Handbook on Austerity, Populism and the Welfare State 2021932709, 9781789906745, 9781789906738

Citation preview

HANDBOOK ON AUSTERITY, POPULISM AND THE WELFARE STATE

ELGAR HANDBOOKS IN SOCIAL POLICY AND WELFARE This exciting series provides a comprehensive overview of cutting edge research on Social Policy and Welfare, forming a definitive guide to the subject. The Handbooks present original contributions by leading authors, selected by an editor internationally recognised as a preeminent authority within the field. Titles in the series are truly international in their scope and coverage, and use a comparative approach to analyse key research themes. Equally useful as reference tools or high-level introductions to specific topics, methods and debates, these Handbooks will be a vital resource for academic researchers and postgraduate students.

Handbook on Austerity, Populism and the Welfare State Edited by

Bent Greve Professor of Welfare State Analysis, Department of Social Sciences and Business, Roskilde University, Denmark

ELGAR HANDBOOKS IN SOCIAL POLICY AND WELFARE

Cheltenham, UK • Northampton, MA, USA

© Bent Greve 2021

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, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA A catalogue record for this book is available from the British Library Library of Congress Control Number: 2021932709 This book is available electronically in the Sociology, Social Policy and Education subject collection http://dx.doi.org/10.4337/9781789906745

06

ISBN 978 1 78990 673 8 (cased) ISBN 978 1 78990 674 5 (eBook)

Contents

List of contributorsviii PART I 1

INTRODUCTION

Introduction to the Handbook on Austerity, Populism and the Welfare State2 Bent Greve

PART II

CONCEPTUAL AND METHODOLOGICAL ISSUES

2

What is austerity? Kevin Farnsworth and Zoë Irving

11

3

The politics of retrenchment Peter Starke

25

4

Populism and the welfare state Cristóbal Rovira Kaltwasser and Lisa Zanotti

41

5

Measuring retrenchment in welfare states: overcoming the challenges to the definition, operationalization and measurement of welfare policy change Elisa Helena Xiol Y Ferreira and Michael Howlett

54

6

The dependent variable problem revisited: methods, concepts, and scope in the welfare retrenchment literature Mehmet Fuat Kına and Erdem Yörük

64

7

Understanding the ‘welfare state’ in the context of austerity and populism Sonja Blum and Johanna Kuhlmann

81

8

Austerity, populism, and the politics of blame: an ideational perspective  Daniel Béland and Alex Waddan

94

9

The social legitimacy of European welfare states after “the age of austerity” Femke Roosma

110

10

Austerity and poverty Paul Spicker

130

PART III COUNTRY AND WELFARE REGIMES – ANALYSIS OF AUSTERITY/POPULISM 11

Nordic welfare state changes especially in the light of migration and the financial crisis Bent Greve and Jon Kvist v

143

vi  Handbook on austerity, populism and the welfare state 12

Fiscal austerity, welfare retrenchment and political populism in Continental European welfare states Jan-Ocko Heuer

13

The United Kingdom before and after Brexit Benjamin Leruth and Peter Taylor-Gooby

170

14

South Europe: reclaiming welfare post-crisis? Maria Petmesidou and Ana Marta Guillén

186

15

Austerity, populism and welfare retrenchment in Central and South Eastern Europe Noémi Lendvai-Bainton and Paul Stubbs

207

16

Support to families with children in the Baltic States: pathways of expansion and retrenchment from 2004 to 2019 Jolanta Aidukaite

221

155

PART IV ARE SPECIFIC WELFARE PROGRAMS MORE PRONE TO AUSTERITY – AND IF SO, WHY? 17

Incremental or paradigm shifting? Evidence about the retrenchment of public pension schemes in the industrialised world from expenditure and replacement rate data, 1980–2015 Paul Bridgen

244

18

Unemployment benefits in the 21st century: new dimensions of retrenchment and the roles of austerity and populism Axel Cronert

19

Austerity and its corresponding effects on public safety and crime Adegbola Ojo

281

20

Family policy in Europe in the era of austerity and populism Mikael Nygård and Mikko Kuisma

294

21

Long-term care policies meet austerity Barbara Da Roit

312

22

Changes in tax systems Nelly Popova

328

23

Labour markets in post-crisis Europe: liberalisation, deregulation, precarisation 344 Dragoș Adăscăliței and Jason Heyes

24

The impact of austerity on social activism  Shana Cohen

360

25

Gender, austerity and the welfare state Sidita Kushi and Ian P. McManus

375

265

Contents  vii PART V 26

CHANGE TO THE ROLE OF WELFARE STATES?

Reflection upon the development of, and the future for, welfare states Bent Greve

396

Index401

Contributors

Dragoș Adăscăliței is a Lecturer at the University of Sheffield, Management School. His research focuses on labour market policies, welfare reforms and industrial relations systems. His work has appeared in, among others, Economic and Industrial Democracy, European Journal of Industrial Relations, European Urban and Regional Studies, Work, Employment & Society, Employee Relations and Social Policy & Administration. Jolanta Aidukaite (PhD in Sociology from Stockholm University) is a Chief Researcher at the Lithuanian Social Research Centre. She has published extensively on the topics of social policy, family policy, housing policy and community mobilization. Her articles are published in such journals as Communist and Post-Communist Studies, Social Inclusion, International Journal of Sociology and Social Policy, Journal of European Social Policy, East European Politics, Journal of Baltic Studies and GeoJournal. Daniel Béland is Director of the McGill Institute for the Study of Canada and James McGill Professor in the Department of Political Science at McGill University, Canada. A student of social and fiscal policy, he has published 20 books and more than 150 articles in peer-reviewed journals such as Governance, Journal of Social Policy and Social Policy & Administration. Sonja Blum is a Researcher and Lecturer at the FernUniversität in Hagen, and Fellow at the KU Leuven Public Governance Institute, Belgium. She holds a PhD in Political Science from the University of Münster. Her specialization is in public policy and comparative social policy. Paul Bridgen is an Associate Professor of Social Policy in the School of Economic, Social and Political Sciences, University of Southampton, UK. He has published extensively on comparative pensions policy, often with Traute Meyer, with a particular focus on the UK and other liberal systems. He has also worked on the impact of migration on pension outcomes. Shana Cohen is the Director of TASC, an independent think tank based in Dublin that focuses on issues related to inequality and democracy in Ireland and the European Union, and an Associate Researcher with the Department of Sociology at the University of Cambridge, UK. She is the author of Transforming Social Action into Social Change: Improving Policy and Practice (2017, Routledge) and the forthcoming book, Re-making Society and the State from Below: Civil Society in an Age of Discord and Disorder (2022, Policy Press). She also has extensive experience working with non-governmental organizations (NGOs) and community-based organizations in a number of countries, including Morocco, the USA, the UK and India. Her work has involved project design, management, and evaluation, as well as helping NGOs respond to policy. Axel Cronert is a Political Scientist at Uppsala University, Sweden, who defended his PhD thesis on the political economy of active labour market policy in 2018. His current research agenda includes the political economy of public investment, the migration–welfare state nexus and coalition politics. viii

Contributors  ix Barbara Da Roit is Professor of Sociology at Ca’ Foscari University of Venice, Italy. Her research, embedded in comparative welfare studies, focuses on the relationship between changes in social policies and social practices with a specific interest in the field of care for elderly people and young children. She has participated in and coordinated a number of national and international research projects on, among other topics, formal and informal care work and their gender and class dimension, the conciliation between paid and unpaid work, the nexus between the transformation of care and gendered work migration, the changing ideas and practices about care work. She regularly publishes in international top sociology and social policy journals. Kevin Farnsworth is Reader in Social Policy at the University of York, UK. His research interests include the political economy of welfare states, business and public policy and welfare states and economic crisis. He pioneered the development of a searchable database for UK Corporate Welfare to improve transparency of government support for business and inform public debate. He has published numerous books, chapters and articles on the topics of corporate welfare, business power in social policy and crisis and austerity. With Zoë Irving he is founding co-editor of the Journal of International and Comparative Social Policy. Elisa Helena Xiol Y Ferreira is a Brazilian and Canadian PhD student in Political Science at Simon Fraser University, Canada. She received her baccalaureate in Social Sciences with Honours in International Development from the University of Ottawa, also recognized by the University of Brasilia as a degree in Political Science, and her master’s in Political Science from Simon Fraser University. Her work focuses on public policy and comparative studies, primarily education policy, with experience from the BC Council for International Education among other roles in the public and private sectors. Bent Greve is Professor in Social Science with an emphasis on welfare state analysis at Roskilde University, Denmark. His research interest focuses on the welfare state, and social and labour market policy, often from a comparative perspective. He has published extensively on social and labour market policy, social security, tax expenditures, public sector expenditures and financing of the welfare state. He is editor of Social Policy & Administration. Recent books include Myths, Narratives and the Welfare State (2021, Edward Elgar), Austerity, Retrenchment and the Welfare State (2020, Edward Elgar), Poverty. The Basics (2020, Routledge), Routledge International Handbook of Poverty (ed.) (2020, Routledge), Welfare, Populism and Welfare Chauvinism (2019, Policy Press), Routledge Handbook of the Welfare State (ed., 2nd edition) (2019, Routledge). Ana Marta Guillén is Professor of Sociology at the University of Oviedo, Spain. Her research interests include welfare state development, comparative social and labour policy, healthcare policy, political economy and the processes of Europeanization and European integration. She has published articles in journals such as Social Science & Medicine, Social Politics, Journal of European Social Policy, European Societies, International Journal of Health Services, West European Politics, Social Policy and Administration and Comparative European Politics, among others. She has acted as a consultant to the European Commission and several EU presidencies. She is a member of the editorial boards of the Journal of European Social Policy and the International Social Security Review. She is also part of the Advisory Board of the European Social Observatory (OSE, Brussels).

x  Handbook on austerity, populism and the welfare state Jan-Ocko Heuer is a sociologist working as Postdoctoral Researcher at the University of Bremen, Germany. He has researched and published on social policy, consumer insolvency/bankruptcy and research data management. Recent social policy publications include ‘Unravelling Deservingness: Which Criteria Do People Use to Judge the Relative Deservingness of Welfare Target Groups? A Vignette-based Focus Group Study’ with Katharina Zimmermann (Journal of European Social Policy, 2020), ‘Regimes, Social Risks and the Welfare Mix: Unpacking Attitudes to Pensions and Childcare in Germany and the UK Through Deliberative Forums’ with Peter Taylor-Gooby et al. (Journal of Social Policy, 2020), and ‘Changing Preferences Towards Redistribution: How Deliberation Shapes Welfare Attitudes’ with Katharina Zimmermann and Steffen Mau (Social Policy & Administration, 2018). Jason Heyes is Professor of Employment Relations and Director of the Centre for Decent Work (CDW) at Sheffield University Management School, UK. He has published on topics including flexicurity, European labour market policy, vocational education and training and employment protection legislation in journals such as Economic & Industrial Democracy, European Journal of Industrial Relations and Work, Employment & Society. His current research interests include dependent self-employment, underemployment, labour administration and working carers. Michael Howlett is Burnaby Mountain Professor and Canada Research Chair (Tier 1) in the Department of Political Science at Simon Fraser University, Canada. He specializes in policy design and studies of policy-making processes. Zoë Irving is Reader in International and Comparative Social Policy at the University of York, UK. Her research interests are in the areas of work and employment, and the relationship between social policy and population size with a particular interest in the social politics of small island states and especially Iceland. She has published widely, with Kevin Farnsworth, on welfare states, economic crisis and austerity, including two co-edited collections with Policy Press, Social Policy in Challenging Times (2011) and Social Policy in Times of Austerity (2015). She is co-author of Exploring the World of Social Policy: An International Approach (Policy Press, 2020), with Michael Hill, and co-editor of the Journal of International and Comparative Social Policy. Mehmet Fuat Kına is a PhD candidate in the Sociology Department at Koç University, and a researcher at the ERC-funded Emerging Welfare project. He studies the relationship between social movements and social welfare provisions. Johanna Kuhlmann is a Postdoctoral Researcher at SOCIUM Research Center on Inequality and Social Policy and the Collaborative Research Centre ‘Global Dynamics of Social Policy’ (CRC 1342), University of Bremen, Germany. She received her PhD from the University of Münster. Her research focuses on comparative social policy and theories of the policy process, with a focus on the role of causal mechanisms. Mikko Kuisma is Program Lead and Lecturer at the Institute of Political Science, University of Tübingen, Germany. His research interests lie in the comparative political economy of welfare states, social rights and citizenship, the role of ideas and discourses in welfare reform, and European populist radical right parties. His work has been published in a number of edited collections and peer-reviewed journals, such as Public Administration, New Political

Contributors  xi Economy, Critical Social Policy and Critical Policy Studies. Since 2018, he has been the managing editor of the Journal of European Social Policy. Sidita Kushi, PhD, is Assistant Professor of Political Science at Bridgewater State University, USA. She recently served as a fellow and research director at the Center for Strategic Studies, the Fletcher School of Law and Diplomacy at Tufts University, where she led the Military Intervention Project (MIP). She is the author of a range of academic articles on military interventions, intrastate conflict, as well as the gendered dynamics of economic crises, published in Comparative European Politics, European Security, International Labour Review, Mediterranean Quarterly, the Journal of Science Policy & Governance, amongst others. Jon Kvist is Professor of European Public Policies and Welfare Studies at the Department of Social Science and Business, Roskilde University, Denmark. He has participated in various Danish and other government committees and research councils. Currently, he is a member of the government commission on Minimum Income Benefit System and the European Commission’s ESPN – European Social Policy Network. He has published widely on the Nordic welfare model, social investments, Europeanization and comparative methods. Noémi Lendvai‐Bainton is a Hungarian‐born academic who has been working in the UK since 2002. She is currently Senior Lecturer in Comparative Public Policy at the School for Policy Studies, University of Bristol, UK. Her research interests include post‐communist welfare states, EU integration, EU social policy, global social policy and the impact of international organizations, East–West migration, comparative research methods and critical policy studies. Her 2015 book co‐authored with John Clarke, David Bainton and Paul Stubbs, Making Policy Move: Towards a Politics of Translation and Assemblage (Policy Press), explores ‘translation’ as a possible new theoretical lens for critical policy studies. She has also widely published on the transformation and Europeanization of post‐communist welfare states. Benjamin Leruth is Assistant Professor in European Politics and Society at the University of Groningen, the Netherlands. His research focuses on Euroscepticism, differentiation in the EU and the politics of the welfare state. His publications on welfare politics include After Austerity (Oxford University Press, 2017, with Peter Taylor-Gooby and Heejung Chung) and Attitudes, Aspirations and Welfare (Palgrave, 2018, with Peter Taylor-Gooby). Ian P. McManus is Assistant Professor of Political Science at Emerson College, USA, whose work focuses on the effects of macroeconomic changes on political competition and social well-being. His research interests include welfare state politics, social inequality, gender equality, economic crises, labour markets, European politics and the political economy of technology. He holds a PhD in Political Science from Northeastern University and was an LSE Fellow in Social Policy at the London School of Economics and a Postdoctoral Research Fellow at the University of Lisbon. He was the recipient of a Doctoral Research Grant from the German Academic Exchange Service (DAAD) and was a visiting scholar at the Free University of Berlin. Mikael Nygård, PhD, is Professor of Social Policy at Åbo Akademi University, Finland. His research interest ranges from research on societal participation of various groups, such as older adults, youth and people with disabilities, to health-related research and comparative welfare state analysis, with a special focus on family policy reforms.

xii  Handbook on austerity, populism and the welfare state Adegbola Ojo is Director of Teaching and Learning; Programme Leader; and Senior Lecturer in Urban Geography and Applications of Big Data at the School of Geography, University of Lincoln, UK. He received his PhD in Quantitative Human Geography from the University of Sheffield; his MSc in Geographic Information Science from University College London; and his BSc in Geography and Planning Sciences from the University of Ado-Ekiti, Nigeria. His research interests are focused on understanding and representing social and spatial dynamics and intricacies of population behaviour within a framework of interdisciplinary studies, population geography, quantitative social science and computer modelling. Maria Petmesidou is Professor Emerita at Democritus University, Greece. For several years she was fellow/scientific committee member of CROP (Comparative Research on Poverty) sponsored by the International Social Science Council and the University of Bergen. She has published widely on comparative social policy. Her recent publications include, among others: Economic Crisis and Austerity in Southern Europe (co-editor, Routledge, 2015); Child Poverty and Youth (Un)Employment and Social Inclusion (co-editor, CROP/Ibidem, 2016); ‘Welfare Reform in Greece’ in After Austerity, edited by P. Taylor-Gooby et al. (Oxford University Press, 2017); ‘Policy Transfer and Innovation for Building Resilient Bridges to Labour Market’ (co-author), in Youth Labour in Transition, edited by J. O’Reilly et al. (Oxford University Press, 2019); ‘Southern Europe’, in Routledge Handbook of the Welfare State (2nd edition), edited by B. Greve (2019); ‘Health Policy and Politics’, in The Oxford Handbook on Modern Greek Politics, edited by K. Featherstone and D. Sotiropoulos (Oxford University Press, 2020). Nelly Popova obtained a PhD degree in 2011 from the Institute of Economic Research of the Bulgarian Academy of Sciences. She is Assistant Professor at the Department of Finance, University of National and World Economy, Sofia, Bulgaria (since 2012). She has published a number of conference papers and articles. Her research interests are in the field of public finance, tax policy and international tax coordination. Femke Roosma is Assistant Professor of Sociology at the Department of Sociology, Tilburg University, the Netherlands. Her main research interests include solidarity and welfare deservingness, social policy, the social legitimacy of welfare states and the cross-national analysis of public opinions and attitudes. Her PhD thesis (2016) focused on the multidimensionality of welfare state legitimacy. Cristóbal Rovira Kaltwasser is Professor of Political Science at Diego Portales University in Santiago de Chile and Associate Researcher of the Centre for Social Conflict and Cohesion Studies (COES). His main area of research is comparative politics and he has a special interest in the ambivalent relationship between populism and democracy. Paul Spicker is Emeritus Professor of Public Policy at Robert Gordon University, Aberdeen, Scotland, now an independent writer and commentator on social policy. His published work include 20 books, several shorter works and 90 academic papers; his most recent work has been published as The Poverty of Nations: A Relational Perspective (Policy Press, 2020). He is a consultant on social welfare in practice and has done work for a range of agencies at local, national and international levels. Peter Starke is Professor (with special responsibilities) at the University of Southern Denmark (SDU). He works on the politics of welfare state change in comparative perspective,

Contributors  xiii including the politics of retrenchment, the welfare state and globalization, war, crisis management as well as the relationship between welfare, crime and punishment. He has published in journals such as Policy Studies Journal, Journal of European Public Policy, Journal of Criminal Justice and Journal of European Social Policy and co-edited Warfare and Welfare (Oxford University Press, 2018). Paul Stubbs is a UK‐born sociologist who has lived and worked in Croatia since 1993. He is currently Senior Research Fellow in the Institute of Economics, Zagreb. His work focuses on policy translation, social protection and the history of the non‐aligned movement. His latest book, co‐edited with Sofiya An, Bob Deacon and Tatiana Chubarova, is Social Policy, Poverty, and Inequality in Central and Eastern Europe and the Former Soviet Union: Agency and Institutions in Flux (CROP/Ibidem, 2019). He is on the editorial board of Critical Policy Studies and the Croatian Journal of Social Policy. He is a former Co-President of the Association for the Anthropology of Policy (ASAP) of the American Anthropological Association. Peter Taylor-Gooby, FBA, FASS, FRSA, OBE, is Research Professor of Social Policy at the University of Kent, UK. He led the Norface ‘Welfare State Futures: Our Children’s Europe’ project and a number of previous EU and ESRC programmes. He is currently advising the UK Government on responses to Covid-19. His most recent novel is Blood Ties (Matador, 2020). Alex Waddan is Associate Professor in Politics at the University of Leicester, UK. He specializes in US and comparative social policy and has published five books and numerous articles in peer-reviewed journals such as Political Studies, Policy and Politics, Journal of Social Policy and Policy Studies Journal. Erdem Yörük is Associate Professor in the Department of Sociology at Koç University in Istanbul, Turkey, and Associate Member in the Department of Social Policy and Intervention at the University of Oxford. He is also the principal investigator of an ERC-funded project, Emerging Welfare (emw.ku.edu.tr). Combining computational social science, quantitative and qualitative methods as well as comparative and historical perspectives, his work so far has focused on welfare systems and social policy, social movements and politics in emerging market economies. He earned his PhD in Sociology from Johns Hopkins University in 2012. He has published in New Left Review, World Development, Governance, Politics & Society, Social Policy & Administration, South Atlantic Quarterly, Current Sociology, Social Indicators Research and International Journal of Communication. His book, The Politics of the Welfare State in Turkey, is forthcoming from the University of Michigan Press. He has been awarded several competitive research grants, including a European Commission Twinning Action Grant, Turkish Science Academy BAGEP Award, ERC-Starting Grant, Marie Curie CIG, NSF Doctoral Dissertation Improvement Grant, and the Ford Foundation MERC Dissertation Grant. He is an associate editor of the SSCI journal European Review. Lisa Zanotti is a Postdoctoral Fellow at Diego Portales University in Santiago de Chile. She holds a PhD in Political Science from Diego Portales University and in Humanities from Leiden University. Her academic publications focus on populism, party systems and attitudes toward democracy. She is co-author of VOX: The Rise of Spanish Populist Radical Right (Routledge, forthcoming 2021).

