​31 company trainee places (virksomhedspraktik), 226 early, 55–​56 job search, 57 occupational, 265 on-​the-​job (OJT), 55, 242–​43, 250t, 258–​59, 260, 263–​64, 265 privatization of, 18 skills, 67–​68 in social assistance programs, 103f, 114t, 116 in unemployment insurance (UI) programs, 54t, 55, 71 vocational, 19, 55, 191, 210–​11, 257, 265–​66 workforce development services, 248–​ 65, 250t Transitional Medical Assistance (US), 124 Troubled Asset Relief Fund (US), 311–​12 Trump, Donald, 311–​12 Tucker, Walter, 128 Turkey, 310 unemployment, 61 American U-​3, 73 American U-​6, 73 during COVID-​19 crisis, 25–​27, 73 during Great Recession, 72–​78

348 Index unemployment (cont.) Harmonized Unemployment Rate, 72–​73 high, 8–​9, 292 long-​term, 321–​22, 322f mothers “needed at home or not actively seeking work,” 125 not employed but actively searching for work, 67 not in employment, education, or training (NEETs), 26, 40b temporary layoffs, 284n.3 unemployment insurance (UI) and, 77–​78 unemployment benefits, 11–​13. See also unemployment insurance (UI) Unemployment Benefits-​II (UB-​II), 98 unemployment insurance (UI), 11–​13 administration of services, 7b Arbeitslosengeld II (ALG-​II), 94, 95, 98–​99, 100t, 209 arguments for, 65–​66 basic, 70 benefit limits, 7b benefit step-​downs, 7b case monitoring, 7b combined with activation schemes, 326 conditional benefits, 289–​90 during COVID-​19 crisis, 66–​67, 142n.14 definition of, 61 duration of benefits, 28, 69, 70–​71, 73–​74, 75, 79, 315 early activation, 38–​60 earnings disregards and incentives, 7b effects of, 68–​72 eligibility criteria, 7b, 11–​12, 28–​29, 290 Emergency Unemployment Compensation (EUC), 64, 75, 79, 313 e-​services, 300 European programs, 7b, 38–​60, 102t, 114t exit rates, 69 expansions or extensions of benefits, 70, 313–​15 Extended Benefits (EB), 64, 75, 313–​14 extending benefits to recipients of other welfare benefits, 295–​96 during Great Recession, 61–​90, 138, 313, 315–​17 job readiness assistance, 7b job search requirements, 7b, 29, 72, 319–​20 maximum duration, 69 Pandemic Emergency Unemployment Compensation (PEUC), 65 participation rates, 15, 317–​18 policy-​induced changes, 76 potential changes, 78–​80 role in ongoing low employment, 80

role in unemployment, 77–​78 rules and requirements, 29–​30 spending on, 15, 34–​35, 294 strategies to lower durations, 78–​79 temporary extensions, 314–​15 theory, 65–​68 and unemployment, 72–​75, 77–​78 in US, 13, 62–​65, 72–​75 United Kingdom. See also Great Britain active labor market policies (ALMPs), 29–​ 30, 214–​15 apprenticeships, 265 Benefits Agency, 104, 217 Brexit, 311–​12 Cameron government, 8 COVID-​19 crisis, 25–​26, 293 Delegated Flexibility Pilots, 216 disability benefits recipients, 295–​96, 319–​20 economic indicators, 328t Employment and Support Alliance, 111–​12 employment protection legislation (EPL), 291 Employment Retention and Advancement (ERA) program, 250t, 297 Employment Service, 103–​4, 217 employment-​to-​population ratio, 328t Employment Zones, 95 Flexible New Deal (FND), 95, 100t, 103–​4, 111–​12, 218–​19, 221, 222–​23 furlough/​wage subsidy schemes, 284–​85 Future Jobs Fund (FJF), 222 Great Recession, 206, 221–​22, 321–​22 gross domestic product (GDP), 293n.11, 328t housing prices, 308–​9 JobCentre Plus (JCP), 95, 104, 214–​15, 216–​ 22, 233, 293 Job Entry Target (JET), 215–​16 Job Outcome Target (JOT), 215–​16 (Income-​Related) Jobseekers Allowance (JSA), 98–​99, 100t, 102–​3, 111–​12, 220, 220t, 221 Job Seekers Regime, 218–​19 job training, 18 Kick-​Start program, 222 labor activation, 29, 288, 292–​93, 296, 298, 321, 326 labor force participation, 7–​8 long-​term unemployment, 321–​22, 322f Mandatory Work Activity, 111–​12 minimum income program, 102, 102t, 114t New Deal for Long-​Term Unemployed (ND25+​), 65, 93 New Deal for Young People (NDYP), 93, 95, 319–​20 One DWP, 217