PART I INTRODUCTION

1. Introduction to the Handbook on Austerity, Populism and the Welfare State Bent Greve

INTRODUCTION This chapter sets the scene for the book, including the reasons for the choice of concepts and how and why the link between these are important elements in order to understand societal development in welfare states. It also argues that the presentation and discussion of the development in a variety of different welfare state regimes can be important in order to understand the impact of ideas, ideologies and variations in policies and decisions in a range of countries. Furthermore, the analysis of countries and groups of countries gives one kind of information, but analysis across a number of different sub-sections of the welfare states aims at making it possible to understand in a broader way why there might have been austerity/retrenchment in certain parts of the welfare states, but not necessarily in others. This also includes why populism can have had an impact in certain social policy areas, although not in all of them, as some welfare areas also have, from time to time, strong support from populist parties. The argument for the choice of framework for the country/regime analysis is based upon classical depictions on welfare regimes (Greve 2019a). Thus, it includes the historical three worlds of welfare capitalism, but other regimes are also included. As retrenchment/austerity has especially been a European discussion, this is the main region to be covered, although other countries are also included to a more limited extent. In order to understand changes, there is also the need to interpret in a methodological way when and how austerity takes place as well as how this might be influenced by populist traits. This is done in several chapters throughout the book, for example, Chapter 5, but is also touched upon below when setting the scene for the book. Lastly, this first chapter gives an overview of the chapters in the book.

SETTING THE SCENE Austerity has been an issue in welfare state analysis for a long time, and it has even been argued to be a permanent condition (Pierson 2001). Presumably, this reflects the fact that from the 1930 Crisis to 2008, the use of the word ‘austerity’ in books was at its highest level in the late 1980s, and gradually reduced to a very low level in 2008 just before the last financial crisis. A recent book on austerity (Blyth 2013) has been cited more than 2500 times as of April 2020, indicating that austerity is a word often used in many books and articles. There has also been increasing interest in the related concept of ‘retrenchment’. The difference between these is not always clear-cut (Greve 2020), but more detail can be found in the book’s two chapters discussing these concepts. 2

Introduction to the Handbook on Austerity, Populism and the Welfare State  3 Earlier studies have indicated that despite the discussions very few welfare states have seen fundamental shifts (Hacker 2004). It has even been argued that welfare states are often a frozen landscape (Palier and Martin 2007). One problem with the analysis of austerity and retrenchment is that even if the same amount of, or more, resources are available, there is still also the question of how to prioritize resources and what to focus on in the different welfare states’ development (Judd et al. 2015). Thus, it might be that there is reduction in one part of the welfare state, while at the same time expansion in other parts; however, in such cases, the use of the word austerity as a generic term for the development as a whole is misleading. I have elsewhere defined austerity as both having a possible permanent and more restricted understanding, such that: ‘Permanent austerity will be defined as: a constant reduction in the welfare state’s position in society and a continuously stronger focus on the market and civil society’s role. In contrast to permanent austerity, there can be restricted austerity, causing one or more parts of the welfare states to witness cut-backs – whether in spending or eligibility conditions’ (Greve 2020, 21). This indicates that methodological analysis of austerity needs to have a timeframe so that, for example, a one-year reduction in spending is not understood as austerity given that the following year a return to the previous spending level is possible. It might further be the case that the welfare states continue at the same level, but that the democratic system changes priorities so that one area has less money, but another has been expanded. Therefore, just arguing that there has been austerity or retrenchment without being more precise about what has actually taken place would be misleading. This also includes, for example, taking change in demography and price development into consideration. This might be a reason why ‘more generally there is surprisingly little evidence that international economic crisis has led to sudden and fundamental welfare state reform’ (Pierson et al. 2016, 264). There is a long tradition in welfare state analysis for arguing that there has been austerity and/or retrenchment. Thatcher and Reagan are often mentioned as having an agenda of dismantling welfare states; however, they were seemingly not able to make considerable change, as ‘on the whole, welfare state programs demonstrated considerable resilience during the tenures of both leaders’ (Pierson 1994, 164). The way one can analyse changes related to austerity is to look into change in long-term as well as short-term spending; structure of policy spending; and systemic and programmatic retrenchment (Pierson 1994). This is also an argument for why this book has a number of chapters discussing the concepts and also how they relate to different welfare state regimes both across the board and individually, but also specific sub-sectors of the welfare states. The reason why populism is also included in the analysis, both conceptually and analytically, is that it may have contributed to a number of changes in welfare states in recent years. This is true in relation to welfare chauvinism, that is, a tendency for welfare to go especially to the citizens of the nation-state and to those considered as deserving to receive these benefits (Greve 2019b). Thus, even if there is not permanent austerity then there might be retrenchment in certain sectors, which is seen by populists as benefits going to non-deserving people, sometimes argued to be welfare scroungers, skivers, lazy unemployed and so on (Hobbins 2016; Trlifajová and Hurrle 2019). Names such as welfare queens, scroungers and skivers indicate partly that the debates are not only related to populist approaches, but also to the narratives’ possible impact on welfare states’ development (Jurado and Brochmann 2013; Cromby and Willis 2014).

4  Handbook on austerity, populism and the welfare state

OVERVIEW OF THE STRUCTURE OF THE BOOK AND INDIVIDUAL CHAPTERS The book is divided into five parts, of which this chapter is the first part. The aim in Part II is to present conceptual and methodological issues, followed by welfare state development based upon a country/regime-specific approach in Part III. Part IV then aims to discuss whether one has been witnessing austerity/retrenchment in central welfare state programmes, combined with whether they are more likely than others to be changed. National examples are included. The regime chapters, in Part III, have the overall intention to describe briefly how to understand the welfare regimes, followed by analysis and interpretation of the changes in the last 10–15 years in order to discuss whether there has been austerity in the different regimes and countries within these. This will be a general discussion but also with examples from different sub-sections of the welfare states. The chapters on the different policy sectors present the main characteristics and principles of the policy sector, and thereafter discuss changes in recent years, including, depending on the sector, the main drivers for the change ranging from demography, the financial crisis, use of new technology, to the possible impact of populism. Given the variety of authors and topics, it is not surprising that different positions and interpretations of what has happened can be found. Furthermore, as a consequence of the format of the book, there will also be some overlap between the different chapters. A short overview of the individual chapters is given below. The book starts with a number of chapters with a focus on definitional and measurement issues. Chapter 2 by Farnsworth and Irving sets the scene for an understanding of the concept of austerity, although, as argued, it is a slippery concept. This also implies that the measurement hereof might be difficult (see also Chapter 5). However, it is pointed out that it is not only an economic approach that is needed, but also a political one, by referring to the debate about the ‘necessary politics’. Finally, what has been done in several countries after the last financial crisis has influenced different sections of society very differently. Chapter 3 by Starke discusses the concept of retrenchment, arguing that there can be systemic and programmatic retrenchment, and in relation to retrenchment also how key actors and policy-makers react and reflect hereupon. Furthermore, there can be motivational issues, decision making but also informational elements at stake. Finally, the chapter discusses framing and blame-avoidance strategies, including referring to populism in relation to retrenchment, while also concluding that individual attitudes and government strategies are the important aspects related to changes in the welfare states. Rovira Kaltwasser and Zanotti in Chapter 4 present the concept of populism and link it to the debates on the development of the welfare states, including that populism can imply a challenge for welfare states, especially from welfare chauvinism. A reason being that there can be a difference between who shall be given benefits and who shall pay for them. Finally, they reflect that given populist forces might be more exclusionary or inclusionary this can also have an impact on the direction of changes in welfare states. Chapter 5 by Ferreira and Howlett focuses on how to measure retrenchment. This is a complex subject given that there are many issues to be aware of in this endeavour including whether to look at institutional change, entitlements, aggregate expenditure data or replacement rates as key measures. The chapter notes how different policy elements can have an impact on outputs and outcomes, also implying that one needs to be aware of how one meas-

Introduction to the Handbook on Austerity, Populism and the Welfare State  5 ures developments, including what to and what not to include in the analysis. In this way the chapter also highlights the classical issue of what is the dependent variable, which is the focus of Chapter 6. Chapter 6 by Kına and Yörük looks precisely into the issue of the dependent variable by first making a systematic literature review, including an overview of the use of different international datasets to see what kind of data is used in order to analyse whether there has been austerity/retrenchment or not, and showing that expenditure is still the most used variable. The analysis of a number of existing studies looking at expenditure, entitlement/rights and legislative impact shows that there is no common knowledge about whether there has actually been retrenchment or not, and in some cases there has only been mixed or partial retrenchment. Chapter 7 by Blum and Kuhlmann explores the understanding of the concept of welfare in a context of austerity and populism. Definitions of ‘the welfare state’ are discussed, and key developments of the welfare state throughout the 20th and 21st centuries are highlighted. Many of the changes since the ‘Great Recession’ have implied a stronger selectivity and tendencies to exclude certain persons from access to different parts of the welfare states. Blum and Kuhlmann highlight three contemporary faces of the austerity–populism nexus: a welfare-hostile, a welfare-friendly, as well as a welfare-ambiguous face. Based on a literature review, they conclude that populist parties today have largely disbanded a welfare-hostile austerity face, but neither are they welfare-friendly per se: the dominant face is that of welfare chauvinism. Chapter 8 by Béland and Waddan investigates how ideas can be linked to the issues of austerity, populism and the politics of blame. Focusing especially on the UK, there seem to be clear examples, including in relation to Brexit, that there is an interaction between ideas and the use of the options related to austerity. If linked to populist stances, this will be part of how to understand the development. Further, if it can be linked to the issues of not only blame avoidance, but also blame generation to understand the interaction between austerity and populism. Changes to the welfare state can have an impact on the legitimacy. This is precisely the focus of Chapter 9 where Roosma investigates the change of legitimacy after what is labelled ‘the age of austerity’. This is done by looking into how a number of conditions are attached to measuring and understanding legitimacy. Data from the European Social Survey indicate that welfare attitudes, and thereby the legitimacy of the welfare states, are, in fact, rather stable, except that fewer people today believe the welfare states are too costly for companies. Chapter 10 by Spicker then looks into the relation between austerity and poverty, including a discussion of the ambiguities of the concepts that are being used, highlighting that ‘austerity’ is often identified, not just with reduced public spending, but with minimal services. As part hereof it shows that the relation is not necessarily the same in all countries between spending on one side and poverty on the other. It is argued that this is only partly due to austerity, as the balance, for example, between public and private provision also has an impact. Part III deals with how different welfare states in different welfare regimes have coped with the crisis. This part opens with Chapter 11 on the Nordic welfare states by Greve and Kvist, especially how the financial crisis and migration have influenced the development, by looking into change in inequality and a number of policy areas. The chapter concludes that at the overall level there has not been austerity and retrenchment in the Nordic welfare states, while not neglecting that populism has had an impact, especially in relation to migration, and

6  Handbook on austerity, populism and the welfare state thereby change in some social benefits such as unemployment benefits and social assistance has taken place. Chapter 12 by Heuer analyses the Continental welfare states, and shows that even if they have been described as frozen landscapes there have been strong changes, though mainly before the financial crisis. Despite the relatively strong economy there has been a growth in the populist parties, and they have had to balance their (increasingly working class) voters’ wish for more welfare, resulting in opposition to retrenchment, with the aims of their potential (centre-right) coalition partners, resulting in support for retrenchment when in government. Presumably a conflict that also can be witnessed in other welfare states for more right-wing populist parties. Chapter 13, by Leruth and Taylor-Gooby, looks at the UK as an example of a liberal welfare state where there is a long tradition of having a lean welfare state, with only a limited level of spending. The chapter shows that retrenchment has a long history, and thus is not a new aspect of the development. It further relates the development to the debate on Brexit as well as the issue of migration. Brexit and migration have both added to the uncertainties of the future of the welfare state in the UK, and it is argued that UK welfare chauvinism is shaping a welfare state with strong discrimination between immigrants and denizens. Southern Europe is one of the regions that bore the brunt of the financial and sovereign debt crisis. In Chapter 14 Petmesidou and Guillén explore reform paths prior to and in the wake of the crisis in this region. Variation in welfare recalibration in the 1990s and 2000s among South European countries was followed by a convergent trend of austerity and retrenchment. In the last few years, recovery has varyingly facilitated the reversal of some unpopular reforms and a slight rise in spending. Yet South European countries have still not returned back to the pre-crisis level in several areas. Trust in institutions has declined and (with the exception of Portugal) populist parties and ideas have surged. Whether the social and economic aftershocks of the COVID-19 pandemic will provide stronger leeway for a populist impact in the future is an open question. Chapter 15 by Lendvai-Bainton and Stubbs considers developments in Central and South Eastern Europe. The authors show how austerity politics and policies have grown apace in most of the countries in recent years, and that this is driven by right-wing populist parties, including a strong nationalist approach using the window of opportunity to make change supporting a ‘deserving’, ‘ethnicized’ middle class, but with negative impact on poverty and inequality. This has also included a stronger role for the market in the delivery of welfare. In Chapter 16 by Aidukaite the focus is on the Baltic states, which in fact witnessed an expansion in the welfare states shortly after the financial crisis, and since then, however, have seen a number of retrenchment initiatives. In the chapter this is shown by a focus on family policy and especially how benefits in place for a short time are easier to retrench than benefits in place for a long time, thus indicating that welfare policies strongly embedded in a society with support are less likely to witness austerity. After these presentations of the impact in different welfare regimes, Part IV then moves to the issues of how and whether change has been witnessed in a number of different policy fields within the welfare states. Chapter 17 by Bridgen looks into pension systems and considers the scale of retrenchment in this area since the 1980s. Using pension expenditure data, adjusted for population, and social rights data for 18 OECD (Organisation for Economic Co-operation and Development countries), he finds retrenchment has been incremental in the large majority of settings.

Introduction to the Handbook on Austerity, Populism and the Welfare State  7 Pensioner support has continued strongly in most countries, with respect both to resource allocation and benefit generosity, although there has been a slight reduction in this commitment more recently. This despite the many pensions reforms over the last 30–40 years due to the demographic changes. This does not imply that the changes have been similar across welfare regimes and countries, for example, there has been a reduction in countries such as Germany and Sweden, thereby also questioning regime theory as a basic for understanding change in welfare states. This is followed by Chapter 18 by Cronert on one of the other central welfare state benefits, that is, unemployment benefits, where the picture is different from the one related to pensions. The trend towards retrenchment in the 1980s and 1990s has continued into the 21st century, but has taken new forms. While patterns vary across welfare regimes, the overall trend is towards more degressive replacement rates, reduced inclusiveness and stricter conditionality. These changes can be related to both fiscal austerity and populist parties. Chapter 19 by Ojo then puts the focus on public safety and crime, looking specifically into the case of the UK. Blaming others seems to be a common trend, while there have also been spending cutbacks within the field, which also have an impact on the possibilities and options for doing preventative work. Naturally, this is one of the fields where the cause and effect of change in measures can give rise to a number of different issues to be aware of. Chapter 20 by Nygård and Kuisma focuses on family policy, a newcomer to social policy. The authors point out that while most European countries increased spending immediately after the financial crisis, there have since been reductions in most countries with regard to cash benefit, whereas the opposite is the case with benefits in kind. This indicates that many countries have adopted an approach that emphasizes the social investment perspective and support towards more gender equality. Remarkably, within one policy area both retrenchment and expansion can take place at the same time. Furthermore, the authors also demonstrate that family policy change in Europe has not been purely about spending. Family policy is also being used as a political tool by European populists, for instance in Hungary and Poland, who have taken an anti-austerity turn by using family policy in promoting nationalist and conservative family values. Long-term care is developed very differently across welfare states, and it is important to analyse whether it has been influenced by the crisis. This is done in Chapter 21 by Da Roit. The chapter shows that changes differ across welfare regimes, partly due to long-term care being a latecomer to social policy in many welfare states. Still, it is also a part of the welfare state where the change in demography has implied an increase in need of welfare state support. Overall, it seems that the crisis has not implied any substantial change in policy. Chapter 22 by Popova then looks into whether there has been any change in the tax and duty system. It does this by presenting approaches to taxes and duties, and the core challenges to the systems such as globalization, digital economy and population ageing. It is argued that there has been a decline in corporate taxes, less progressivity than earlier, but also that indirect taxes have increased. In Chapter 23 Adăscăliței and Heyes analyse labour market policies and show that retrenchment to deregulation have not solved the problems of unemployment, underemployment and more broadly the issue of precarity in Europe. Underemployment has even increased in many European countries. The authors also note that in the context of the East–West migration, wages have grown faster than productivity in the new member states. Not only have classical welfare state areas been influenced by the development in recent years, and in Chapter 24, by Cohen, there is a presentation of the consequences of social

8  Handbook on austerity, populism and the welfare state activism using a UK case study to show how the crisis has implied the need for stronger local activism, including elements such as food-banks and other activities, which might contradict national policies. However, there has also been the pressure of having fewer resources available locally. It is not obvious that all individuals and groups will be influenced by austerity to the same degree, and this is discussed in Chapter 25 by Kushi and McManus who focus on the impact of austerity measures on gender. They begin by showing the strong differences in unpaid labour across welfare regimes and other data on social spending, public employment and labour market participation rates that affect gender outcomes. They further show that in several welfare states, women have been left more vulnerable than before the crisis, partly due to the fact that gender issues have not been adequately incorporated into policymaking decisions after the Great Recession in several welfare states. The chapter ends with a discussion of gender and welfare regimes in the context of the coronavirus pandemic. The final Part V comprises Chapter 26 by Greve. It is based upon the analyses within the different chapters and with reference to the recent economic crisis as a consequence of the COVID-19 crisis. It discusses whether this will imply retrenchment and austerity, and if austerity, is it more likely to be partial austerity, given the support in welfare states to have a solid and well-functioning healthcare sector. Naturally, this is a preliminary assessment.

CONCLUDING REMARKS Austerity and retrenchment have been central issues in many publications over the last 10–20 years. Populism has more recently become a core topic for analysis of many and very different societies’ development. Part of this reflects the constant ongoing changes in societies and also that there are winners and losers from the changes. This is not to imply that everything was better in the good old days and that we should return to a previous time. Instead, there is a need to reflect upon what the consequences have been and might be in the future. Hopefully, this research handbook will help to inform about how to interpret and understand the concepts, how the development has been and what we can learn from the development in the last 10–15 years. This does not imply that future crises will have the same consequences. Thus, for example, the ongoing, at the time of writing, COVID-19 crisis and its impact on welfare states is difficult to depict. This is due to several reasons, one being that it is still not possible to know how long it will last, but also that the learning effect from the last financial crisis might imply another way of coping with this crisis. Furthermore, a crisis due to natural circumstances might also be perceived differently by voters than an understanding, at least by some, that a crisis was due to greed among the elites of societies.

REFERENCES Blyth, Mark. 2013. Austerity: The History of a Dangerous Idea. Oxford: Oxford University Press. Cromby, John, and Martin E.H. Willis. 2014. ‘Nudging into Subjectification: Governmentality and Psychometrics.’ Critical Social Policy 34 (2): 241–59. Greve, B. 2019a. The Routledge Handbook of the Welfare State. Edited by B. Greve. 2nd ed. Oxon: Routledge.

Introduction to the Handbook on Austerity, Populism and the Welfare State  9 Greve, B. 2019b. Welfare, Populism and Welfare Chauvinism. Bristol: Policy Press. Greve, B. 2020. Austerity, Retrenchment and the Welfare State. Truth or Fiction? Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Hacker, Jacob S. 2004. ‘Privatizing Risk without Privatizing the Welfare State: The Hidden Politics of Social Policy Retrenchment in the United States.’ American Political Science Review 98 (2): 243–60. https://​doi​.org/​10​.1017/​S0003055404001121. Hobbins, Jennifer. 2016. ‘Young Long-Term Unemployed and the Individualization of Responsibility.’ Nordic Journal of Working Life Studies 6 (2): 43–59. Judd, Dawn, Jurgen Boeck, and Aase Madsen. 2015. ‘Chicken or Egg? Global Economic Crisis or Ideological Retrenchment from Welfare in Three European Countries.’ Critical and Radical Social Work 3 (3): 339–55. https://​doi​.org/​10​.1332/​2049​86015X1439​2797857418. Jurado, Elena, and Grete Brochmann. 2013. Europe’s Immigration Challenge: Reconciling Work, Welfare and Mobility. London: IB Tauris. Palier, Bruno, and Claude Martin. 2007. ‘Editorial Introduction from “a Frozen Landscape” to Structural Reforms: The Sequential Transformation of Bismarckian Welfare Systems.’ Social Policy & Administration 41 (6): 535–54. https://​doi​.org/​10​.1111/​j​.1467​-9515​.2007​.00571​.x. Pierson, C., L. Humpage, P. Fisher, L. Buckner, S. Ronchi, and C. Mills. 2016. ‘Coming Together or Drifting Apart? Income Maintenance in Australia, New Zealand, and the United Kingdom.’ Politics & Policy 44 (2): 261–93. https://​doi​.org/​10​.1111/​polp​.12150. Pierson, P. 1994. Dismantling the Welfare State?: Reagan, Thatcher and the Politics of Retrenchment. Cambridge Studies in Comparative Politics. Cambridge: Cambridge University Press. https://​doi​.org/​ DOI: 10.1017/CBO9780511805288.002. Pierson, Paul. 2001. ‘Coping with Permanent Austerity Welfare State Restructuring in Affluent Democracies.’ In The New Politics of the Welfare State. https://​doi​.org/​10​.1109/​TELSKS​.2011​ .6143211. Trlifajová, Lucie, and Jakob Hurrle. 2019. ‘Work Must Pay: Does It? Precarious Employment and Employment Motivation for Low-income Households.’ Journal of European Social Policy 29 (3): 376–95. https://​doi​.org/​10​.1177/​0958928718805870.

PART II CONCEPTUAL AND METHODOLOGICAL ISSUES

2. What is austerity? Kevin Farnsworth and Zoë Irving

INTRODUCTION Although it is an idea that has defined the 2010s, there is nothing new about austerity. Historically, times of austerity have played as much of a role in shaping contemporary welfare states as times of prosperity. Welfare states have expanded and contracted, and periodic self-imposed cuts in welfare spending and welfare ‘effort’ represent an ongoing struggle waged along ideological lines. Nevertheless, the present period that began with the 2008 economic crisis has cemented austerity in the contemporary political economy, first as a descriptor for an ‘essential’ but temporary adjustment in welfare spending, but second, and most importantly, as a political project aimed at transforming the welfare state and realising ambitions with much longer roots. Twenty-first-century austerity – or neo-austerity – represents the latest iteration of the ongoing struggle between politics and markets. Austerity occupies a particular place in political economy that operates at many levels and in different ways in this contemporary struggle. Austerity is also a slippery idea (Farnsworth and Irving, 2015; Greve, 2020). It draws together particular views of the economy (the impact of public spending and borrowing on private sector investment and growth), with views of public finance (the need to balance expenditure with taxation receipts) and ideological viewpoints (to shrink the state) with views of society (that locate austerity politically and sociologically in contemporary neoliberalism and the global reconfiguration of state–citizen relations). Austerity is also about power and how it is wielded, by powerful individuals, national and international institutions and governments. Thus, what is often portrayed as a relatively simple idea – a voluntary national contraction in expenditure (cf. Blyth, 2013) – is more complex than this. Much of the debate about austerity therefore reflects disagreement about what it is, how the concept (or conceptions) of austerity can be defined and how it can then be identified, measured, compared and evaluated. The argument presented here is that austerity is determined by politics as much as it is by economics. Austerity is both a political (ideological) project and an economic one, driven by international and world regional bodies as well as national governments. It is about the unwillingness or inability of citizens and policy makers to defend the welfare state in the face of ongoing economic, political and social crises. And it is about the underlying structural weaknesses of contemporary capitalist economies, the delegitimisation of government, the ineffectiveness of governance and the fact that successive reforms have both steered social policies in less progressive directions and have rendered social policy less, rather than more, compatible with welfare states. The precise impact on developed welfare states is, as argued elsewhere (Farnsworth and Irving, 2012, 2015), uneven and variable, but most have encountered austerity in one form or another. Some welfare states are being retrenched, others are undergoing a process of restructuring, and still others are experiencing wholesale reconfiguration. This chapter begins by exploring the notion of austerity and then goes on to consider how it is identified and subsequently measured. The chapter ultimately presents the argument that 11

12  Handbook on austerity, populism and the welfare state austerity has unnecessarily delegitimised the political welfare state, undermined its positive economic effects and will leave societal scars that will last even if public spending commitments are restored.

DEFINING AND MEASURING AUSTERITY Given the above, austerity is much more than a simple indicator of how pared back or expansive welfare states are during a particular period of time. Time is itself a dimension of austerity’s contested definition because how it is defined determines the extent to which it can be identified historically, and during which specific periods (Blyth, 2013; Hill, 2015). As it has been most often applied to the analysis of nation states, austerity is detached from its political context and presented as an economic ‘truth’. In economic terms it is understood as a means to reduce the structural deficit – the difference between revenues and expenditure that are not expected in the standard economic cycle. Mark Blyth (2013) extends this by defining it as a ‘voluntary correction’. This latter conceptualisation is, in fact, incorrect since some states have austerity imposed on them (Greece from 2010 would be a good example here). In such cases, power is exercised and ‘choices’ are imposed by others, including creditor organisations. The value of Blyth’s definition, however, is that it is useful in pointing to austerity as an exercise in politics rather than economics. There are two triggers for austerity. On the one hand, it represents active, endogenous political efforts to reduce the state at the national level. On the other, it represents the outcome of exogenous forces that are exerted on governments which, in turn, implement involuntary cuts to their national welfare states. The economic orthodoxy suggests that austerity has to prevail because ‘unsustainable’ public debt levels require such a response, despite the lack of objective and robust measures that might determine when/how debt is too high. Although austerity is sold on the basis that there is no alternative, a range of economic choices are available, especially to countries with sovereign currencies. The extensive use of QE (quantitative easing, or printing money) by a number of countries and, eventually, the European Union, is just one example. All of this suggests that the politics of austerity matters far more than the economics. Voluntary corrections do not always take the same form, target the same areas of spending or apply to all areas of spending at the same time. These differences tend to make the empirical reality of austerity quite murky and, where spending is the indicator, contested. There is much scholarship within the fields of public and social policy observing that the politics of welfare has enabled welfare states to stand the test of time (van Dyke, 2003) and that they are consequently remarkably resilient in the face of internal and external economic pressures. Van Kersbergen et al. (2014), for example, present a direct challenge to the idea that, post-2008, austerity was the only game in town (see also Armingeon, 2013). As some budgets were increasing and since overall public expenditure in most nations had actually increased during the so-called period of austerity, van Kersbergen et al. (2014) argue that austerity, by definition, had not occurred. Focusing on particular cuts associated with austerity and their easing over time, others have also indicated at various points since 2010, when austerity measures replaced the emergency stimulus packages introduced in many countries, that austerity is ‘over’. Eurofound (2016, p. 10) for example, indicated this was the case as job losses in public administration in 2013–15 slowed, and by 2020, even in the UK where austerity had been central to public

What is austerity?  13 discourse on the welfare state, the ‘end of austerity’ had been signalled in budget statements in both 2018 and 2019. In Europe too, the turning on of spending taps was regarded as a sign that austerity was over (Financial Times, 2019). It is clear from these debates that although austerity is a loose concept, used in different ways to describe different ends, it is most commonly regarded as an economic instrument that can be identified and applied through adjustments in spending and regulatory rules (including on benefits). The following section presents data on movements in expenditure in order to locate which countries have cut expenditure and which areas of expenditure have been cut the deepest.

MEASURING ECONOMIC AUSTERITY It is well established in comparative research that expenditure data often hides more than it reveals, and that the true value of proportions of gross domestic product (GDP) spent are disguised by changes in GDP. Nevertheless, the resilience argument does appear to be supported by spending data. Data compiled by the IMF and other international organisations on actual and projected expenditure suggests that it has remained remarkably buoyant over time. Figure 2.1 presents reported expenditure as a percentage of GDP between 2006 and 2019, and planned expenditure in 2020 to 2024, in selected high-income countries. In these countries, expenditure increases in the immediate aftermath of the crisis, before falling back to 2019. Figure 2.1 shows that in most of these countries, expenditure levels off from 2020, although cuts are still planned in Ireland, Greece and, more modestly, France. Most countries, with the

Source: IMF Fiscal Monitor.

Figure 2.1

Public expenditure: actual (2006–19) and planned (2020–24)

14  Handbook on austerity, populism and the welfare state

Figure 2.2

Expanding states, 2007–17

exception of Ireland, Sweden and the UK, are planning to spend a higher proportion of GDP in 2024 than they were spending in 2006. Comparing expenditure levels before the crisis hit (2007), in the midst of austerity (2012), and in the most recent available year (2017) the shifts appear more considerable (Figures 2.2 and 2.3). Taken together, these figures show that all states increased expenditure 2007–12, but not all states subsequently reduced expenditure to their pre-crisis levels. The countries spending more at the end of the period than at the start – the expanding states – are mapped in Figure 2.2, while those spending less at the end of the period than at the beginning – the austerity states – are mapped in Figure 2.3. All the countries included – made up of OECD (Organisation for Economic Co-operation and Development) nations where data across the periods is available – report increases in expenditure in 2012 compared with 2007. But some nations are marked by subsequent deep cuts in order to revert back to pre-crisis expenditure levels, in particular Ireland, Greece, the Czech Republic and the UK. Despite such shifts in expenditure throughout this period, however, many ended the period recording higher

What is austerity?  15

Figure 2.3

Austerity states, 2007–17

expenditure in 2017 compared with 2007. In sum, these figures illustrate the complexity of determining whether austerity has taken place based on expenditure data alone. To move beyond a simple comparison of expenditure and achieve a more nuanced illustration of the presence or absence of economic austerity, it is better to compare across additional dimensions. Taken together, four variables provide an indication of the strength of austerity within states. First, pre-crisis expenditure as average expenditure, as a percentage of GDP, over several years in the 2004–07 period would provide some indication of settled expenditure in ‘normal’ times (variable A). Second, a measure of how expenditure changes between 2008 and 2011 is required as this captures the immediate and sustained impact on public spending of the banking crisis and early economic slowdown (variable B). Third, the level of movement in expenditure during the 2012–16 period captures the period of austerity (variable C). Fourth, expenditure across the most recent three years, 2017–19, provides an indication of where public expenditure has settled compared with pre-crisis levels (variable D). The

16  Handbook on austerity, populism and the welfare state results are presented in Figure 2.4.1 All countries with the full range of available relevant data are included. Negative scores illustrate that some degree of expenditure austerity has been imposed on nations. At the extremes, Ireland experienced the highest level of austerity, while Switzerland appears not to have experienced austerity at all. Pre-crisis, the expenditure level in Ireland was 32.5 per cent GDP. The banking crisis pushed expenditure up to over 50 per cent GDP, and the more recent figures suggest that public expenditure in Ireland is now at 26 per cent of GDP. Switzerland in contrast, had pre-crisis expenditure of 34 per cent GDP which fell during the banking crisis to a little over 32 per cent GDP. The 2017–19 average places public expenditure in Switzerland at 34 per cent GDP, identical to the pre-crisis level. This figure suggests that nearly all the states considered undertook economic austerity during the period and that, perhaps predictably given their specific economic exposure to the banking and sovereign debt crises, Ireland, Iceland and Greece are the most austere, followed closely by the UK and New Zealand.

Source: Author calculations based on IMF and OECD data.

Figure 2.4

Austerity scores

Despite signs of varying actual austerity, expenditure plans in most OECD countries are projected to converge rather than diverge. Differentials in public expenditure in 1995 broadly fit with welfare or particular variety of capitalist or welfare regime type (Esping-Andersen, 1990; Hall and Soskice, 2001), but by 2024, the differences in expenditure are projected to become smaller within a downward trend (Figure 2.5). Expenditure across all economies is projected to fall, except for the lowest spending comparator economy here, Japan. Expenditure

What is austerity?  17 in the highest spending economies in 1995 is projected to fall fastest. This suggests that economic austerity may be reducing some of the wider expenditure differentials between welfare states, although almost 30 GDP percentage points will still separate the most from the least well-funded welfare states. The implications for comparative studies of welfare states, and especially for those who study welfare regimes and/or families of nations, are potentially very significant.

Source: IMF Fiscal Monitor and OECD data.

Figure 2.5

Public expenditure as a percentage of GDP 1995 and projected 2024

Although expenditure levels are important, as noted earlier it is the way in which spending is distributed – the regime type – that matters in achieving welfare goals. This is as much about how as how much the state raises taxes and spends revenues. Cuts in one part of the welfare state cannot be compensated for by large increases in another part. Increasing expenditure on defence would not compensate citizens for equivalent cuts in public health services. This level of detail can really only be examined with reference to the broad political economy of nation states, and much of the work undertaken analysing the post-crisis landscape has indicated that austerity has not been adopted as a ‘one size fits all’ strategy.

18  Handbook on austerity, populism and the welfare state

VARIETIES OF AUSTERITIES In early reflections on welfare state change since 2008, it has been argued (Farnsworth and Irving, 2011, 2012) that there are varieties of austerity that are linked directly to the variety of crisis experienced by particular states (depending on financial exposure, strength of export markets, position within the Eurozone for example). This variation is apparent in both the combinations and timing of stimulus/austerity measures in individual countries, the outcomes of national elections in the early years, and the extent to which a fundamental belief in cutting their way to growth was a feature of national government thinking (see, for example, Bermeo and Pontusson, 2012; Starke et al., 2013; Kahler and Lake, 2013; Schäfer and Streeck, 2013). The patterns of austerity within the OECD countries have not necessarily followed the traditional welfare regime models as countries such as Denmark have diverged from the principles of generosity and citizenship that characterise ‘social democratic’ regimes, while Australia has not fully adhered to what might have been expected in the ‘liberal’ model which is characterised by lower expenditure and less generous welfare provision (Esping-Andersen, 1990). This has raised further questions about the predictive capacity of regime theory, particularly in an age where the power and influence of international organisations and the relevance of global social policy has broadly strengthened, and the operation of national social and economic policy has been reshaped. The welfare arrangements in emerging and low-income economies have also become more visible and subject to greater analysis, giving a much broader canvas on which patterns can be discerned. Despite the proposition that austerity may not be a significant feature of all European or OECD states, a study by Ortiz and Cummins (2021) demonstrates not only that austerity measures can take many forms other than direct cuts, but also that it is in fact a significant feature of welfare policy in the Global South. Analysing hundreds of International Monetary Fund (IMF) country reports and spending projections for 183 countries for the period 2005–20, the authors argue that the period 2016–20 represents a second, and deeper phase of austerian budget cuts, which by 2020 had impacted on the lives of 83 per cent of populations in developing countries and 61 per cent of those in high-income countries. The austerity measures identified further in this analysis are not restricted to budget cuts alone (for example, the reduction of the public sector wage bill, healthcare and pension reforms, and the contraction of social protection), and it considers the added effects of the reduction or elimination of subsidies (on food and fuel, for example), labour market flexibilisation, the increase of consumption taxes and the privatisation of public assets. Many of these measures and reforms to existing provision have a traceable history in the social policy advice of international organisations such as the World Bank (WB) and the IMF, highlighting the heritage of austerity within the practice of ‘structural adjustment’. This raises a further question relevant to the notion of austerity, and its identification and measurement – the role of ideas in shaping and maintaining hegemony, and how these feed into public understandings and expectations of the state–citizen relationship and the guarantees, rights and obligations that follow on from this. The role of international organisations in shaping ideas in social policy has become an important avenue of study in the 2000s (Broome and Seabrooke, 2007; Chwieroth, 2008; Clift and Tomlinson, 2012; Belande and Orenstein, 2013; Ban, 2015; Moschella, 2015). The changing of welfare relations at the level of ideas and practices is a dimension that is less easy to test empirically than changes in spending patterns, but no less important because of this. While both the WB and the IMF have made efforts to

What is austerity?  19 distance the organisations from the idea of ‘structural adjustment’ since the early 2000s, it should not be assumed that this signifies a reversal of policy preferences or a post-2008 shift to a new economic paradigm. The IMF in particular has shown some inconsistency in its support for and advice against the pursuit of austerity as a route to growth. The IMF has explicitly recognised that rising inequality as an outcome of the financial crisis and subsequent policy responses to it is a threat to both stability and growth. And its staff have pointed to the way in which privatisations and the running-down of state assets has undermined the sustainability of borrowing (Ostry et al., 2016). However, this recognition must be contextualised within the organisational imperatives of legitimacy and survival; greater nuance in approaches to the global economy versus the specific actions perceived to be required within individual countries; and the capacity of organisations to assimilate dissonant emergent thinking into existing paradigmatic frameworks (Ban, 2015; Farnsworth and Irving, 2018). These frameworks are obviously part of the existing global economic system which has also survived yet another economic crisis with relatively little systemic damage. Given the indisputable negative effects of austerity – even in its simplest economic terms, there may have been some expectation that it would meet substantial political resistance – the immovable objects identified by Paul Pierson (1998), but this has not been the case. The explanation for the success of an objectively bad, and demonstrably failed idea (Stuckler and Basu, 2013; Schui, 2014; Seymour, 2014) can be found in the politics of austerity.

THE POLITICS OF AUSTERITY In considering the politics of austerity and how these have shaped ‘what it is’, it is necessary to look further back in time to the last economic crisis and the point where confidence in the post-war social settlement began to erode. The essence of the 1970s economic crisis (high inflation linked to increases in oil prices, falling economic demand, rising unemployment) was somewhat different to the drivers in 2008, but the debate surrounding the role of the welfare state in precipitating and proliferating the crisis was similar during both periods. Proponents of an austerian response were able to present a convincing narrative that despite the culpability of the financial sector in catalysing the crisis, it was the concentration of national wealth invested in the unproductive state rather than the productive private sector that was the real culprit (cf. Bacon and Eltis, 1976; Gough, 1979). Post-2008, this led to widespread calls for a rebalancing of economies, with greater emphasis on the private (wealth creating, industrial) sector and a politically successful message that public spending had to be curtailed. The earlier pathologisation of the welfare state that took hold in the 1980s came to define the periods of Thatcherism and Reaganomics. It enabled a punitive social politics of ‘welfare’ to become entrenched in the already residualised US system of social protection. In the UK particularly, but also filtering out to the other OECD countries and the European Union, it allowed the focus of social policy to be narrowed to the individual incentives required of social provision – independence from state support and flexibility in labour supply, for example. This period also saw the ramping up of market solutions to policy problems, deregulation and efficiency drives and softened the ground for the further excavation of the foundations of welfare states that has occurred since 2008. Even though by the mid-2010s debt in most countries remained at lower levels than experienced in the past, political actors were able to convince

20  Handbook on austerity, populism and the welfare state sufficiently large numbers amongst electorates that social policies helped to cause the rise in debt, and that increasing levels of debt mean that social policy will be unaffordable in future. This narrative on the welfare state as ‘villain’ in the economic crisis story drew on both international organisation advice and academic studies. Reinhart and Rogoff’s (2010) work on the problem of government indebtedness found favour amongst austerity advocates and echoed the IMF stance that public borrowing above a certain level is unsustainable because it begins to have a negative impact on growth. For Reinhart and Rogoff, the ‘cut-off’ is 90 per cent of GDP, but in its 2009 Fiscal Monitor, the IMF bases most of its calculations on sustainable debt to GDP ratios set at 60 per cent GDP (IMF, 2009). Both measures are almost arbitrary. The counter-arguments to the success of austerity measures in securing economic growth were routinely ignored (Krugman, 2012; Wren Lewis, 2018), as was the immediate and long-term public and social cost of debt-determined welfare state reconfiguration. There is ample economic and historical evidence to support the claim that austerity does not result in economic growth (Schui, 2014), and certainly not the kind of growth that is compatible with welfare goals and social justice. Senior IMF economists have more recently acknowledged that neoliberal policies, including privatisation and austerity, helped to drive the economic crisis, increased inequality, reduced growth and contribution to national fiscal fragility (Ostry et al., 2016). Particularly where austerity is ‘discretionary’ in the sense that it is chosen rather than enforced by economic conditions, it is argued to be deflationary and therefore depress growth (Guajardo et al., 2014). There is a simple explanation for why this is the case. In discretionary austerity situations, politicians respond to past policy decisions and actions rather than current and/or prospective economic challenges. In such cases, for example the UK, the economic failings of austerity are less relevant than the deeper policy ambitions. Here the ‘problem’ that austerity promised to solve was not the deficit nor even economic slowdown, but the welfare state itself. Austerity does not just describe economic policy instruments aimed at reducing expenditure. It also describes an approach to public services, and the role of the state. As with most policy ideas, the idea of austerity required careful framing to emphasise its necessity, and a nuanced political packaging of its constituent measures to render it palatable to national electorates. In particular, a key element of framing austerity has been its dissociation from ‘ideology’. In the UK, David Cameron made this explicit in 2010, stating that ‘We’re tackling the deficit because we have to – not out of some ideological zeal. This is a government led by people with a practical desire to sort out this country’s problems, not by ideology,’2 but by 2013, once public acceptance of austerity had been established, was able to argue the merits of the resultant ‘leaner state’. Austerity discourse has built upon the visions of the economy that were promoted during the 1980s and this powerful imagery was reused by austerity governments from 2010 with reference to the burden of debt, and the threat of economic collapse. Although austerity’s advocates have denied its connection with ideology, the processes at work have clearly been ideational at both state and supranational level. Fiscal consolidation, while framed as an economic necessity, has also represented an effective tool to deliver the more residualised welfare model long preferred by market liberals. Although post-2008 crisis austerity has fitted well within neoliberalism’s privileging of markets and private sector solutions, the ‘age of austerity’ has starkly revealed the inconsistencies of ‘neoliberalism’ as an economic and political project. First, that although an expansive public sector makes demands upon the state, the private sector also makes demands upon the state that reach well beyond the minimalist regulation presented in market liberal thinking

What is austerity?  21 (Farnsworth, 2012). Second, that under austerity, the wider and long-term social and political consequences of protecting economic elites at the cost of rising inequalities have been vastly underestimated. Regardless of their social goals, welfare states have served an essential economic function – enabling stability in challenging times and ensuring that living and working conditions are sufficient for economies to thrive and compete. The social goals have always been terrain that is politically determined, but the logic of austerity also poses a systemic threat. The reduction of social policy to its productivist uses is a defining feature of what we have termed neo-austerity (Farnsworth and Irving, 2018) which is more about capitalising on crisis in order to shrink the state than it is about cutting expenditure in order to resolve economic crisis. Part of the success of this new form of austerity has been its capacity to overcome a key contradiction in neoliberal thinking on the activity of the state vis-à-vis the market. Contrary to the idea that only a ‘small state’ can provide markets with the freedom they need, markets do need states to provide the regulatory frameworks in which they operate, and in modern welfare states these systems are both extensive and expensive. Neo-austerity has enabled states to impose constraints on public spending while at the same time reconfiguring social spending to target support to measures that promote private sector interests. While ‘social investment’ in its original sense – the investment in human capital, for example – has itself been the subject of austerity cuts, the demand and supply for state support for private business has not been subject to the same economising (IMF, 2016, p. 36). Post-2008 austerity is neither the post-Second World War acceptance of time-limited privation as a trade-off for ‘jam tomorrow’ (see Hill, 2015), nor is it the ‘permanent austerity’ (Pierson, 1998) of the 1980s which was an ideological attack on the state. The ‘neo-austerity’ of the 2010s was presented as a non-ideological, ‘no-alternative’ solution to a set of economic circumstances that were borne out of a global financial crisis. It was constructed through technical economic arguments that masked a reinvigorated ideological reframing of the social settlement which succeeded in changing the narrative of welfare state expansion from one which prioritised social ‘improvement’ to one which further narrowed the terms of social ‘investment’. The limits on neo-austerity therefore are far less obvious since it undermines many of the pre-existing avenues of resistance to welfare retrenchment. The welfare reform measures introduced from 2010 have engineered greater divisions between social groups – private sector versus public sector workers, home-owners and those who rent, for example, intergenerational divisions – those of retirement age versus those entering the labour market and increased divisions between those who rely materially on state support and those whose reliance is hidden in services in kind. The neo-austerian model is one where the bonds of solidarity that characterised post-war welfare state building and endured through the 1980s to 2000s are dissolved. In this way austerity has a pernicious effect on socially integrative function of social policy. This is experienced both directly through increasingly residualised public provision and indirectly through the absence of recognition of or concern for the need for the ‘welfare commons’ (Williams, 2015). With this absence, the political demand for more and better welfare is also diminished and a brake is applied to any further extension or enhancement of rights and guarantees. Reducing capacity for welfare state renewal is the most significant achievement of neo-austerity. This has been achieved not only through changing the narrative and sowing deeper divisions but also through the more traditional means of mobilising demands for welfare. The power of labour, measured through union density, had been declining prior to

22  Handbook on austerity, populism and the welfare state the 2008 crisis, but in the social context of recession, falling incomes, job and housing insecurity both the voice and exit options for workers are affected. Intra-class solidarities amongst precarious workers are limited by atomised employment patterns and austerity measures that seek to further flexibilise workers and block pathways towards the reconstitution of labour power (McBride and Mitrea, 2017). These measures are evidenced in lowered standards of compensation for labour market losses, the rising incidence of in-work poverty and the focus on ‘supply-side reform’ (see Hastings and Heyes, 2018). Supply-side reform is largely seen in employees adjusting to the reality of ‘marginal’ part-time employment (Messenger and Wallot, 2015) with all the long-term labour market disadvantages that are consequently accrued. Class solidarities too are at risk. The OECD (2019) has reported that middle class economic power has weakened as the highest income decile has gained since 2010 while those on median incomes have seen little change. A weakening middle class is associated with a weakening of democracy, but also has consequences for the strength of support for public welfare, through the coalescence of class interests and the extent of class investment in the welfare state. The twenty-first-century politics of austerity therefore look very different to those of previous periods of contraction. What Hemerijck (2012) has termed the ‘stress-testing’ of welfare states has failed to deliver the heralded economic growth and has contributed to a legitimised lack of restraint for capital, political instability in previously stable democracies and in many, a populist rejection of globalised markets from which there is more limited political exit (Schäfer and Streeck, 2013).