Index 349 Pathways to Work, 155f, 167–​68, 218 Pensions Service, 217 Poor Laws, 113 public employment services (PES), 17–​18, 206, 214–​23, 231–​33, 293 Sector Skills Development Agency, 268 social assistance programs, 8, 92, 93, 98–​ 99, 100t, 102–​4, 102t, 106–​7, 111–​12, 114t, 325 spending on labor market programs, 245–​46, 245f, 318–​19, 328t spending on social assistance, 325 unemployment assistance, 29, 52, 65, 69, 71, 73, 100t, 114t, 289–​90, 319–​20, 326 unemployment rates, 6–​7, 206, 222, 291, 293, 321–​22, 322f, 328t Universal Credit, 169, 297–​98 Work and Health Programme, 221 Working Families’ Tax Credit, 95 Working Tax Credit, 95, 103 Work Programme, 65, 111–​12, 218–​21, 220t, 222–​23, 299 Youth Contract (YC), 222 United States Affordable Care Act, 313 aging of the population, 188, 188f Aid to Dependent Children, 2–​3, 123–​26 Aid to Families with Dependent Children (AFDC), 2–​3, 14, 123–​26, 129 American Families Plan, 4 American Job Centers (AJCs), 244, 264–​ 65, 270 American Recovery and Reinvestment Act (ARRA, Obama stimulus law), 64, 76, 137–​38, 312–​13 American Rescue Plan (ARP), 4, 142, 242–​ 43, 266–​67 Americans with Disabilities Act (ADA), 187, 196–​97 Biden administration, 4, 143 Bipartisan Budget Act, 184, 186–​87 cash assistance programs, 2–​5, 13–​16, 92 child poverty rates, 134–​35, 135f, 136 Child Tax Credit (CTC), 1, 2, 3–​5, 15–​16, 123, 134, 143–​44, 266–​67 Civilian Conservation Corps, 243 Clinton administration, 124, 127, 130 Comprehensive Employment and Training Act (CETA), 243 Consolidated Omnibus Budget Reconciliation Act of 1985 (COBRA), 62–​63 Coronavirus Aid, Relief, and Economic Security (CARES) Act, 64–​65, 142, 321n.7