CONCLUSIONS This chapter has explored the ways in which austerity has been represented and the contexts within which these representations have operated. It has demonstrated that taking a simple economic approach to the definition and measurement of austerity based on spending alone provides some of the picture, but that this partial representation raises questions for existing theories of welfare that require consideration of both supranational factors and an analysis of politics that goes beyond quantitative assessment. The political economy of austerity is more than the sum of varying austerity measures in individual nation states. Austerity remains a theoretically and empirically contested notion, and one for which there are both more and less optimistic prognoses (Gamble, 2016; Streeck, 2017). These analyses share an assessment of an increasingly untenable pact between capitalism and democracy where the effects of a decade of welfare state depletion have contributed to significant social harm. As the data presented in the first section of this chapter indicates, many governments appear to have bypassed or at least curtailed austerity measures but as the discussion of the politics of austerity has argued, this does not imply that they have also avoided austerity. From this it can be concluded that even if the ‘end of austerity’ were to be declared, the reparation required for the social, political and economic damage done would stretch beyond a simple increase in spending. Public services have been sold-off or outsourced. Expenditure cuts have undermined investment in services which is undermining their effectiveness. Benefit claimants have been castigated. Immigrants have been blamed for unsustainable expenditure. Public confidence in public services has taken a knock. The welfare state may yet prove to be far less resilient than past scholars assumed.

What is austerity?  23

NOTES 1. 2.

These four variables have been used to create a proxy for austerity (A–B)+(C–B)+(D–C)+(D–A). https://​www​.theguardian​.com/​politics/​2013/​nov/​11/​david​-cameron​-policy​-shift​-leaner​-efficient​ -state (accessed 4 January 2021).

REFERENCES Armingeon, K. (2013) Breaking with the past? Why the Global Financial crisis led to austerity policies but not to the modernization of the welfare state. In C. Pierson, F.G. Castles, and I. K. Naumann (eds), The Welfare State Reader, 3rd edn. London: Blackwell, pp. 214–26. Bacon, R. and Eltis, W. (1976) Britain’s Economic Problem: Too Few Producers. London: Macmillan. Ban, C. (2015) Austerity versus Stimulus? Understanding fiscal policy change at the International Monetary Fund since the Great Recession. Governance 28(2): 167–83. Belande, D. and Orenstein, M. (2013) International organizations as policy actors: An ideational approach. Global Social Policy 13(2): 125–43. Bermeo, N. and Pontusson, J. (eds) (2012) Coping with Crisis, Government Reactions to the Great Recession. New York: Russell Sage Foundation. Blyth, M. (2013) Austerity, the History of a Dangerous Idea. Oxford: Oxford University Press. Broome, A. and Seabrooke, L. (2007) Seeing like the IMF: Institutional change in small open economies. Review of International Political Economy 14(4): 576–601. Chwieroth, J. (2008) Normative change from within: The International Monetary Fund’s approach to capital account liberalization. International Studies Quarterly 52: 129–58. Clift, B. and Tomlinson, J. (2012) When rules started to rule: The IMF, neo-liberal economic ideas and economic policy change in Britain. Review of International Political Economy 19(3): 477–500. Esping-Andersen, G. (1990) The Three Worlds of Welfare Capitalism. Cambridge: Polity Press. Eurofound (2016) Eurofound Yearbook 2015: Living and Working in Europe. Dublin: European Foundation for the Improvement of Living and Working Conditions. Farnsworth, K. (2012) Social versus Corporate Welfare, Competing Needs and Interests within the Welfare State. Basingstoke: Palgrave Macmillan. Farnsworth, K. and Irving, Z. (eds) (2011) Social Policy in Challenging Times: Economic Crisis and Welfare Systems. Bristol: Policy Press. Farnsworth, K. and Irving, Z. (2012) Varieties of crisis, varieties of austerity: Social policy in challenging times. Journal of Poverty and Social Justice 20(2): 135–49. Farnsworth, K. and Irving, Z. (2015) Austerity: More than the sum of its parts. In K. Farnsworth and Z. Irving (eds), Social Policy in Times of Austerity: Global Economic Crisis and the New Politics of Welfare. Bristol: Policy Press, pp. 9–41. Farnsworth, K. and Irving, Z. (2018) Austerity: Neoliberal dreams come true? Critical Social Policy 38(3): 461–81. Financial Times (2019, 10 February) Europe turns on spending taps as austerity comes to an end. Gamble, A. (2016) Can the Welfare State Survive? Cambridge: Polity Press. Gough, I. (1979) The Political Economy of the Welfare State. Basingstoke: Macmillan. Greve, B. (2020) Austerity, Retrenchment and the Welfare State: Truth or Fiction? Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Guajardo, J., Leigh, D., and Pescatori, A. (2014) Expansionary austerity? International evidence. Journal of the European Economic Association 12(4): 949–68. Hall, P.A. and Soskice, D. (eds) (2001) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage. Oxford: Oxford University Press. Hastings, T. and Heyes, J. (2018) Farewell to flexicurity? Austerity and labour policies in the European Union. Economic and Industrial Democracy 39(3): 458–80. Hemerijck, A. (2012) Changing Welfare States. Oxford: Oxford University Press.

24  Handbook on austerity, populism and the welfare state Hill, M. (2015) Conventional wisdom on government austerity: UK politics since the 1920s. In K. Farnsworth and Z. Irving (eds), Social Policy in Times of Austerity: Global Economic Crisis and the New Politics of Welfare. Bristol: Policy Press, pp. 43–66. IMF (2009) Fiscal Monitor. Washington, DC: IMF. IMF (2016, April) Acting Now, Acting Together. Fiscal Monitor. Washington, DC: International Monetary Fund. Kahler, M. and Lake, D. (2013) Politics in the New Hard Times. Ithaca, NY: Cornell University Press. Krugman, P. (2012) End this Depresssion Now! New York: W.W. Norton and Co. McBride, S. and Mitrea, S. (2017) Internalizing neoliberalism and austerity. In S. McBride and B. Evans (eds), The Austerity State. Toronto: University of Toronto Press. Messenger, J. and Wallot, P. (2015, May) The diversity of ‘marginal’ part-time employment. INWORK Policy Brief No.7. Geneva: International Labour Office. Moschella, M. (2015) The institutional roots of incremental ideational change: The IMF and capital controls after global financial crisis. British Journal of Politics and International Relations 17: 442–60. OECD (2019) Under Pressure: The Squeezed Middle Class. Paris: OECD Publishing. Ortiz, I. and Cummins, M. (2021) The austerity decade 2010–20. Social Policy and Society 20(1): 142–57. Ostry, J., Loungani, P., and Furceri, D. (2016, June) Neoliberalism oversold? Finance and Development. Washington, DC: International Monetary Fund, pp. 38–41. Pierson, P. (1998) Irresistable forces, immovable objects: Post-industrial welfare states confront permanent austerity. Journal of European Public Policy 5(4): 539–60. Reinhart, C.M. and Rogoff, K.S. (2010) Growth in a time of debt. American Economic Review 100(2): 573–8. Schäfer, A. and Streeck, W. (2013) Politics in the Age of Austerity. Cambridge: Polity Press. Schui, F. (2014) Austerity, The Great Failure. New Haven, CT: Yale University Press. Seymour, R. (2014) Against Austerity, How We Can Fix the Crisis They Made. London: Pluto Press. Starke, P., Kaasch, A., and Van Hooren, F. (2013) The Welfare State as Crisis Manager: Explaining the Diversity of Policy Responses to Economic Crisis. Basingstoke: Palgrave Macmillan. Streeck, W. (2017) How Will Capitalism End? London: Verso. Stuckler, D. and Basu, S. (2013) The Body Economic, Eight Experiments in Economic Recovery from Iceland to Greece. London: Penguin Books. van Dyke, L. (2003) Milking the system – how corporate welfare fleeces the taxpayer, destroys the environment and corrupts public policy (Australia). Overland 170, 110. Van Kersbergen, K., Vis, B., and Hemerijck, A. (2014) The great recession and welfare state reform: Is retrenchment really the only game left in town? Social Policy & Administration, 48(7): 883–904. Williams, F. (2015) Towards the welfare commons: Contestation, critique 30 and criticality in social policy. In Z. Irving, M. Fenger, and J. Hudson (eds), Social Policy Review 27: Analysis and Debate in Social Policy. Bristol: Policy Press, p. 31. Wren-Lewis, S. (2018) The Lies We Were Told, Politics, Economics, Austerity and Brexit. Bristol: Bristol University Press.

3. The politics of retrenchment Peter Starke

INTRODUCTION Welfare state retrenchment has been an object of research in political science for over 25 years now. The defining text remains Paul Pierson’s book Dismantling the Welfare State? Reagan, Thatcher, and the Politics of Retrenchment (1994).1 This is not to say that other approaches, including power resources theory, have not contributed to the study of cutbacks (see Korpi and Palme 2003). Yet in terms of the depth of understanding of specific political dynamics surrounding retrenchment and the influence on the whole field, Dismantling remains the key reference. Each of its different causal mechanisms and sub-mechanisms has become subject to empirical testing and theoretical refinement in subsequent years and, as will be shown, most have withstood scrutiny, even though certain blind spots remain in the literature. The book’s starting point is the puzzle that, despite the economic crisis and a perceived mandate to roll back the postwar intervention state, it seemed that Reagan and Thatcher had failed when it came to the welfare state. Systems of social protection had survived the attack from the political right relatively unscathed. This is a puzzle especially from the perspective of power resources theory which would expect the victory of conservative parties and the weakening of trade unions to have a sizeable effect on the welfare state.2 Pierson tried to solve the puzzle by conceptualizing retrenchment as a genuinely political phenomenon within the setting of advanced democratic welfare states, or what he called the ‘new politics of the welfare state.’ The origins of the retrenchment debate can be found in the context of the early to mid-1990s, a watershed moment for the discussion about the welfare state. The world was slowly recovering from an economic crisis that had hit some advanced welfare states – notably archetypal Sweden – particularly hard. The left was out of office in most wealthy countries at the time, notably all the big OECD (Organisation for Economic Co-operation and Development) countries (except for Japan). The end of the Cold War, the beginning of the globalization debate and deepening European integration cast doubt on the sustainability of large welfare states too. Meanwhile, in academia, comparative welfare state research was flourishing, kickstarted by Esping-Andersen’s similarly influential Three Worlds of Welfare Capitalism (1990). Amid this moment of heightened attention and uncertainty came Pierson, making a counterintuitive case for stability and for taking the domestic political defense of large welfare states seriously. Subsequent writing has often boiled Dismantling down to two highly misleading messages: first, partisan politics (sometimes called ‘old politics’) do not matter anymore and, second, welfare state retrenchment is nearly impossible. I hope it will become clear in this chapter that these are the wrong conclusions to take away from the book (see also Jensen et al. 2019). In what follows I will rather focus on the causal mechanisms about the politics of cutbacks and ask how well Pierson’s original theory of policymaking has stood up to empirical study. The chapter is structured as follows. I begin with an account of the concept of welfare state retrenchment in the comparative social science literature in the aftermath of Pierson. 25

26  Handbook on austerity, populism and the welfare state Then I reconstruct the theoretical explanation, especially in terms of the causal mechanisms used. Based on this, I point to elements of the explanation that have been empirically tested, challenged or extended by subsequent research, including electoral punishment, the role of unconventional political participation, interest groups politics and blame avoidance. The final subsection sums up and concludes.

THE CONCEPT OF RETRENCHMENT There is no universally accepted definition of the concept welfare state retrenchment. Even Pierson does not offer a definition, even though he discusses two important sub-types of retrenchment. It becomes clear though that he uses ‘welfare cutbacks’ (p. 14) and retrenchment synonymously and regards mere social spending reductions as an inadequate measure of retrenchment (see chapters 5 and 6 on measurement). In line with this, I defined welfare state retrenchment narrowly as a ‘political decision to reduce the level of social protection guaranteed by the state’ (Starke 2008: 13). In other words, retrenchment is a conscious policy decision to cut back cash benefits and services guaranteed – but not necessarily directly provided – by the state. It is about individual social rights or entitlements rather than aggregate spending (which always varies with the social and economic problem cycle). Typically, policymakers reduce benefit levels, their duration or tighten eligibility by adding conditionalities or changing qualification age for benefits. Only rarely are programs abolished wholesale. Note that the material consequences of the retrenchment decision can lie far in the future – which can be relevant if policymakers want to avoid an immediate backlash in popularity. Most of the literature on welfare state retrenchment is based on this narrow conceptualization of (programmatic) retrenchment. There are at least two kinds of retrenchment that this narrow conceptualization – or what Pierson calls ‘programmatic retrenchment’ (1994: 15) – does not capture, however:3 first, systemic retrenchment (Pierson 1994: 13–17). Systemic retrenchment refers to policy changes that, while leaving social welfare programs intact in the short run, weaken their funding arrangements, manipulate public opinion and weaken support groups or key features of the institutional environment that form the basis of the welfare state or individual schemes. If they can overcome the institutional path dependence of these support systems of the welfare state (and that can be a big if), retrenchment advocates might be quite successful in the long run. Klitgaard and Elmelund-Præstekær (2014) show how right-wing governments in Denmark since the mid-1970s used tax cuts strategically to limit social spending (see also Elmelund-Præstekær and Klitgaard 2012). In related research, Klitgaard et al. (2015) investigate the likelihood of institutional welfare state reforms across Sweden, Denmark, Spain and the Netherlands in the period between 1982 and 2011. These reforms, which may include changes in the participation of trade unions in the administration of social insurance funds, are often ‘inconsequential for voters in the short run’ but may have large repercussions on the long-term trajectory of social provision. Overall, systemic retrenchment has become far less influential than programmatic retrenchment as a concept (but see Lo Vuolo 1997; Taylor‐Gooby, 1999; Marier, 2016). Perhaps this has to do with the difficulty of judging future probabilities of retrenchment given contemporary policy changes and whether these can really be seen as intentional acts of retrenchment. Yet, because the existing studies all indicate that systemic retrenchment is widespread, less dangerous from an electoral perspective and subject

The politics of retrenchment  27 to stronger partisan bias than programmatic retrenchment, systemic retrenchment certainly warrants more systematic research. The second kind of welfare state decline not covered by programmatic retrenchment is policy drift (Streeck and Thelen 2005). In an early critique of Pierson, Clayton and Pontusson pointed out that measures of retrenchment used by Pierson tended to be insensitive to a changing social environment. In other words, in times of increasing needs or even just economic growth, it is sufficient for retrenchment advocates to do nothing in order to achieve their goal of shrinking the relative importance of welfare benefits. In a widely cited article, Hacker (2004) claims that the resilience of the US welfare state to right-wing attacks is more apparent than real. In the face of institutional stickiness and political polarization, anti-welfare politicians simply reverted to a strategy of non-updating benefits to rising labor market risks and changing family patterns.

THE ‘NEW POLITICS’ AS A THEORY OF POLICYMAKING: KEY ACTORS AND MECHANISMS The most influential parts of Pierson’s book are perhaps not the empirical analysis of the UK and the US, but his theoretical account of the politics of retrenchment. Although it is sometimes portrayed in more simplistic terms, in the original formulation Pierson’s theory is a complex structure, built from elements at various analytical levels and borrowed from different social scientific disciplines (including psychology, policy studies, economics). Instead of a simple list of independent variables such as partisan composition of government or veto points, what the theory offers is a series of interlocking causal mechanisms operating at the level of voters, interest groups and policymakers, notably governments. These three types of actors, together with the policies themselves, are the main ingredients to the explanation, and they are assumed to be equipped with certain characteristics. I focus here on the political mechanisms of welfare state retrenchment. Purely institutional path dependency of the increasing returns type is an additional mechanism constraining retrenchment – and policy change more generally – but one that I will have to leave out for the sake of simplicity (Pierson 1994: 39–46, 2000). What do I mean by mechanism? Mechanisms to explain a phenomenon consist of ‘entities (or parts) whose activities and interactions are organized so as to be responsible for the phenomenon’ (Glennan and Illari 2017: 2; see also Machamer et al. 2000). The main entities are governments, interest groups and voters as well as the policies themselves. There are three main clusters of activities (with sub-activities) that structure the theory in the sense that we can talk of three distinct mechanisms that, taken together, result in the ‘new politics of the welfare state’: (1) electoral punishment, (2) interest group mobilization and (3) blame avoidance. The three mechanisms are arranged in a feedback loop (Figure 3.1) and together produce the phenomenon of retrenchment politics (or ‘new politics’). Let us consider each mechanism separately: 1. Electoral accountability or punishment mechanism. Voters – the main entities in this mechanism – are assumed to be materialistic and self-interested (Table 3.1). Their choices are retrospective and based on policy changes and their material consequences for themselves. While there are net beneficiaries and net taxpayers, there are reasons to think that the political resistance of beneficiaries will be more consequential than taxpayer support.

28  Handbook on austerity, populism and the welfare state First, voters are boundedly rational and subject to cognitive biases, especially loss aversion (Kahneman and Tversky 1979). This means that they tend to be more willing to take risks (e.g. voting, switching parties) in response to perceived losses than gains, which makes retrenchment structurally much more difficult than expansion (or mobilizing for tax cuts, for that matter). On the other hand, voters are also imperfectly informed, which means that their perceptions of policy change can be manipulated by policymakers. 2. Interest group mobilization. The interest group channel has been the stepchild of retrenchment studies. While Dismantling takes the power of organized providers and beneficiaries into account, subsequent research has often ignored this channel. Pierson’s account of interest group politics is rationalist, broadly following Mancur Olson’s (1965) collective action theory and related theories such as James Q. Wilson’s policy typology (Wilson 1980). Interest groups are self-interested and care about the net cost of policies to their members, but it is also the distribution of costs and benefits that matters. In the best case, retrenchment has diffuse and often long-term benefits such as lower taxes and fiscal deficits, but the costs imposed on beneficiaries and service providers are immediate, visible and concentrated.4 Once mobilized, organized interests lobby the government and, especially groups with a lot of members can ‘offer or withhold electoral support’ (Pierson 1994: 30).5 Whether interest groups are fully or boundedly rational remains unclear. 3. Blame avoidance. In contrast to voters, politicians are portrayed as fully rational, highly strategic actors.6 Their motivations can be understood through the lens of the policy vs. office (and vote) heuristic (Müller and Strøm 1999): what motivates them is gaining and keeping office and getting their preferred policies implemented, at least in the long run (Pierson 1994: 17). Much of the time, both can be achieved in tandem. But in circumstances of a strong policy–office trade-off – as in the case of retrenchment, because of the pro-welfare attitudes and biases of voters – this reconciliation is not so straightforward anymore. To get out of the dilemma, politicians employ strategies of blame avoidance and systemic retrenchment. It’s not just that they can do this, they must. Given the high risks of counter-mobilization and electoral punishment, blame avoidance is a necessary condition for retrenchment (Starke 2008: 35). It is important to note that in Pierson’s account, blame avoidance is not an exercise in political communication or propaganda. The key activities policymakers engage in for blame avoidance are about policy design (deciding who loses what and when). However, it is crucial that policymakers are also acting under varying constraints stemming from the institutional environment (i.e. the structure of the political system) and the inherited policy design (e.g. means-tested vs. universal benefits). These two kinds of constraints have subsequently often been ignored (but see Jensen and Mortensen 2014; Jensen et al. 2019: 687), even though they are, in fact, the main independent variables, ultimately explaining the success or failure of retrenchment (Pierson 1994: 50). The three mechanisms are linked through a feedback loop. Given the potential for electoral punishment and interest group mobilization, rational policymakers have strong incentives to use strategies of blame avoidance. And the constraints from institutional structure and policy design determine whether they are likely to be successful in cutting back and the distribution and design of these cutbacks. Note that policy change remains the dependent variable and not whether pensioner groups mobilize or angry voters throw out their incumbent government. The context in which all this happens is a democratic political system with mature social protection schemes. This is important because mechanisms cannot usually be taken out of

The politics of retrenchment  29 Table 3.1

Actor characteristics in Dismantling the Welfare State Voters

Interest groups

Policymakers

Motivation

Disposable income

Membership

Office and policy

Decision-making

Boundedly rational

Fully rational (?)

Fully rational

Strongly limited

Somewhat limited

Somewhat limited

Information

Note: Own compilation, based on Pierson (1994).

context; they are ‘switched on’ only under certain conditions (Falleti and Lynch 2009). For example, voters can effectively punish governments in liberal democracies with real electoral competition and free elections. Similarly, in such a context interest groups can freely mobilize, launch campaigns and influence members and other voters. But welfare state maturation also matters, because mature welfare states – via policy feedback – effectively produce their own constituencies, activating the second mechanism, which is why ‘retrenchment is not simply the mirror image of welfare state expansion’ (Pierson 1994: 29). Simply applying the framework to an autocratic system and/or one with only rudimentary systems of social protection would thus be inadequate (Castiglioni 2001: 38; Jung and Walker 2009). Budgetary crisis and what Pierson calls ‘electoral slack’ in his 1996 article, by contrast, may attenuate some of these mechanisms (1996: 176–7). Exceptionally good election results or a weak or ineffective opposition may buffer the government somewhat from electoral retribution (but probably not from interest group feedback).

EMPIRICAL TESTS, CRITICISMS AND EXTENSIONS OF THE ‘NEW POLITICS’ Does Retrenchment Remain Unpunished? With regard to the electoral connection, a string of papers has tested whether retrenchment really leads to punishment at the ballot box. The result is, surprisingly: no (Armingeon and Giger 2008). Electoral punishment is highly conditional at best. For example, parties of the right are typically not punished or can even gain votes in the aftermath of cutbacks (Giger and Nelson 2011; Schumacher et al. 2013). Many voters are conflicted between welfare state support and concerns about fiscal and economic policies (Giger and Nelson 2012; Marx and Schumacher 2016) or not that interested in social policy issues as the retrenchment literature often assumes (Giger 2012). On the one hand, these findings could be read as a resounding refutation of the ‘new politics.’ They seem more in line with a vision of voters as highly partisan, uninterested in policymaking and only ‘blindly’ retrospective in their behavior (Achen and Bartels 2017). Yet on the other hand, this would be an oversimplified version of Pierson’s theory. After all, the idea is that governments are punished unless they employ blame avoidance – which is something none of the punishment studies measures directly. In other words, these studies ignore the feedback loop from blame avoidance back to elections (see Figure 3.1). More recent work on blame avoidance has generally confirmed that the popularity of reforms varies with the use of such strategies (see below). Finally, studies of instances of quite visible retrenchment have confirmed – often via natural experiments – that such cutbacks do have negative ramifications for electoral support and voting (Lindbom 2014 on hospital clo-

30  Handbook on austerity, populism and the welfare state sures in Sweden; Larsen 2017 on Danish student grant cuts; Fervers 2019 on the German Hartz IV legislation). This, however, raises the question of why in some exceptional cases policymakers refrain from blame avoidance in the first place, a question that will be discussed below. Unconventional Politics Politics happens outside of elections too. And yet, unconventional political participation as a potential activity in the feedback mechanism has only recently received some attention from retrenchment scholars. Events such as the protests and strikes against planned pension reforms in France in 2020 and the riots sparked by rising Metro ticket prices in Chile in 2019 vividly demonstrate that political behavior such as going on strike, protesting, boycotting, civil disobedience or even rioting may have a role to play in explaining retrenchment and the absence thereof. Austerity and welfare state cutbacks often lead to protests, that much is clear (Giugni and Grasso 2016). These acts of contention take place in the form of political or general strikes (Lindvall 2011, 2013; Hamann et al. 2013), and perhaps even riots (Kawalerowicz and Biggs 2015; Ponticelli and Voth 2020). So far, this literature has not been well integrated with mainstream retrenchment research. We have limited knowledge about systematic reactions (or the anticipation) of protests by politicians (but see Béland and Marier 2006 on protest avoidance; also Vanhuysse 2009) or the effectiveness of protests regarding reform outcomes (but see Bailey 2014; Taghizadeh and Lindbom 2014; Bailey and Shibata 2019). What is more, protests are often charged with strong negative emotions, but we lack research on the specific negative emotions produced by retrenchment plans (e.g. anger, sadness, fear, empathy) and their respective implications for the behavior of specific groups and the political dynamics. Where Are the Interest Groups? The second, meso-level interest group mechanism has not been as extensively studied as individual-level politics, which is in line with the general biases of contemporary political science toward individual preferences and choices rather than on organizational behavior (see Hacker and Pierson 2014). Interest groups should matter most for retrenchment in social services (such as healthcare and childcare), for at least two reasons. First, in contrast to transfers, services involve providers such as doctors, nurses and childcare workers and provider organizations such as nonprofit and for-profit hospitals which are often relatively autonomous from the state. These actors are represented by powerful lobbying organizations in the political arena, including hospital and trade unions. Second, social service beneficiaries might also be easier to organize than transfer recipients since they tend to be more geographically concentrated around a physical point of provision (e.g. via housing associations, parent groups etc.). Nevertheless, we know very little about these dynamics and the effects of interest group feedback on retrenchment. Interest groups played a certain role in Dismantling, but this aspect of retrenchment politics has since been neglected by scholars (but see Taghizadeh and Lindbom 2014 on client groups; Moe 2011; Anzia and Moe 2015; Hertel-Fernandez 2018 on public sector unions). This should be a highly fruitful field of future research. It would be interesting to see, for example, how interest group feedback varies in line with the service intensity of a welfare state and to what extent increased interest group activity can thwart political strategies of obfuscation and thereby resist retrenchment.

Figure 3.1

Key mechanisms in Dismantling the Welfare State

Notes: Own depiction. Main entities in bold, mechanisms as gray arrows, independent variables in italics.

The politics of retrenchment  31

32  Handbook on austerity, populism and the welfare state The New Politics as a ‘Retrenchment Manual’? The Importance of Blame Avoidance Dismantling has sparked a small industry of blame avoidance studies. Do its findings amount to an ‘retrenchment manual,’ a kind of modern-day ‘mirrors for princes’? Perhaps that would be going too far, but it is striking that research on ways to resist or block retrenchment is not nearly as widespread as research on how to avoid electoral blame for it (a notable exception is Bailey and Shibata 2019). Of course, findings on blame avoidance can potentially also be used by retrenchment critics, but, given the strong information asymmetries between governments and voters, it is doubtful that enlightening the public alone will be enough to counteract crafty politicians. For better or worse, scholars have studied the assumptions, policy patterns and effects of blame avoidance and largely confirmed that mechanism.7 First of all, it seems that policymakers believe that their actions may be unpopular and punishable at the ballot box, as shown by Wenzelburger (2014), based on interviews with former high-level political actors from Belgium, Canada, France and Sweden. Even if actual punishment is weak or non-existent, this is enough for ‘strategies … to minimize political resistance’ (Pierson 1994: 19) to appear.8 And we do indeed find a lot of evidence of the phenomenon. As mentioned, Pierson’s strategies in Dismantling are: (1) obfuscation, for example, by using technically complex benefit formulas and broad reform coalitions to conceal the extent of retrenchment as well as who was responsible; (2) compensation in the form of often symbolic or short-term transfers to those affected by retrenchment; and (3) division, in which the opposition is split, for example, by focusing cutbacks on beneficiaries who are widely considered less deserving. This weakens the solidarity of majority voters, as exemplified by the racist stereotype of the ‘welfare queen’ in the US (Gilens 1999). Other authors have added strategies, and tried to consolidate existing lists of strategies with the help of typologies (Pal and Weaver 2003; Hood 2010; Vis 2016). For reasons of space constraints, I won’t cover all innovations in the blame avoidance toolbox but focus largely on Pierson’s three-fold typology plus some examples from the literature on strategic timing and framing. In her recent literature review, Barbara Vis (2016) provides a more comprehensive overview of the literature on blame avoidance. There is thus no doubt that governments make use of these strategies when cutting back benefits. And yet there are also instances in which governments appear to commit ‘electoral suicide.’ Examples include New Zealand in the 1990s, perhaps Germany and Austria in the early 2000s and the Conservative–Liberal coalition in the UK (Fink and Talos 2004; Starke 2008). During those moments, governments are often straightforward about their goals and means and benefit cuts can turn out to be highly visible. Some scholars and politicians also seem to believe in the virtue of high-risk strategies of ‘shock therapy’ and policy ‘blitzkrieg’ (Williamson 1994; Mitchell 2005). Why are some policymakers seemingly unwilling or unable to employ blame avoidance? Potential answers are that certain politicians are simply more policy-oriented than others, overly optimistic about the economic effects of retrenchment or under the spell of unelected bureaucrats and advisors. The most comprehensive attempt to date to map the distribution of strategies is by Jensen and Wenzelburger (forthcoming) who coded expansionary and restrictive legislative changes in social policy across several European countries since the mid-1970s. Their data show, among other things, that whereas expansionary reforms often increased nominal benefits, retrenchment was typically introduced by stealth, especially through changes in benefit indexation, a classical method of obfuscation (Jensen et al. 2018; see also Lindbom 2007; Green-Pedersen

The politics of retrenchment  33 et al. 2012). Retrenchment also tends to go hand in hand with selective expansionary reforms, thereby confirming Pierson’s compensation strategy (see also Dorlach 2019; Lee et al. 2020). According to Knotz and Lindvall (2015), coalition governments are prone to compensate for the pain of unemployment benefit cuts with increased training expenditure. The strategy of division is arguably more context-dependent and hence more difficult to measure directly across policies and countries. Pierson hypothesized that, due to opportunities for division tactics, means-tested benefits should be more vulnerable to cutbacks than universal benefits. Interestingly, he finds almost the opposite pattern in his book, which he explains with the small size of many targeted programs (generating small fiscal savings) (1994: 170). Yet more recent cross-country evidence suggests that means-tested programs may have been more vulnerable to cuts (Nelson 2007). Other studies also indicate that while pension insurance has often been shielded from cutbacks, this is much less the case for unemployment and social assistance (van Vliet and Wang 2019). An additional blame avoidance technique relates to the strategic timing of reforms. It is well known that policymakers try to introduce less popular measures at the beginning of their term in the hope that voters might forget the pain until the next election. There is some qualitative and quantitative evidence for such effects, but they seem to be conditional on institutional context (Zohlnhöfer 2007; Fernández 2012; König and Wenzelburger 2017). Framing through strategic messaging around retrenchment was not part of Pierson’s original account, but has become an independent strand of the literature. Framing happens when the presentation or communication around a policy has effects on public support. As mentioned earlier, voters care about multiple things beyond social welfare benefits and are often poorly informed. This opens up the opportunity for politicians to frame the issue of retrenchment as a trade-off between welfare state generosity and economic performance (or another issue voters care about) or frame retrenchment as welfare state ‘modernization.’ Scholars argue that a convincing economic justification or framing of retrenchment could dampen electoral feedbacks and may even allow for credit-claiming (Green-Pedersen 2002; Schmidt 2002; Elmelund‐Præstekær and Emmenegger 2013). Yet the effect of economic crisis and crisis discourse might be more ambiguous given that support for welfare benefits (including unemployment benefits) increases during recessions (Jensen 2012; Blekesaune 2013; Margalit 2013; Starke et al. 2013). Hence the government frame can well be met with an effective counter-frame based on these patterns. Another way of framing is by portraying benefit recipients as less deserving and/or morally questionable (Esmark and Schoop 2017). It has also long been known that support for welfare programs is highly sensitive to racial priming (e.g. Gilens 1999), so it is evident that this can be exploited by policymakers. In sum, research has shown that the use of blame avoidance and framing strategies is widespread and systematically linked to the content of policies (i.e. their restrictive direction). Moreover, some of these studies have tried to link the predominance of certain strategies to institutional context conditions à la Pierson (e.g. Jensen and Mortensen 2014). Much work still needs to be done to consolidate the relatively fragmented findings on these comparative opportunities for blame avoidance, however. In addition, blame avoidance has been exclusively applied to electoral politics, but largely left out social movements and interest groups.

34  Handbook on austerity, populism and the welfare state Does Blame Avoidance Work? Addressing the Feedback Loop The blame avoidance literature has increasingly looked at the electoral effectiveness of blame avoidance, hence addressing the feedback arrow in Figure 3.1. Most studies use experimental designs (including survey designs) to test the effect of blame avoidance and framing on support for cutbacks. For example, Wenzelburger and Hörisch (2016) demonstrate that relatively generic blame avoidance and framing strategies affect students’ attitudes toward (hypothetical) pension and student grant cutbacks. Häusermann et al. (2019) find evidence that voters are more likely to support pension cutbacks if reform packages include compensatory spending. However, this is only true for some kinds of compensation (such as including low-wage and atypical workers in the system) and not for others. And Marx and Schumacher (2016) find that several specific economic framings can lower the resistance to retrenchment. Framing recipients as undeserving also has an impact on retrenchment support (e.g. Slothuus 2007) as does linking problems to scapegoats such as immigrants (Goerres et al. 2020). Similarly, case studies of the effect of blame avoidance and framing on government popularity show that strategic communication works, perhaps even better than obfuscation (Elmelund-Præstekær et al. 2015). But these studies typically also make clear that whether framing works outside of the lab in the messy world of politics depends on the context, for example, a weak or ineffective opposition or a specific constellation in the party competition (Davidsson and Marx 2013). Moreover, almost all studies on the effects of blame avoidance use policy or government support as their dependent variable, not vote choice. Whether voters care enough (and long enough) about retrenchment and the framing under which it was introduced for it to have an influence on reelection is therefore still largely unclear. We still lack empirical studies that fully connect the feedback loop between blame avoidance and electoral punishment (see also Vis 2016).

CONCLUSION: BLIND SPOTS AND FUTURE AVENUES This literature review has demonstrated the striking durability of the ‘new politics of the welfare state’ framework as developed by Paul Pierson over 25 years ago. Most of that research has been quantitative and comparative, but there have also been influential small-N and single case studies. The absence of a coherent alternative theoretical account of the politics of retrenchment is remarkable, and despite extensions and revisions here and there, the overall framework remains intact. None of the mechanisms has been clearly disconfirmed by extensive empirical research. And yet, some actors and processes have received more attention than others, which has led to certain blind spots in the research program. I can see at least five issues that merit additional research.9 First, what is really the reason for the absence of strong electoral punishment? Is it uninformed voters, smart politicians using blame avoidance, party system effects and the salience of competing issues or a combination of those factors? We simply don’t know. Future studies will need to combine electoral effects with information on blame avoidance used, with variables on the party-political context and so on in order to solve that puzzle. Second, however, we need to broaden our vision and look beyond the electoral arena too. What is the role of organized interests in thwarting blame avoidance and mobilizing voters? And how does their impact vary across countries and policy areas, including social services which have often been ignored by studies of retrenchment? How important

The politics of retrenchment  35 are unconventional kinds of participation such as political protests and their anticipation for the comparative politics of retrenchment? When and how do protests succeed and when do they fail? The relative lack of studies on protests and interest groups seems to be partly the result of a predominance of micro-level models of electoral politics in political science more generally. Third, ‘systemic retrenchment’ and more generally, the intended and unintended long-term political consequences of welfare state cutbacks at the level of public opinion and political power are still unclear. Can we generally expect a political backlash after a retrenchment period or, on the contrary, a political ‘entrenchment’ of retrenchment? Research on the political effects of economic inequality seems to suggest the latter, but we also witness a disintegration of the former neoliberal elite consensus. Fourth, what role do supra and international dynamics play for welfare state retrenchment and how does this relate to the ‘new politics’? The literature on welfare state retrenchment sometimes seems stuck in a pre-globalized setting of isolated nation states. Yet the wave of austerity packages after the Eurozone crisis demonstrates the very limited explanatory role of domestic actors and that institutions sometimes have vis-à-vis outside interference (Petmesidou and Glatzer 2015; Frieden and Walter 2017). Finally, how do more recent political shifts, often summed up as ‘populism,’ relate to the politics of retrenchment? There is already research suggesting that the rise of populism may be due to past welfare state retrenchment (Berman and Snegovaya 2019; Hopkin and Blyth 2019), but can we also expect the same kind of ‘new politics’ under conditions of multi-dimensional political competition and populist challenges? After decades of clear scientific progress, there remains work to be done for students of welfare state retrenchment.

NOTES 1. Earlier work on retrenchment includes Therborn and Roebroek (1986) and Ramesh (1990). 2. All of this depends on whether one thinks that retrenchment was in fact minor, which is something that authors like Korpi and Palme dispute (2003). 3. The definition does not capture reduction in ‘social policy by other means’ either. Most countries, at least at some point in history, provided material support that is functionally equivalent to formal social protection, for example, in the form of producer or consumer subsidies (Seelkopf and Starke 2019). Many of these programs have become subject to retrenchment too (Dorlach 2019). 4. This is, of course, quite stylized, as the benefits stemming from cutbacks can sometimes be narrowly distributed too. 5. Hence the small arrow between interest groups and voters in Figure 3.1. 6. With regard to information, Pierson acknowledges that there is an informational asymmetry between voters and governments, but even policymakers have only imperfect knowledge, for example, about electoral risks (Pierson 1994: 20). Vis and van Kersbergen later extended the bounded rationality assumption from voters to policymakers (2007). 7. It would be more accurate to talk about several blame avoidance mechanisms, as each strategy presupposes its own set of entities and activities. Spelling this out in detail would go beyond the constraints of this chapter, however. 8. Some authors have claimed that the ‘new politics’ implies the end of left–right politics (e.g. Korpi and Palme 2003) but that conclusion is far from straightforward. Granted, Pierson himself is ambiguous on this point. On the one hand, he claims that ‘there appears to be little correlation between declines in left power resources and the magnitude of retrenchment’ and that ‘there are good reasons to believe that the centrality of left party and union confederation strength to welfare state outcomes has declined’ (1996: 150, 151). On the other hand, his assumption is that policymakers are equally motivated by office (and vote) and policy concerns. As long as Thatcher and Reagan and other ideologically driven actors use strategies of blame avoidance, they can in fact put a partisan mark on

36  Handbook on austerity, populism and the welfare state welfare changes without contradicting the ‘new politics.’ In other words, findings that parties still matter for social policy outputs (see Bandau and Ahrens 2020 for an overview) may not constitute very strong disconfirming evidence. 9. This list applies to welfare states in rich democracies in the first place. Retrenchment in different political and economic settings might raise a whole different set of issues.