COVID-​19 pandemic, 284–​85, 310 debt-​to-​GDP ratio, 310–​11 Deficit Reduction Act (DRA), 136–​37 Department of Education, 244 Department of Health and Human Services (DHHS), 267–​68, 272 Department of Labor (USDOL), 243, 267–​ 68, 272, 321 disability benefits programs, 16–​17, 185–​87, 193–​95, 200–​1, 313 disability benefits recipients, 17, 152f, 184, 187–​93, 188f, 194–​95, 194f, 195t, 199, 201t, 295, 315 Disability Insurance, 80 Earned Income Tax Credit (EITC), 2–​3, 15, 133–​34, 297, 313 economic indicators, 328t economic mobility, 325–​26 Economic Stimulus Act, 312 Emergency Unemployment Compensation (EUC), 64, 75, 79, 313 employment policy, 306 employment protection, 325 employment protection legislation (EPL), 322 employment rate, 80 employment-​to-​population ratio, 306–​8, 307f, 328t Family Support Act, 14, 124 federal deficit, 310–​11 fiscal policy, 312 Great Recession, 8n.4, 15, 64, 72–​75, 137–​41, 284, 306–​9, 309f, 321–​22 gross domestic product (GDP), 306–​8, 310, 328t health insurance, 62–​63 Health Profession Opportunity Grants (HPOG), 267–​68 housing prices, 308–​9 Job Opportunities and Basic Skills (JOBS) program, 124–​25, 127–​28 Job Training Partnership Act (JTPA), 243–​ 44, 250t labor activation, 1–​24, 112–​13, 288, 297, 320–​21 labor force participation, 131, 132f, 133 labor markets, 323, 325–​26 labor underutilization rate, 306–​8, 308f long-​term unemployed, 306–​8, 308f long-​term unemployment, 321–​22, 322f Manpower Development and Training Act (MDTA), 243 Medicaid, 124, 313 Middle-​Class Relief and Job Creation Act, 321

350 Index United States (cont.) Middle Class Tax Relief and Job Creation Act (February 2012), 79 National Job Corps, 250t, 261–​62, 263 National Youth Administration, 243 New Deal, 123–​24 Pandemic Emergency Assistance Fund, 142 Pandemic Emergency Unemployment Compensation (PEUC), 65 Pandemic Unemployment Assistance (PUA), 65 Pandemic Unemployment Compensation (PUC), 65 Personal Responsibility and Work Opportunity Reconciliation Act, 129, 130–​36 political conflict, 3–​5 political economy, 306 public service employment (PSE), 243 Public Works Administration, 243 Reagan administration, 124, 243–​44 Re-​employment and Eligibility Assessment (REA), 63, 71–​72, 78–​79 re-​employment services, 320 Re-​employment Services and Eligibility Assessment (RESEA) Program, 63, 78–​79 Service Delivery Areas, 244 social assistance programs, 14–​16, 92, 112–​ 13, 122–​49, 325 Social Security, 17 Social Security Administration, 184 Social Security Disability Amendments, 190 Social Security Disability Insurance (SSDI), 17, 183–​89, 188f, 197, 198, 199 spending on labor market programs, 31, 245–​46, 245f, 318–​19, 325, 328t State Workforce Investment Boards, 244 Supplemental Nutrition Assistance Program (SNAP, food stamps), 15, 80, 113, 122–​23, 128–​29, 138, 139–​40, 141, 143, 313, 315–​ 18, 316f Supplemental Security Income (SSI), 80, 129, 184 TANF Emergency Contingency Fund, 15, 137–​38 Taxpayer Relief Act (1997), 3 Temporary Assistance for Needy Families (TANF), 2–​3, 5, 14–​16, 122–​23, 127–​33, 131f, 135–​43, 139f, 140f, 144, 244, 315–​ 18, 316f Transitional Medical Assistance program, 124 Troubled Asset Relief Fund, 311–​12 unemployment benefits, 13, 15, 28, 29, 62–​65, 68–​69, 70–​71, 72–​75, 76–​78, 80, 142n.14, 313, 315–​18