REFERENCES Achen, Christopher H., and Larry M. Bartels. 2017. Democracy for Realists: Why Elections Do Not Produce Responsive Government. Vol. 4. Princeton, NJ: Princeton University Press. Anzia, Sarah F., and Terry M. Moe. 2015. ‘Public sector unions and the costs of government.’ The Journal of Politics 77 (1): 114–27. Armingeon, Klaus, and Nathalie Giger. 2008. ‘Conditional punishment: a comparative analysis of the electoral consequences of welfare state retrenchment in OECD nations, 1980–2003.’ West European Politics 31 (3): 558–80. Bailey, David J. 2014. ‘Resistance is futile? The impact of disruptive protest in the “silver age of permanent austerity”.’ Socio-Economic Review 13 (1): 5–32. Bailey, David J., and Saori Shibata. 2019. ‘Austerity and anti-austerity: the political economy of refusal in “low-resistance” models of capitalism.’ British Journal of Political Science 49 (2): 683–709. Bandau, Frank, and Leo Ahrens. 2020. ‘The impact of partisanship in the era of retrenchment: insights from quantitative welfare state research.’ Journal of European Social Policy 30(1): 34–47. Béland, Daniel, and Patrik Marier. 2006. ‘Protest avoidance: labor mobilization and social policy reform in France.’ Mobilization 11 (3): 377–91. Berman, Sheri, and Maria Snegovaya. 2019. ‘Populism and the decline of social democracy.’ Journal of Democracy 30 (3): 5–19. Blekesaune, Morten. 2013. ‘Economic strain and public support for redistribution: a comparative analysis of 28 European countries.’ Journal of Social Policy 42 (1): 57–72. Castiglioni, Rossana. 2001. ‘The politics of retrenchment: the quandaries of social protection under military rule in Chile, 1973–1990.’ Latin American Politics and Society 43 (4): 37–66. Davidsson, Johan Bo, and Paul Marx. 2013. ‘Losing the issue, losing the vote: issue competition and the reform of unemployment insurance in Germany and Sweden.’ Political Studies 61 (3): 505–22. Dorlach, Tim. 2019. ‘Retrenchment of social policy by other means: a comparison of agricultural and housing policy in Turkey.’ Journal of Comparative Policy Analysis: Research and Practice 21 (3): 270–86. Elmelund-Præstekær, Christian, and Michael Baggesen Klitgaard. 2012. ‘Policy or institution? The political choice of retrenchment strategy.’ Journal of European Public Policy 19 (7): 1089–107. Elmelund‐Præstekær, Christian, and Patrick Emmenegger. 2013. ‘Strategic re‐framing as a vote winner: why vote‐seeking governments pursue unpopular reforms.’ Scandinavian Political Studies 36 (1): 23–42. Elmelund-Præstekær, Christian, Michael Baggesen Klitgaard, and Gijs Schumacher. 2015. ‘What wins public support? Communicating or obfuscating welfare state retrenchment.’ European Political Science Review 7 (3): 427–50. Esmark, Anders, and Sarah R. Schoop. 2017. ‘Deserving social benefits? Political framing and media framing of “deservingness” in two welfare reforms in Denmark.’ Journal of European Social Policy 27 (5): 417–32. Esping-Andersen, Gøsta. 1990. The Three Worlds of Welfare Capitalism. Oxford: Polity Press. Falleti, T.G., and J.F. Lynch. 2009. ‘Context and causal mechanisms in political analysis.’ Comparative Political Studies 42 (9): 1143–66. Fernández, Juan J. 2012. ‘Economic crises, population aging and the electoral cycle: explaining pension policy retrenchments in 19 OECD countries, 1981–2004.’ International Journal of Comparative Sociology 53 (2): 77–96. Fervers, Lukas. 2019. ‘Economic miracle, political disaster? Political consequences of Hartz IV.’ Journal of European Social Policy 29 (3): 411–27.

The politics of retrenchment  37 Fink, M., and E. Talos. 2004. ‘Welfare state retrenchment in Austria: ignoring the logic of blame avoidance?’ Journal of Societal & Social Policy 3 (1): 1–21. Frieden, Jeffry, and Stefanie Walter. 2017. ‘Understanding the political economy of the Eurozone crisis.’ Annual Review of Political Science 20 (1): 371–390. Giger, Nathalie. 2012. ‘Is social policy retrenchment unpopular? How welfare reforms affect government popularity.’ European Sociological Review 28 (5): 691–700. Giger, Nathalie, and Moira Nelson. 2011. ‘The electoral consequences of welfare state retrenchment: blame avoidance or credit claiming in the era of permanent austerity?’ European Journal of Political Research 50 (1): 1–23. Giger, Nathalie, and Moira Nelson. 2012. ‘The welfare state or the economy? Preferences, constituencies, and strategies for retrenchment.’ European Sociological Review 29 (5): 1083–94. Gilens, Martin. 1999. Why Americans Hate Welfare: Race, Media, and the Politics of Antipoverty Policy. Chicago, IL: University of Chicago Press. Giugni, Marco, and Maria T. Grasso (eds). 2016. Austerity and Protest: Popular Contention in Times of Economic Crisis. London: Routledge. Glennan, Stuart, and Phyllis Illari. 2017. ‘Introduction: mechanisms and mechanical philosophies.’ In Stuart Glennan and Phyllis Illari (eds), The Routledge Handbook of Mechanisms and Mechanical Philosophy. London: Routledge, pp. 1–9. Goerres, Achim, Rune Karlsen, and Staffan Kumlin. 2020. ‘What makes people worry about the welfare state? A three-country experiment.’ British Journal of Political Science 50 (4): 1519–37. Green-Pedersen, Christoffer. 2002. The Politics of Justification: Party Competition and Welfare-state Retrenchment in Denmark and the Netherlands from 1982 to 1998. Amsterdam: Amsterdam University Press. Green-Pedersen, Christoffer, Flemming Juul Christiansen, Eva-Maria Euchner, Carsten Jensen, and John Turnpenny. 2012. ‘Dismantling by default? The indexation of social benefits in four countries.’ In Michael W. Bauer, Christoffer Green-Pedersen, Adrienne Héritier, and Andrew Jordan (eds), Dismantling Public Policies: Preferences, Strategies, and Effects. Oxford: Oxford University Press, pp. 129–54. Hacker, Jacob S. 2004. ‘Privatizing risk without privatizing the welfare state: the hidden politics of social policy retrenchment in the United States.’ American Political Science Review 98 (2): 243–60. Hacker, Jacob S., and Paul Pierson. 2014. ‘After the “Master Theory”: Downs, Schattschneider, and the rebirth of policy-focused analysis.’ Perspectives on Politics 12 (3): 643–62. Hamann, Kerstin, Alison Johnston, and John Kelly. 2013. ‘Unions against governments: explaining general strikes in Western Europe, 1980–2006.’ Comparative Political Studies 46 (9): 1030–57. Hertel-Fernandez, Alexander. 2018. ‘Policy feedback as political weapon: conservative advocacy and the demobilization of the public sector labor movement.’ Perspectives on Politics 16 (2): 364–79. Hood, Christopher. 2010. The Blame Game: Spin, Bureaucracy, and Self-preservation in Government. Princeton, NJ: Princeton University Press. Hopkin, Jonathan, and Mark Blyth. 2019. ‘The global economics of European populism: growth regimes and party system change in Europe.’ Government and Opposition 54 (2): 193–225. Häusermann, Silja, Thomas Kurer, and Denise Traber. 2019. ‘The politics of trade-offs: studying the dynamics of welfare state reform with conjoint experiments.’ Comparative Political Studies 52 (7): 1059–95. Jensen, Carsten. 2012. ‘Labour market- versus life course-related social policies: understanding cross-programme differences.’ Journal of European Public Policy 19 (2): 275–91. Jensen, Carsten, and Peter B. Mortensen. 2014. ‘Government responses to fiscal austerity: the effect of institutional fragmentation and partisanship.’ Comparative Political Studies 47 (2): 143–70. Jensen, Carsten, and Georg Wenzelburger. forthcoming. Reforming the Welfare State. London: Routledge. Jensen, Carsten, Christoph Arndt, Seonghui Lee, and Georg Wenzelburger. 2018. ‘Policy instruments and welfare state reform.’ Journal of European Social Policy 28 (2): 161–76. Jensen, Carsten, Georg Wenzelburger, and Reimut Zohlnhöfer. 2019. ‘Dismantling the welfare state? After twenty-five years: what have we learned and what should we learn?’ Journal of European Social Policy 29 (5): 681–91.

38  Handbook on austerity, populism and the welfare state Jung, Chang Lyul, and Alan Walker. 2009. ‘The impact of neo‐liberalism on South Korea’s public pension: a political economy of pension reform.’ Social Policy & Administration 43 (5): 425–44. Kahneman, Daniel, and Aaron Tversky. 1979. ‘Prospect theory: an analysis of decision under risk.’ Econometrica 47 (2): 263–92. Kawalerowicz, Juta, and Michael Biggs. 2015. ‘Anarchy in the UK: economic deprivation, social disorganization, and political grievances in the London Riot of 2011.’ Social Forces 94 (2): 673–98. Klitgaard, Michael B., and Christian Elmelund-Præstekær. 2014. ‘The partisanship of systemic retrenchment: tax policy and welfare reform in Denmark 1975–2008.’ European Political Science Review 6 (1): 1–19. Klitgaard, Michael Baggesen, Gijs Schumacher, and Menno Soentken. 2015. ‘The partisan politics of institutional welfare state reform.’ Journal of European Public Policy 22 (7): 948–66. Knotz, Carlo, and Johannes Lindvall. 2015. ‘Coalitions and compensation: the case of unemployment benefit duration.’ Comparative Political Studies 48 (5): 586–615. Korpi, Walter, and Joakim Palme. 2003. ‘New politics and class politics in the context of austerity and globalization: welfare state regress in 18 countries, 1975–95.’ American Political Science Review 97 (3): 425–46. König, Pascal, and Georg Wenzelburger. 2017. ‘Honeymoon in the crisis? A comparative analysis of the strategic timing of austerity policies and their effect on government popularity in three countries.’ Comparative European Politics 15 (6): 991–1015. Larsen, Erik Gahner. 2017. ‘Welfare retrenchments and government support: evidence from a natural experiment.’ European Sociological Review 34 (1): 40–51. Lee, Seonghui, Carsten Jensen, Christoph Arndt, and Georg Wenzelburger. 2020. ‘Risky business? Welfare state reforms and government support in Britain and Denmark.’ British Journal of Political Science 50 (1): 165–84. Lindbom, Anders. 2007. ‘Obfuscating retrenchment: Swedish welfare policy in the 1990s.’ Journal of Public Policy 27 (2): 129–50. Lindbom, Anders. 2014. ‘Waking up the giant? Hospital closures and electoral punishment in Sweden.’ In Staffan Kumlin and Isabelle Stadelmann-Steffen (eds), How Welfare States Shape the Democratic Public: Policy Feedback, Participation, Voting, and Attitudes. Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing, pp. 156–78. Lindvall, Johannes. 2011. ‘The political foundations of trust and distrust: reforms and protests in France.’ West European Politics 34 (2): 296–316. Lindvall, Johannes. 2013. ‘Union density and political strikes.’ World Politics 65 (3): 539–69. Lo Vuolo, Rubén M. 1997. ‘The retrenchment of the welfare state in Latin America: the case of Argentina.’ Social Policy & Administration 31 (4): 390–409. Machamer, Peter, Lindley Darden, and Carl F. Craver. 2000. ‘Thinking about mechanisms.’ Philosophy of Science 67 (1): 1–25. Margalit, Yotam. 2013. ‘Explaining social policy preferences: evidence from the Great Recession.’ American Political Science Review 107 (1): 80–103. Marier, Patrik. 2016. ‘Devine intervention? Lessons in systemic retrenchment from Canada’s most generous welfare state.’ Regional & Federal Studies 26 (2): 221–41. Marx, Paul, and Gijs Schumacher. 2016. ‘The effect of economic change and elite framing on support for welfare state retrenchment: a survey experiment.’ Journal of European Social Policy 26 (1): 20–31. Mitchell, A. 2005. ‘Roger’s revolution: Blitzkrieg in Kiwiland.’ Journal of Legislative Studies 11 (1): 1–15. Moe, Terry M. 2011. Special Interest: Teachers Unions and America’s Public Schools. Washington, DC: Brookings Institution Press. Müller, Wolfgang C., and Kaare Strøm (eds). 1999. Policy, Office, or Votes? How Political Parties in Western Europe Make Hard Decisions. Cambridge: Cambridge University Press. Nelson, K. 2007. ‘Universalism versus targeting: the vulnerability of social insurance and means-tested minimum income protection in 18 countries, 1990–2002.’ International Social Security Review 60 (1): 33–58. Olson, Mancur. 1965. The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, MA: Harvard University Press.

The politics of retrenchment  39 Pal, Leslie A., and R. Kent Weaver. 2003. ‘The politics of pain.’ In Leslie A. Pal and R Kent Weaver (eds), The Government Taketh Away: The Politics of Pain in the United States and Canada. Washington, DC: Georgetown University Press, pp. 1–40. Petmesidou, Maria, and Miguel Glatzer. 2015. ‘The crisis imperative, reform dynamics and rescaling in Greece and Portugal.’ European Journal of Social Security 17 (2): 158–81. Pierson, Paul. 1994. Dismantling the Welfare State? Reagan, Thatcher, and the Politics of Retrenchment. Cambridge: Cambridge University Press. Pierson, Paul. 1996. ‘The new politics of the welfare state.’ World Politics 48 (2): 143–79. Pierson, Paul. 2000. ‘Increasing returns, path dependence, and the study of politics.’ American Political Science Review 94 (2): 251–67. Ponticelli, Jacopo, and Hans-Joachim Voth. 2020. ‘Austerity and anarchy: budget cuts and social unrest in Europe, 1919–2008.’ Journal of Comparative Economics 48 (1): 1–19. Ramesh, Mishra. 1990. The Welfare State in Capitalist Society – Policies of Retrenchment and Maintenance in Europe, North America and Australia. New York: Harvester Wheatsheaf. Schmidt, Vivien A. 2002. ‘Does discourse matter in the politics of welfare state adjustment?’ Comparative Political Studies 35 (2): 168–93. Schumacher, Gijs, Barbara Vis, and Kees Van Kersbergen. 2013. ‘Political parties’ welfare image, electoral punishment and welfare state retrenchment.’ Comparative European Politics 11 (1): 1–21. Seelkopf, Laura, and Peter Starke. 2019. ‘Social policy by other means: theorizing unconventional forms of welfare production.’ Journal of Comparative Policy Analysis: Research and Practice 21 (3): 219–34. Slothuus, Rune. 2007. ‘Framing deservingness to win support for welfare state retrenchment.’ Scandinavian Political Studies 30 (3): 323–44. Starke, Peter. 2008. Radical Welfare State Retrenchment: A Comparative Analysis. Basingstoke: Palgrave Macmillan. Starke, Peter, Alexandra Kaasch, and Franca Van Hooren. 2013. The Welfare State as Crisis Manager: Explaining the Diversity of Policy Responses to Economic Crisis. Basingstoke: Palgrave Macmillan. Streeck, Wolfgang, and Kathleen Thelen. 2005. ‘Introduction: institutional change in advanced political economies.’ In Wolfgang Streeck and Kathleen Thelen (eds), Beyond Continuity: Institutional Change in Advanced Political Economies. Oxford: Oxford University Press, pp. 1–39. Taghizadeh, Jonas Larsson, and Anders Lindbom. 2014. ‘Protests against welfare retrenchment: healthcare restructuring in Sweden.’ Scandinavian Political Studies 37 (1): 1–20. Taylor‐Gooby, Peter. 1999. ‘Policy change at a time of retrenchment: recent pension reform in France, Germany, Italy and the UK.’ Social Policy & Administration 33 (1): 1–19. Therborn, Göran, and Joop Roebroek. 1986. ‘The irreversible welfare state: its recent maturation, its encounter with the economic crisis, and its future prospects.’ International Journal of Health Services 16 (3): 319–38. van Vliet, Olaf, and Jinxian Wang. 2019. ‘The political economy of social assistance and minimum income benefits: a comparative analysis across 26 OECD countries.’ Comparative European Politics 17 (1): 49–71. Vanhuysse, Pieter. 2009. ‘Power, order and the politics of social policy in Central and Eastern Europe.’ In Alfio Cerami and Pieter Vanhuysse (eds), Post-Communist Welfare Pathways. Basingstoke: Palgrave Macmillan, pp. 53–70. Vis, Barbara. 2016. ‘Taking stock of the comparative literature on the role of blame avoidance strategies in social policy reform.’ Journal of Comparative Policy Analysis: Research and Practice 18 (2): 122–37. Vis, Barbara, and Kees Van Kersbergen. 2007. ‘Why and how do political actors pursue risky reforms?’ Journal of Theoretical Politics 19 (2): 153–72. Wenzelburger, Georg. 2014. ‘Blame avoidance, electoral punishment and the perceptions of risk.’ Journal of European Social Policy 24 (1): 80–91. Wenzelburger, Georg, and Felix Hörisch. 2016. ‘Framing effects and comparative social policy reform: comparing blame avoidance evidence from two experiments.’ Journal of Comparative Policy Analysis: Research and Practice 18 (2): 157–75. Williamson, John (ed.) 1994. The Political Economy of Policy Reform. Washington, DC: Institute for International Economics.

40  Handbook on austerity, populism and the welfare state Wilson, James Q. 1980. ‘The politics of regulation.’ In James Q. Wilson (ed.), The Politics of Regulation. New York: Basic Books, pp. 357–74. Zohlnhöfer, Reimut. 2007. ‘The politics of budget consolidation in Britain and Germany: the impact of blame avoidance opportunities.’ West European Politics 30 (5): 1120–38.

4. Populism and the welfare state Cristóbal Rovira Kaltwasser and Lisa Zanotti

INTRODUCTION1 Much of the contemporary debate about the welfare state is related to its change, reform and retrenchment. After all, there is little doubt that the welfare state today confronts several internal and external challenges. Take, for instance, issues such as a growing, ageing population and changing family structures as well as economic globalization, mass migration and impact of the Great Recession (Bermeo and Bartels 2014). Within this debate about the challenges of the welfare state, populism has received very little attention.2 This is a curious omission as a large part of the current academic and public debate focusses on the rise of populist forces that are transforming the democratic order. In other words, we have limited knowledge about the relationship between populism and the welfare state, especially about the ways in which the former might have an impact on the latter. In this contribution, we address this research gap. As we will argue in this chapter, the link between populism and the welfare state is anything but straightforward. To better understand this complex relationship, it is important to acknowledge that populism is first and foremost a set of ideas based on the moral distinction between ‘the pure people’ and ‘the corrupt elite’, while the welfare state is commonly defined as a democratic regime that aims to guarantee social rights, in addition to civil and political rights. Seen from a theoretical perspective, it is plausible to argue that in general terms populism challenges the welfare state, because it seeks to (re)define who the ‘true’ members of the political community are and therefore who should have social rights. By advancing a moral understanding of the political world (Müller 2016), populist forces indeed bring to the fore the debate about deservingness/non-deservingness that it is inherent to the welfare state (Greve 2019). This is why, at an abstract level, one can argue that populist forces are prone to somehow challenging the welfare state, since they seek to favour the social rights of those conceived of as ‘the pure people’ in opposition to those who are seen as ‘impure’ and to limit the influence of those conceived as ‘the corrupt elite’. However, seen from an empirical level, existing research has shown that populism normally appears attached to other ideologies (e.g., nativism, producerism, socialism, etc.), which play a pivotal role in advancing political projects of different political colour (Mudde and Rovira Kaltwasser 2013, 2017). Consequently, we rely on the scholarly work that identifies two subtypes of populism – inclusionary and exclusionary populism – that employ different ideologies, which lead to different understandings of the welfare state. In summary, the main aim of this chapter consists in demonstrating that when studying the relationship between populism and welfare state, it is important to distinguish the theoretical from the empirical discussion. On the one hand, the theoretical discussion reveals that, by (re) politicizing the debate about who the members of the demos are, populism ends up putting the welfare state under stress, specifically in terms of establishing who should receive social benefits and who should pay for them. On the other hand, the empirical discussion demonstrates that two different subtypes of populism are predominant today and that each of them 41

42  Handbook on austerity, populism and the welfare state challenges the welfare state in different ways. Therefore, when it comes to thinking about the link between populism and the welfare state, instead of providing generalizations, scholars and practitioners should rather examine how specific subtypes of populism propel different redefinitions of the functioning of the welfare state. The rest of this chapter is divided in five sections. We start by presenting a definition of populism that is dominant within the contemporary debate, particularly in the political science literature. After this, we analyse the relationship between populism and the welfare state. Particular emphasis is given here to the development of an analytical framework for understanding the ways in which populist forces can affect the welfare state. The next two sections deal with two subtypes of populism and their relationship with the welfare state. Here we show that while exclusionary populism hinges on nativist ideas to affect the interests of some elites and help certain segments of the people, inclusionary populism relies on socialist ideas to target certain elites and benefit some segments of the people. Finally, we close with a short conclusion, in which the main arguments of the chapter are presented and a couple of ideas for future research are advanced.

WHAT IS POPULISM? Populism has been and, to a certain extent, still is a contested concept. However, it seems that some consensus has arisen around the so-called ideational approach, according to which populism should be thought of as a set of ideas that not only portrays society as divided between ‘the corrupt elite’ and ‘the pure people’ but also defends popular sovereignty at any cost. The ideational approach encompasses all the definitions of populism that conceive it as an ideology (Mudde 2004, 2007; Mudde and Rovira Kaltwasser 2017), a discourse (Laclau 2005; Stavrakakis 2014), a frame (Aslanidis 2016) or a heartland (Taggart 2000). In practice, populism rarely appears in its pure form, but rather in articulation with other ‘host ideologies’ such as nativism, producerism, socialism, and so on. As Mudde and Rovira Kaltwasser (2017) pointed out, these ‘host ideologies’ are crucial for the promotion of political projects that are appealing to a broader public. In general terms, all ideational definitions of populism have at least three elements in common: the concept of society divided between (i) the people and (ii) the elite, and (iii) that the people express their general will through politics. First, in the ideational definition of populism, the people are at the centre of the populist discourse. ‘The people’ are a homogeneous subject and are morally pure by definition. It is also important to notice that the concept is not associated with any predetermined group, but it is an ‘imagined community’ or construct. In Laclau’s (2005) words, ‘the people’ is an empty signifier, that is, an empty box that can assume different meanings depending on the context. In this sense, we can observe that the discourse of populist radical right parties, which are typically studied in Western Europe, is characterized by limiting the concept of ‘the pure people’ to the native population of the country (Mudde 2007). On the contrary, radical left populist forces, like the ones that emerged in Latin America towards the end of the 1990s and the beginning of the 2000s, had a more inclusionary concept of ‘the pure people’ (Mudde and Rovira Kaltwasser 2013). The same stands for ‘the elite’. Just as ‘the people’ are conceived as a homogeneous, single group in the ideational definition of populism, so are ‘the elite’. Different kinds of populist discourse entail a different framing of who the members of ‘the corrupt elite’ are. To con-

Populism and the welfare state  43 tinue using our previous example, for populist radical right parties in Western Europe, the elite category is represented by a supranational entity such as the European Union, which is responsible for making decisions that affect the life of ‘the pure people’, reducing their power. On the contrary, for radical left populism, ‘the corrupt elite’ is represented by groups, such as international financial institutions, that respond to interests that are different from those of ‘the pure people’. In this regard, while the people are always morally pure, the elite are always depicted as corrupt. It is important to understand that corruption here is not, or at least not primarily, defined in monetary terms, but rather in moral terms. The elite is corrupt because it is not willing to act in line with the will of the people and therefore is not inclined to respect the (silent) majority. This leads us to the third characteristic of the populist ideology. For populists, politics should always be an expression of the people’s general will. Populist political actors then embody the people, so that the actions of the former are the reflection of the will of the latter. Since populists conceive ‘the pure people’ as a homogeneous and undivided entity, the general will is also a clear and homogeneous idea that cannot be doubted but should be rather strictly obeyed in order to enact a model of government that allegedly is truly democratic. Using the ideational definition of populism has three main advantages. First, as briefly mentioned above, conceiving populism as an ideology or discourse forces us to differentiate between different types of populism, depending on the ‘host ideology’ to which populism is associated. As we will argue below, this allows us to distinguish between exclusionary and inclusionary populism – a taxonomy that is particularly relevant when it comes to thinking about the relationship between different types of populism and the welfare state. The second advantage is that by employing the ideational definition of populism, scholars can analyse both the supply and demand side of the phenomenon. While some scholars have described populism as a top-down political strategy, in which charismatic leaders mobilize unorganized masses (Weyland 2001; Roberts 2006), the ideational approach allows for the examination of both the supply side – the populist leaders or parties – alongside the demand side – the activation of the populist attitudes in the electorate. This last point refers to the fact that to be successful, populism needs leaders or parties that articulate the discourse as well as support for those ideas at the mass level. Last, since populism is a set of ideas, it can be exploited not only by leaders, but also by other political agents, such as political parties or social movements (Mudde and Rovira Kaltwasser 2017). In fact, while populism in Latin America has been mainly embodied by a leader due to the presidentialist features of the subregion, in contemporary Western Europe there are plenty of examples of populist parties such as the National Rally (the former National Front) in France, the Swiss People’s Party in Switzerland and SYRIZA in Greece, just to mention a few. Having defined the features of populism, it is important to underline an important corollary. Since populism is a thin-centred ideology, a discourse or a frame, it does not entail any specific policy. In other words, there is no such thing as populist policies. This does not undermine the fact that certain subtypes of populism that are characterized by certain ideological traits tend to implement similar types of policies. For instance, if we analyse the populist radical right party family, we cannot deny that they all tend to push for the implementation of restrictive immigration policies. However, a consideration is in order. Populist radical right parties share more than just the populist features. The more relevant ideological trait they share is nativism, that is, the belief that countries should be inhabited only by the native group (the nation) and that non-natives are fundamentally threatening the homogeneous nation-state (Mudde 2007).

44  Handbook on austerity, populism and the welfare state In this sense, the anti-immigration discourse and the policies that populist radical right eventually seek to implement are not an expression of their populist ideology, but instead of their nativism.

POPULISM AND THE WELFARE STATE At this stage, we do have clarity about the notion of populism. In order to understand its relationship with the welfare state it is important to present a working definition of the latter. Following the work of Van Kersbergen and Manow (2008, 522), one can say that ‘the welfare state is first and foremost a democratic state that – in addition to civil and political rights … guarantees social protection as a right attached to citizenship’. Of course, there is great variation between welfare states, since they differ with respect to generosity and scope of social protection. In fact, the scholarly literature identifies different types of welfare states, such as: the Scandinavian social democratic regime marked by generous funding and social protection to all citizens; the Continental conservative regime characterized by substantial benefits for workers; the Anglo-Saxon liberal regime distinguished by relatively low benefits and usually targeted at clearly delineated social groups; and the Southern European regime typified by a highly segmented job market with high protection for the few state workers and limited protection for the rest.3 Despite the existence of these different types of welfare state, there is little doubt that it is in all its different forms under pressure from many sides today, particularly because of the ageing population, economic globalization, transformation of family structures and, in recent times, the aftereffects of the ‘Great Recession’. However, there is not much debate about the extent to which populism represents a challenge to the welfare state and if this is the case, which aspects of the welfare state are under threat because of populism. To answer these questions properly, we need to pause for a second and reflect on the theoretical arguments that have been advanced to understand the rise and evolution of the welfare state. By doing this, we are better situated to then examine the impact of populism. One of the most common arguments in the academic literature is that the growth of the welfare state is directly related to the democratic class struggle. This means that collective actors such as the labour movement, special interest groups and political parties are the ones who fight for social policies in favour of their clientele. Seen in this light, the different level of power of the collective actors fighting for social policies helps to explain the variation in the generosity of the welfare state. Indeed, the classic work of Huber and Stephens (2001) maintains that electorally strong left-wing forces lead to an expansion of the welfare state, while the opposite occurs when right-wing forces are able to obtain a sizeable share of the vote and therefore attain the executive office. The logic behind the so-called ‘partisan politics’ argument is that ‘the working class is the main constituency of the left, [which] holds strong preferences for redistribution and social insurance, and – consequently – the left fights for welfare state expansion in the democratic class struggle’ (Häusermann et al. 2013, 226). However, a growing body of scholarly work reveals that electoral constituencies have been changing in recent decades and this is why traditional political forces – such as Christian democratic and social democratic parties – have declined in electoral power, while at the same time new political forces – such as the greens and the populist radical right – seem to have growing electoral support (e.g., Kitschelt 1994; Mudde 2007; Kriesi et al. 2008). Part of this change has

Populism and the welfare state  45 to do with the fact that cultural conflicts have been gaining relevance and mainstream political parties have serious difficulties adapting to this new scenario, in which the classic economic struggle is less relevant than before. At the same time, middle-class sectors have been losing their historical position in societies, something well documented in new studies that reveal that in most advanced democracies the levels of subjective social status have been diminishing, particularly for working-class men (e.g., Gidron and Hall 2017, 2020). It is not a coincidence then that the political landscape has become much more diverse today as various political parties coexist and compete for the representation of the interests of different social groups. On the one hand, there is a new electorate of younger, highly educated and strongly open-minded voters, who push both old and new parties to devote less attention to the traditional redistributive policy agenda (Inglehart 1990). On the other hand, workers and low-income voters are increasingly worried about immigration, status loss and deindustrialization (Häusermann et al. 2013). In this debate about the transformation of the political landscape, populism plays an important role. After all, populist forces normally attack mainstream political parties for their alleged lack of interest in understanding the real concerns of ‘the pure people’. Moreover, by (re)politicizing issues that are relevant to certain electoral constituencies that do not feel well represented by mainstream political parties, populist forces are able to challenge the ways the party system works. More often than not, mainstream political parties feel obliged to rethink their programmatic agendas, especially if populist forces have the capacity to consolidate in the electoral arena by stealing some of the voters that had previously supported mainstream parties. Part of the success of populist forces lies in their ability to open a debate about who the members of ‘the pure people’ are, and this certainly has an impact on the welfare state. Remember that the latter assumes that citizens do have not only civil and political but also social rights. However, the populist discourse seeks to (re)define who the ‘true’ members of the community are and this certainly has an effect on the ways in which the welfare state should operate, especially when it comes to thinking about the groups that should have and should not have social rights. With respect to this point it is necessary to underline that the discussion on who does and who does not deserve benefits and services is an issue intrinsic to the welfare state itself. In fact, it is argued that opinions on social policy are strongly influenced by whether or not recipients are seen as deserving. However, it is also true that the rise of populist forces somehow push further this classic social policy discussion of the deserving and undeserving, thereby increasing its saliency (Greve 2019, 3). This is because populists reframe the deservingness/undeservingness dilemma as a moral scheme, since they consider only some members of the community members of ‘the pure people’, while their opponents are depicted as corrupt and impure (Müller 2016, 25). As we will see in the next section, it is important to analyse in empirical terms the existence of different subtypes of populism, since each of them relates to the welfare state in their own way. Nevertheless, at a theoretical level, it is possible to identify that populism puts the welfare state under stress. Given that the populist set of ideas is based on the distinction between ‘us’ (the good ones) and ‘them’ (the bad ones), it advances a moral understanding about the members of ‘the pure people’. The latter is seen as a homogeneous community that works hard and is being exploited by the establishment. In this very process of identifying ‘the pure people’, populist forces play a pivotal role since they define the criteria for considering someone as an insider or an outsider. Therefore, populist forces usually claim that ‘the pure people’ should receive public assistance and ‘the corrupt elite’ should pay for it. Yet, this

46  Handbook on austerity, populism and the welfare state approach defence of the welfare state is problematic in democratic terms because establishing objective criteria for determining the boundaries between the two groups is anything but simple (Näsström 2007; Ochoa Espejo 2011; Rovira Kaltwasser 2014).

EXCLUSIONARY/INCLUSIONARY POPULISM AND THEIR RELATIONSHIP WITH THE WELFARE STATE The distinction between inclusionary and exclusionary populism has become increasingly common in the academic literature, since it helps to characterize the two different versions of populism that are predominant in the contemporary world. This distinction is based on three dimensions: material, political and symbolic (Betz 2001; Filc 2010; Mudde and Rovira Kaltwasser 2013). The material dimension is linked to the distribution of material resources to various groups in society. While in most cases exclusionary populism tends to exclude people from the distribution of these resources, under inclusionary populism some groups are targeted to receive more state resources (Mudde and Rovira Kaltwasser 2013, 158). The political dimension refers to Robert Dahl’s (1971) concept of democracy, which is based on political participation and public contestation. While exclusionary populism prevents the full participation and contestation of some groups, inclusionary populism targets some groups to extend their participation and contestation. Last, the symbolic dimension alludes to the boundaries set on both ‘the people’ and ‘the elite’. Although this dimension looks less straightforward, it does not mean that it is less relevant (Mudde and Rovira Katwasser 2013). In fact, when populist actors define ‘the people’ with certain symbols and values, those who do not share these symbols and values are implicitly excluded and therefore become ‘the others’. Even if right-wing and left-wing populism share both exclusionary and inclusionary traits, we can say that, in general terms, exclusionary populism is usually right-wing and that inclusionary populism is normally left-wing. Before exploring the relationship between each of these subtypes of populism and the welfare state, one clarification needs to be made. Conceiving populism as an ideology or discourse means that this can be different from the policy actually implemented by administrations supported by populist forces. This is particularly relevant if we think that in many Western European countries, populist parties have access to government when they form a coalition with mainstream parties. In these cases, populist parties need to bargain with other political forces, and in consequence, the government outputs are the result of the interaction between populist and non-populist actors. In sum, our analysis focusses on the relationship between populism and the welfare state at a discursive level, but at a practical level it is important to consider the extent to which populist forces are able to access power by themselves or in coalition with non-populist forces. Exclusionary Populism and the Welfare State In this section we analyse the characteristics of exclusionary populism and the relation it holds with the welfare state. The focus lies in cases in both Western and Eastern Europe, two regions that have seen the proliferation of populist radical right forces in the last few years. Let’s begin by closely examining the material, political and symbolic dimensions of exclusionary populism. As mentioned above, the material dimension refers to the distribution of the material

Populism and the welfare state  47 resources of the state to specific groups in society, and in consequence, exclusionary populism seeks to leave out particular groups from access to state resources, such as public sector jobs or welfare provisions (Mudde and Rovira Kaltwasser 2013, 158). This is particularly relevant with reference to the welfare state since, as we explain later, exclusionary populism usually goes hand in hand with welfare chauvinism: the idea that social benefits should be provided only to the natives. Regarding the political sphere, exclusion means that some groups are prevented from fully participating in the democratic system and they are consciously not represented in the arena of public contestation (Mudde and Rovira Kaltwasser 2013, 161). This aspect is linked to the withholding from non-citizens of voting rights or the call for limited political and religious rights for the ‘alien’ groups. Finally, the symbolic dimension also refers to the boundaries of both ‘the pure people’ and ‘the corrupt elite’. In the case of exclusionary populism, ‘the pure people’ is defined in symbolic terms as all those sharing the characteristics and values of the native population. However, it is important to notice that within the category of exclusionary populism, the boundaries of both ‘the pure people’ and ‘the corrupt elite’ are not fixed. In other words, exclusionary forms of populism do not always target the same group. For instance, although it is true that exclusionary populist forces in Western Europe are characterized by presenting the achievements of European civilization against challengers ranging from American popular culture to Islam (Betz 2001, 394), after September 11 and particularly because of the terrorist attacks perpetrated by the terrorist group Al-Qaeda, the ‘aliens’ became a more restricted group: Muslims (Betz and Meret 2009). At the same time, exclusionary populist forces in Eastern Europe have a tendency to include Jews as part of ‘the aliens’, something less common in Western Europe (Mudde 2007). The boundaries of the ‘corrupt elite’ are not fixed either. As Mudde (2013, 7) has pointed out, exclusionary populism in Western Europe depicts the establishment as morally corrupt because elites ‘have hijacked the political system and silenced the voice of the people by making backroom deals and enforcing a conspiracy of silence’. In turn, exclusionary populism in Eastern Europe usually portrays ‘the corrupt elite’ as those powerful actors who controlled the Communist regime as well as the (new) liberal establishment that in the 1990s was in charge of the transition process (Stanley 2017, 144–8). Despite these differences between Eastern and Western Europe, in recent years exclusionary populist forces in both regions have targeted mainly the EU, depicting it as a bureaucratic body guilty of taking away decisional power from the people. However, this was not always the case. In the 1980s, exclusionary populism was not against the European Community. Things started to change at the beginning of the 1990s, particularly with the implementation of the Maastricht Treaty and the transformation of the Community into the Union, which meant a broadening of its prerogatives. Moreover, in more recent times the EU has started to push member states to take responsibility for a share of migrants as a result of the migration crisis, generating anger in exclusionary populist actors. After having described the discourse of exclusionary populism we focus now on its core constituency. Who supports this political project? The academic literature finds that those who vote for exclusionary populist forces are the so-called ‘losers of globalization’: older men with low levels of education (Betz 1994; Goodhart 2017). Nevertheless, there is an open debate about the empirical validity of this argument, since in many countries the voters of exclusionary populist forces are also middle-class and young citizens (Spier 2010; Arzheimer 2011; Mudde 2013). Suffice it to say here that it is important to note that the notion of ‘losers

48  Handbook on austerity, populism and the welfare state of globalization’ should not be equated with individuals who are unemployed and economically marginalized (i.e., labour market outsiders), but rather with individuals who have secure jobs (i.e., labour market insiders) and feel threatened by the arrival of immigrants as well as ongoing societal transformations that threaten their social status (Häusermann et al. 2013; Mudde and Rovira Kaltwasser 2018). As Kriesi et al. (2006, 929) have pointed out, ‘[g]iven the heterogenous economic interests of the “losers” of denationalization, the defence of their national identity and their national community constitutes the smallest common denominator for their mobilization’. Because most of those voters who support populist radical right parties should be seen as labour market insiders, it is not a surprise that exclusionary populist forces increasingly adopt welfare chauvinist stances (Akkerman et al. 2016). Welfare chauvinism can be thought of as a political view guided by nativism (Greve 2019): it maintains that benefits and services should go only to natives, since they are the only ones entitled to receive support from the welfare state (Ennser-Jedenastik 2018). Not by chance, asylum seekers, migrants and refugees are normally seen by exclusionary populists as ‘welfare tourists’, meaning undeserving individuals who allegedly come to steal benefits from the natives who are the ones who deserve protection from the welfare state. Interestingly, the welfare chauvinist discourse is appealing in both Eastern and Western Europe, although these two regions present very different levels of immigration and also of social benefits. How can we understand the attraction of the welfare chauvinist rhetoric? The answer to this question is that by arguing that access to welfare should be restricted to the ‘deserving natives’, exclusionary populist actors have the capacity to mobilize older men with low levels of education and protected jobs, are interested in securing their privileges and increasingly worry about immigration. However, the fact that exclusionary populism tends to be in favour of welfare chauvinism does not mean that it promotes a left-wing policy. After all, the very notion of welfare chauvinism involves a restrictive interpretation of who are the pure people, and in consequence, it is employed to blame the corrupt elite for ‘cutting the welfare rights of deserving “natives” and the non-natives for their excessive claims on the welfare state’ (Schumacher and Van Kersbergen 2016, 302). Since exclusionary populist forces seek to demarcate what they see as the ‘authentic’ and ‘deserving’ members of the community from those ‘others’ who are deemed legitimately excludable, welfare chauvinism is a natural expression of such demarcation whose political logic is sharply at odds with the integrative thrust of left populisms (Roberts 2019, 652). The adoption of welfare chauvinist positions reflects the combination of economic and cultural issues. In effect, welfare chauvinism not only implies the promotion of a state that provides generous economic protection to ‘the pure people’, but also the advancement of cultural demarcations to redefine who should receive the benefits. Interestingly, exclusionary populist forces are particularly successful in the electoral arena of Western European countries marked by economic prosperity, low unemployment and generous social welfare policies (e.g., Austria, the Netherlands and Switzerland). This means that in those contexts where the welfare state is more consolidated and generous, there is fertile soil for the appearance of exclusionary populists who promise to return to an idealized image of the past characterized by ethnic homogeneity, social cohesion and generous welfare policies. The support of populist radical right forces with welfare chauvinist positions is partly driven by catchwords such as ‘welfare tourism’ and ‘welfare magnetism’, which both relate to the free movement of workers within the EU (Greve 2019, 29). In fact, there is a growing conflict within the EU

Populism and the welfare state  49 member states (typically, an east–west divide as well as north–south divide) between sending and receiving countries, as well as a clash of ‘supporters of pan-European free-movement and non-discrimination, on the one hand, against supporters of social and cultural closures, on the other’ (Ferrera 2017, 7). Inclusionary Populism and the Welfare State In contrast to exclusionary populism, inclusionary populism seeks to (re)incorporate large sections of the electorate that are in a precarious situation and/or have been left out. As for exclusionary populism, to better understand the profile of inclusionary populism, it is important to consider three dimensions (Mudde and Rovira Kaltwasser 2013). First, the material dimension of inclusionary populism alludes to the necessity of expanding the role of the state in the provision of social rights. This normally implies the development of a strong critique of the establishment because of the alleged protection of its own interests rather than those of the pure people. Second, the political dimension of inclusionary populism refers to fostering more and better participation by those who are in a marginal position within society, so that they can increase their influence and thus diminish the power of the elite. Third and finally, in the case of inclusionary populism, the symbolic dimension is normally linked to socialist ideas, through which ‘the corrupt elite’ is depicted as a plutocratic minority and ‘the pure people’ is portrayed as the vast majority of citizens who feel excluded and left behind. Given that socialist ideas play an important role in inclusionary populism, the latter tends to define the boundaries of both ‘the pure people’ and ‘the corrupt elite’ in socioeconomic terms. This makes an important difference with exclusionary populism, which relies mainly on sociocultural markers. Nevertheless, inclusionary populism also reveals some variance when it comes to framing the moral distinction between the people and the elite. On the one hand, ‘the pure people’ is usually conceived of as the underdogs, all those who do not profit from the existing economic model and therefore struggle to make ends meet. Yet, depending on the context in which inclusionary populism operates, it emphasizes certain characteristics of ‘the pure people’, such as the disadvantaged position of citizens living in certain regions of the country (e.g., Die Linke in Germany) or the precarious job situation of young people (e.g., SYRIZA in Greece and Podemos in Spain). On the other hand, ‘the corrupt elite’ is generally considered a tiny minority that is very rich and has enough power to control the political system. In some cases, special emphasis is given to the damaging role of financial elites, who allegedly spare no effort in evading taxes and therefore undermining the proper functioning of the welfare state. By contrast, in other cases the main enemy are mainstream political parties, which are portrayed as a cartel of elites (e.g., ‘the caste’) interested in its own interests rather than in respecting the will of the people. When looking at the European context, inclusionary populism is the exception rather than the rule. However, this started to change because of the Great Recession. As the economic crisis of 2008–09 deeply affected Southern European economies, it is not a surprise that exclusionary populist forces started to gain traction in countries such as Spain (Podemos) and Greece (SYRIZA). The rise of these inclusionary populist forces is related to the fact that social democratic parties were in power when the Great Recession began, so they were the parties that implemented austerity measures, which ended up alienating an important segment of those who have historically supported social democratic parties (Roberts 2017). Moreover, empirical studies on inclusionary populism in Western Europe reveal that its core

50  Handbook on austerity, populism and the welfare state constituency is mainly young people, many of them with high levels of education, who favour strong intervention by the state in the economy as well as progressive values on issues such as immigration (Ramiro and Gomez 2017; Rooduijn 2018). What about the relationship between inclusionary populism and the welfare state? There is little doubt that the former is in favour of the latter. Given that inclusionary populism argues that there should be generous social protection for the pure people and that the corrupt elite should pay its fair share, it is at the forefront of the attack against austerity and welfare retrenchment. Despite some differences about the extent to which the establishment should be able to receive benefits or not, inclusionary populist forces spare no effort in supporting social rights and high taxation rates for the wealthy. For instance, Jean-Luc Mélenchon’s la France insoumise and Pablo Iglesias’s Podemos advance a harsh critique of free market economics and demand the enactment of a progressive tax system in order to finance and improve the welfare state (Kioupkiolis 2019; Marlière 2019). At the same time, several of inclusionary populism condemn the austerity policies promoted by the EU and endorse instead the idea that the EU should allocate resources to generate a sort of transnational social citizenship (Katsambekis and Kioupkiolis 2019). As Koch (2020) has recently demonstrated, inclusionary populist forces are prone to advancing a cosmopolitan political project that is characterized by the demand for transnational forms of citizenship that include the protection of social rights. It is worth noting that inclusionary populist forces are much more electorally successful in Southern than in Northern Europe. As Roberts (2019) has indicated, the reason for this lies in the existence of different political economic contexts in Southern Europe and in Northern Europe: whereas in the former material issues are quite relevant for a large segment of the electorate – particularly after the Great Recession – in the latter post-material issues have been gaining preponderance in the last few years – particularly after the 2015–16 refugee crisis. In other words, in the absence of a strong welfare state and with harsh austerity measures, there is fertile soil for the emergence and consolidation of inclusionary populist forces that seek to (re)invigorate the provision of social rights by the welfare state. By contrast, in those countries where the welfare state is well consolidated and the pressure for austerity is less, there is fertile soil for the emergence and consolidation of exclusionary populist forces that promote welfare chauvinist politics with the aim of maintaining the existing social benefits only for the native population and denying immigrants access to the welfare state.