unemployment rates, 72–​75, 284–​85, 306–​9, 309f, 321–​22, 328t wages, 21 Wagner-​Peyser Employment Services, 244 welfare reform, 2–​3, 5, 14–​15, 204–​5, 288 welfare-​to-​work type reforms, 29 worker protections, 21 Workers Compensation, 185 workforce development services, 18–​ 19, 242–​82 Workforce Innovation and Opportunity Act (WIOA), 244, 263–​66, 270, 272–​73 Workforce Innovation Fund (WIF), 267–​68 Workforce Investment Act (WIA), 244–​45, 250t, 262, 263 Workforce Investment Areas, 244 work-​sharing, 321 Works Progress Administration (WPA), 243 Universal Credit (UK), 169, 297–​98 UVW (Employee Insurance Agency), 98 video-​based job counseling, 10. see also job search counseling visitation toolbox with the dialog guide (vis itationsværkstøjskassen with dialog-​ guiden), 224 vocational training, 19, 191, 257, 265–​66. See also training retraining, 55 vouchers for, 210–​11 voluntary programs, 250t vouchers, 209–​12 vulnerable groups, 26 wages, 21 wage subsidies, 27, 226, 284–​85 Wagner-​Peyser Employment Services, 244 Wallonia, 289–​90 Washington, 320 Weber, Andrea, 248–​49 Weidinger, Matt, 3, 14–​16, 122–​49 Weise, Frank-​Jürgen, 210 Weishaupt, J. Timo, 10, 17–​18, 183–​203 Welfare and Employment Agency (NAV), 104 welfare-​to-​work programs, 29, 250t Western Europe, 204–​5. See also specific countries Wilkins, Roger, 16, 150–​82 Wilson, Dan, 312–​13 Wisconsin, 320 Wisconsin Regional Training Partnership (Milwaukee), 268–​69 women disability benefits recipiency rates, 188f, 189 labor force participation, 6–​7, 131, 132f, 133, 188f, 189, 324

Index 351 mobilization of, 287 mothers “needed at home or not actively seeking work,” 125 wage and salary income for working-​age mothers, 133–​34, 134f work, 94–​96, 130. See also employment work ability assessment method (arbejdsevnemetoden), 224 WorkAdvance, 268–​70 Work and Health Programme (UK), 221 Work and Social Assistance Act (WWB), 95–​ 96, 98–​99, 100t, 110–​11 Work Capability Assessment, 168–​69 work disability, 154. See also disability Work Disability Pension (WDP), 169–​71, 172 Worker Profiling and Re-​employment Service, 63 worker protections, 21 Workers Compensation (US), 185 work experience programs, 12 workfare, 91–​92, 112, 113, 214, 288 work-​first approach, 18, 291–​92 in public employment sevices (PES), 215, 216–​17, 224, 230, 233 in social assistance programs, 91, 109, 112–​13 workforce development, 17–​19. See also public employment services (PES) evidence base, 270–​73 private support for, 265–​67 public sector support for, 245–​48, 265–​ 67, 270–​73 targeting employment and training services, 263–​65 training programs, 248–​63, 250t US, 242–​82 for youth and young adults, 265–​67

Workforce Innovation and Opportunity Act (WIOA), 244, 263–​66, 270, 272–​73 Workforce Innovation Fund (WIF), 267–​68 Workforce Investment Act (WIA), 244–​45, 248 effectiveness, 250t, 257–​58, 262, 271–​72 targeting of services, 263 Workforce Investment Areas (US), 244 Workforce Investment Boards (WIBs), 268–​70 workhouse test, 113 Working Families’ Tax Credit (UK), 95 Working Tax Credit (UK), 95, 103 Work Programme (UK), 65, 111–​12, 218–​21, 220t, 222–​23, 299 work-​sharing, 321–​22 Works Progress Administration (WPA), 243 World Bank, 9 World Health Organization (WHO), 154, 196–​97 Year Up, 247–​48, 260–​61, 266–​67, 269–​70 youth and young adults active labor market policies (ALMPs) for, 250t, 259–​60 New Deal for Young People (NDYP), 93, 95, 319–​20 not in employment, education, or training (NEETs), 26, 40b social assistance for, 225 Title I, 265–​66 training programs for, 250t, 259–​61 unemployment, 25–​26, 222, 324 workforce development services for, 247–​ 48, 265–​67 Youth Contract (YC), 222 Youth Guarantee (EU), 40b Youth RSA (France), 109 Ziebarth, Nicholas, 16, 150–​82