CONCLUSION In this chapter we have offered an analysis of the relationship between populism and the welfare state. By providing working definitions of these two concepts, we argue that at an abstract level populism challenges the welfare state, since it seeks to redefine who should receive social benefits and who should pay for this. As mentioned above, populists put more strain on the issue of deservingness versus undeservingness since they conceive society as divided in two moral opposite groups: the pure people versus the corrupt elite. This is a very broad generalization, however. Therefore, we maintain that it is important to work at an empirical level in order to distinguish the existence of subtypes of populism which affect the welfare state in different ways. On the one hand, one can identify exclusionary populist forces that are characterized by promoting welfare chauvinist positions, that is, the defence of generous social rights only for those who are depicted as the natives. This type of populist force is particularly

Populism and the welfare state  51 successful in Northern European countries, where a relatively generous welfare state exists, and in consequence, the combination of populist and nativist ideas are employed to defend the social rights of those who have secure jobs (i.e., labour market insiders) and feel threatened by the arrival of immigrants and ongoing societal transformations that put their social status under threat. On the other hand, one can identify inclusionary populist forces that are marked by the defence of the welfare state, which should be financed by a progressive tax system that targets the wealthy. This subtype of populism is more common in Southern Europe, a region that not only has a relatively weak welfare state (at least in comparison with Northern Europe) but has also experienced serious economic difficulties in the last few years because of the Great Recession and the consequent pressure by the EU to enact austerity measures. Not by chance, exclusionary populist forces are supported mainly by those who are unemployed (i.e., labour market outsiders) and/or in a precarious economic situation (i.e., the young), so their demands are centred on material rather than on cultural issues. Previous studies have indeed shown that the Great Recession has not affected public support for the welfare state but rather has given more saliency to the very existence of a generous system of social rights (Greve 2019). Although it is true that the differing impact of exclusionary and inclusionary populism on the welfare state holds true for Western Europe, this argument needs to be tested in more detail in other world regions. For instance, there is little doubt that with the coming into power of Donald Trump, the United States entered a new political era marked not only by the ascendency of exclusionary populism, but also by the formation of an inclusionary populist wing within the Democratic Party headed by Bernard Sanders. Nevertheless, the welfare state in the United States offers low benefits that are usually targeted at clearly delineated social groups, and in consequence, it would be relevant to look in more detail at the impact of these two subtypes of populism on the US welfare regime. The same type of research could be done for other world regions. In fact, Brazil has seen the emergence of an exclusionary populist president (Jair Bolsonaro) and several inclusionary populists have captured the executive office in South America (e.g., Hugo Chávez in Venezuela, Rafael Correa in Ecuador and Evo Morales in Bolivia).

NOTES 1. Funding. The study was funded by Fondo Nacional de Desarrollo Científico y Tecnológico (FONDECYT Project 1180020), the Centre for Social Conflict and Cohesion Studies - COES (ANID / FONDAP / 15130009) and the Observatory for Socioeconomic Transformations (ANID / PCI / Max Planck Institute for the Study of Societies / MPG190012).  2. A partial exception can be found in the recent literature that examines the development of welfare chauvinist positions by the populist radical right in Europe (e.g., Afonso 2015; Schumacher and Van Kersbergen 2016). However, as we will argue below, this literature is about one specific type of populism rather than about populism per se. 3. We are aware of the fact that there is open debate about the existence of various welfare state types around the world, and in consequence, the four mentioned types do not exhaust the list of classifications.

52  Handbook on austerity, populism and the welfare state

REFERENCES Afonso, A. 2015. ‘Choosing Whom to Betray: Populist Right-wing Parties, Welfare State Reforms and the Trade-off Between Office and Votes.’ European Political Science Review 7, no. 2: 271–92. Akkerman, T., S.L. de Lange, and M. Rooduijn. 2016. ‘Into the Mainstream? A Comparative Analysis of the Programmatic Profiles of Radical Right-wing Populist Parties in Western Europe over Time.’ In Radical Right-wing Populist Parties in Western Europe, 49–70. Oxon and New York: Routledge. Arzheimer, K. 2011. ‘Electoral Sociology – Who Votes for the Extreme Right and Why – and When?’ In The Extreme Right in Europe: Currents, Trends and Perspectives, edited by U. Backes and P. Moreau, 35–50. Göttingen, Germany: Vandenhoeck & Ruprecht. Aslanidis, P. 2016. ‘Is Populism an Ideology? A Refutation and a New Perspective.’ Political Studies 64, no. 1: 88–104. Bermeo, N., and L. Bartels, eds. 2014. Mass Politics in Tough Times: Opinions, Votes and Protest in the Great Recession. New York: Oxford University Press. Betz, H.G. 1994. Radical Right-wing Populism in Western Europe. Basingstoke: Macmillan. Betz, H.G. 2001. ‘Exclusionary Populism in Austria, Italy, and Switzerland.’ International Journal 56, no. 3: 393–420. Betz, H.G., and S. Meret. 2009. ‘Revisiting Lepanto: The Political Mobilization Against Islam in Contemporary Western Europe.’ Patterns of Prejudice 43, nos 3–4: 313–34. Dahl, R. 1971. Polyarchy, Participation and Opposition. New Haven, CT: Yale University Press. Ennser-Jedenastik, L. 2018. ‘Welfare Chauvinism in Populist Radical Right Platforms: The Role of Redistributive Justice Principles.’ Social Policy & Administration 52, no. 1: 293–314. Ferrera, M. 2017. ‘Mission Impossible? Reconciling Economic and Social Europe after the Euro Crisis and Brexit.’ European Journal of Political Research 56, no. 3: 3–22. Filc, D. 2010. The Political Right in Israel: Different Faces of Jewish Populism. London: Routledge. Gidron, N., and P.A. Hall. 2017. ‘The Politics of Social Status: Economic and Cultural Roots of the Populist Right.’ The British Journal of Sociology 68: S57–S84. Gidron, N., and P.A. Hall. 2020. ‘Populism as a Problem of Social Integration.’ Comparative Political Studies 53, no. 7: 1027–59. Goodhart, D. 2017. The Road to Somewhere: The Populist Revolt and the Future of Politics. Oxford: Oxford University Press. Greve, B. 2019. Welfare, Populism and Welfare Chauvinism. Bristol: Policy Press. Häusermann, S., G. Picot, and D. Geering. 2013. ‘Rethinking Party Politics and the Welfare State – Recent Advances in the Literature.’ British Journal of Political Science 43, no. 1: 221–40. Huber, E., and J. Stephens. 2001. Development and Crisis of the Welfare State. Parties and Policies in Global Markets. Chicago, IL: University of Chicago Press. Inglehart, R. 1990. Culture Shift in Advanced Industrial Society. Princeton, NJ: Princeton University Press. Katsambekis, G., and A. Kioupkiolis, eds. 2019. The Populist Radical Left in Europe. London: Routledge. Kioupkiolis, A. 2019. ‘Late Modern Adventures of Leftist Populism in Spain: The Case of Podemos, 2014–2018.’ In The Populist Radical Left in Europe, edited by G. Katsambekis and A. Kioupkiolis, 47–72. London: Routledge. Kitschelt, H. 1994. The Transformation of European Social Democracy. New York: Cambridge University Press. Koch, C. 2020. ‘Varieties of Populism and the Challenges to Global Constitutionalism: Dangers, Promises and Implications.’ Global Constitutionalism: 1–39. doi: doi:​10​.1017/​S2045381719000455. Kriesi, H., E. Grande, R. Lachat, M. Dolezal, S. Bornschier, and T. Frey. 2006. ‘Globalization and the Transformation of the National Political Space: Six European Countries Compared.’ European Journal of Political Research 45, no. 6: 921–56. Kriesi, H., E. Grande, R. Lachat, M. Dolezal, S. Bornschier, and T. Frey. 2008. West European Politics in the Age of Globalization. Cambridge: Cambridge University Press. Laclau, E. 2005. On Populist Reason. London: Verso.

Populism and the welfare state  53 Marlière, P. 2019. ‘Jean-Luc Mélenchon and France Insoumise: The Manufacturing of Populism.’ In The Populist Radical Left in Europe, edited by G. Katsambekis and A. Kioupkiolis, 93–112. London: Routledge. Mudde, C. 2004. ‘The Populist Zeitgeist.’ Government and Opposition 39, no. 4: 542–63. Mudde, C. 2007. Populist Radical Right Parties in Europe. Cambridge: Cambridge University Press. Mudde, C. 2013. ‘Three Decades of Populist Radical Right Parties in Western Europe: So What?’ European Journal of Political Research 52, no. 1: 1–19. Mudde, C., and C. Rovira Kaltwasser. 2013. ‘Exclusionary vs. Inclusionary Populism: Comparing Contemporary Europe and Latin America.’ Government and Opposition 48, no. 2: 147–74. Mudde, C., and C. Rovira Kaltwasser. 2017. Populism: A Very Short Introduction. Oxford: Oxford University Press. Mudde, C., and C. Rovira Kaltwasser. 2018. ‘Studying Populism in Comparative Perspective: Reflections on the Contemporary and Future Research Agenda.’ Comparative Political Studies 51, no. 13: 1667–93. Müller, J.W. 2016. What is Populism? Philadelphia: Pennsylvania University Press. Näsström, S. 2007. ‘The Legitimacy of the People.’ Political Theory 35, no. 5: 624–58. Ochoa Espejo, P. 2011. The Time of Popular Sovereignty. University Park, TX: Pennsylvania University Press. Ramiro, L., and R. Gomez. 2017. ‘Radical-left Populism during the Great Recession: Podemos and Its Competition with the Established Radical Left.’ Political Studies 65, no. 1: 108–26. Roberts, K. 2006. ‘Populism, Political Conflict, and Grass-roots Organization in Latin America.’ Comparative Politics 38, no. 2: 127–48. Roberts, K.M. 2017. ‘State of the Field. Party Politics in Hard Times: Comparative Perspectives on the European and Latin American Economic Crises.’ European Journal of Political Research 56, no. 2: 218–33. Roberts, K.M. 2019. ‘Bipolar Disorders: Varieties of Capitalism and Populist Out-flanking on the Left and Right.’ Polity 51, no. 4: 641–53. Rooduijn, M. 2018. ‘What Unites the Voter Bases of Populist Parties? Comparing the Electorates of 15 Populist Parties.’ European Political Science Review 10, no. 3: 351–68. Rovira Kaltwasser, C. 2014. ‘The Responses of Populism to Dahl’s Democratic Dilemmas.’ Political Studies 62, no. 3: 470–87. Schumacher, G., and K. Van Kersbergen. 2016. ‘Do Mainstream Parties Adapt to the Welfare Chauvinism of Populist Parties?’ Party Politics 22, no. 3: 300–12. Spier, T. 2010. Modernisierungsverlierer?: Die Wählerschaft rechtspopulistischer Parteien in Westeuropa. Wiesbaden: Springer-Verlag. Stanley, B. 2017. ‘Populism in Central and Eastern Europe.’ In The Oxford Handbook of Populism, edited by C. Rovira Kaltwasser, P. Taggart, P. Ochoa Epejo, and P. Ostiguy, 140–60. Oxford: Oxford University Press. Stavrakakis, Y. 2014. ‘The Return of “the People:” Populism and Anti-populism in the Shadow of the European Crisis.’ Constellations 21, no. 4: 505­–17. Taggart, P. 2000. Populism. Buckingham: Open University Press. Van Kersbergen, K., and P. Manow. 2008. ‘The Welfare State.’ In Comparative Politics, edited by D. Caramani, 520–45. Oxford: Oxford University Press. Weyland, K. 2001. ‘Clarifying a Contested Concept. Populism in the Study of Latin American Politics.’ Comparative Politics 34, no. 1: 1–22.

5. Measuring retrenchment in welfare states: overcoming the challenges to the definition, operationalization and measurement of welfare policy change Elisa Helena Xiol Y Ferreira and Michael Howlett

INTRODUCTION Measuring growth and retrenchment in welfare states is needed to understand and analyse the dynamics of welfare regimes. Comparative policy studies of welfare state retrenchment, and studies of policy change in general, require attention to the interplay of several elements and dimensions involved in welfare policy-making. Undermining any of these features and the links between them may significantly impact the validity of the actual patterns of change portrayed. Questions of how to measure welfare efforts, in particular, are not simple and are highly contingent on the theoretical conceptualization of such regimes and upon the types of policies and the levels of government being examined. How to define welfare and well-being from the perspective of outputs or outcomes, whether to focus on the politics of welfare state change (reforms) and/or the policies of welfare state growth, how to categorize policy components, how to define and measure the impacts of fixed and variable costs due to demographic changes are amongst the many considerations which have to made when determining how to estimate welfare state activity. In the past, for example, the measurement of regimes has suffered from what can be termed a ‘dependent variable’ problem, with different authors utilizing different criteria to define regime elements and thus arriving at different estimations of policy effort in this area. The result of this lack of consensus on the definition, operationalization and measurement of policy efforts at the fundamental level of what should be regarded as a ‘welfare state’ means there are different conceptions in the literature on welfare change in general, and also retrenchment in welfare states more specifically. Esping-Andersen’s (1990) work on welfare state regimes, for example, is frequently cited in discussions of retrenchment. Esping-Andersen, however, argues that social expenditure data is not an appropriate measure if one’s definition of welfare state expansion is broad and linked to levels of de-commodification in society (Esping-Andersen et al. 2002). But, according to Green-Pedersen (2004, 2007), if the same case was tackled from a different theoretical perspective, and depending on the research question, growth in social security expenditure may well be a fitting data type to include, even if only as a complement to other measures or indicators (as will be seen in the following sections). Not only does this example illustrate that measurements must be carefully considered vis-à-vis the author’s understanding of the welfare state, it demonstrates how the theoretical confusion in defining the welfare state impacts discussions of both expansion and retrenchment 54

Measuring retrenchment in welfare states  55 (Allan and Scruggs 2004; Esping-Andersen 1990; Green-Pedersen 2004; Green-Pedersen and Haverland 2002). Beyond that, it sheds further light on the challenges of building ‘robust foundations for systematic empirical cross-national analyses’ (Clasen and Siegel 2007, 12) in welfare state research, if there is not a broadly shared empirical basis or shared interpretations of welfare state effort. This is not to say that significant strides have not been made in this area or that they have all been in vain. Pierson’s seminal work on welfare state change (1994, 1996, 2000), for example, paved the way for the decomposition of welfare activity into several policy components which collectively help to determine the substance and direction of social policy (Bauer and Knill 2014). This chapter builds on this insight and focuses on the need to precisely disaggregate the elements of policy in a way that allows for effective measurement of retrenchment and the accumulation of knowledge on the topic. It will first examine the obstacle that is the dependent variable problem in the study of welfare states. Once the challenge for consensus on central theoretical definitions of welfare and retrenchment has been explored, the need for elements of policy to be disentangled is discussed. The importance of proper classification tools and measurements is emphasized. But the chapter also includes other considerations when measuring retrenchment, such as the time selected and measuring replacement rates; and the variety of tactics and methodologies which can be employed in welfare retrenchment studies to address these issues.

PROBLEMS IN EXISTING MEASURES OF SOCIAL WELFARE EFFORT: QUANTITATIVE ANALYSES VERSUS MIXED METHODS There currently has been a lack of consensus in the literature on many fundamental questions around social welfare policy and this is a major obstacle not only for the measurement of welfare state retrenchment, but also for accumulation of knowledge on the topic (Clasen and Siegel 2007; Green-Pedersen 2004; Howlett and Cashore 2009; Pierson 2002). Building on Pierson’s (1994, 1996) seminal discussions of welfare state retrenchment, Green-Pedersen (2004) highlighted this dependent variable problem and emphasized that the obstacle lies in theoretical conceptualization of welfare state activities, not in any lack of data used for empirical investigations per se. The dependent variable selected for the study of retrenchment has often been more a reflection of the researcher’s underlying understanding of not only what is the welfare state but also of the specific kinds of goals and aims which underlie welfare state policy than any kind of objective measure (Bridgen 2018; Howlett and Cashore 2009; Pierson 1994, 1996, 2002; Starke 2006). In terms of retrenchment specifically, the prevalent theoretical perspectives commonly classify retrenchment as amounting to cutbacks in entitlements or as institutional change (Green-Pedersen 2004). In this analysis of retrenchment as cutbacks, average replacement rates in social security and aggregate expenditure data are commonly used; whether outcome or output measures are applied, each possesses strengths and weaknesses (Green-Pedersen 2004). Jahn (2018), for example, analyses redistribution effects during retrenchment using CWED2, a dataset of unemployment replacement rates of low, middle, and high income levels. Spicker’s (2018) work emphasizes how quantitative methods offer ways to organize

56  Handbook on austerity, populism and the welfare state and classify information, yet they are not conducive to the generalization sought after in comparative policy studies. Kühner (2007), however, like Esping-Andersen, has noted that ‘one must be very critical indeed of attempts to operationalize “welfare state change” by utilizing social expenditure data as percentages of GDP [gross domestic product] (even if given as first order differences or as annual percentage growth rates) in particular’ (16). Through their comparison of replacement rates (RR) from prominent datasets, Scruggs’s Comparative Welfare Entitlements Dataset (CWED) and Korpi and Palme’s Social Citizenship Indicator Program (SCIP), Wenzelburger et al. (2013) point out the substantial disagreement about the determinants of welfare state retrenchment found there, and caution against entrusting RR data alone to address the dependent variable problem, as does Scruggs (2013). Greve (2020) also examines implications of using replacement rates, the pitfalls of using a percentage of GDP, assessing both overall spending and that of specific policy fields, and the challenges of measuring the quality of services among several other considerations in measuring retrenchment. Other empirical research on welfare states that employ predominantly quantitative indicators include those issued by the Organisation for European Co-operation and Development (OECD), Eurostat and the United Nations (UN) which deal with outputs and outcomes and the performance of welfare states. How’s Life?2020 (2020), for example, uses over 80 well-being indicators and classifies changes in trends over time via two types of analysis, for indicators with sufficient time series and those without. Like the UN’s Sustainable Development Goals ‘Measuring Progress’ report (2014), it ‘illustrates and quantifies a progress index that is a simple and minimal adjustment to the GDP’ (73) but also uses a composite index, a compilation of individual indicators, to measure and monitor development. These problems led Castles (1998) to claim the main role of quantitative methods is to provide ‘a preliminary sorting process, informing us about combinations of variables which fit together to produce possible accounts’ (20). According to Spicker (2018): ‘the broad generalizations that come from quantitative analysis are worth far less than qualitative, in-depth considerations of the political, economic and social context of a country’ (226). Nevertheless, there is no question that quantitative indicators contribute significantly to discussions of welfare state retrenchment, particularly when conducting cross-national research. The Need for Mixed Methods However, likewise, the notion that quantitative data alone is always sufficient to assess comparative welfare state effort is inaccurate. The use of qualitative data as well can overcome some of the limitations of purely quantitative work, as it can add to understandings of local contexts and policy nuances in a manner that can provide researchers with a more complete view of the mechanisms at play in policy change, further progressing our understandings of the phenomena in question (Del Pino and Ramos 2018). An example of such a combined or mixed-methods approach can be found in the work of Del Pino and Ramos (2018) on welfare retrenchment in Spain. The authors analysed the determinants of reforms through a ‘process-tracing study of decision processes and the contents of the policy documents that justified the reforms, as well as political speeches by policy-makers’ (707). Examining five regional governments in Spain between 2009 and 2014 – the second phase of the economic crisis that significantly impacted the country’s economy – they developed a composite index to analyse healthcare reforms utilizing seven quantitative

Measuring retrenchment in welfare states  57 and qualitative indicators. These indicators were selected with respect to various initiatives that involved a reduction in the rights, benefits or services granted to citizens before the crisis (Del Pino and Ramos 2018). In their study, several figures such as GDP trends, unemployment rates and public deficit were also used. But these were accompanied by 15 in-depth interviews conducted over three years to address questions on the determinants of retrenchment decisions, the processes involved and the obstacles faced during implementation (Del Pino and Ramos 2018). Similarly, in Wenzelburger’s (2011) work on the strategic dimensions of adjustment efforts in Sweden, Belgium, Canada and France, qualitative interviews were conducted with finance ministers, prime ministers and direct collaborators to circumvent the limits of the comparable quantitative data on political strategy across the countries in question. Several interviews were conducted with different actors to understand the different perspectives each held and even confront interviewees with rival interpretations (Wenzelburger 2011). Detailed process-tracing of fiscal adjustments were conducted as a basis for interviews and qualitative content analysis of key speeches on fiscal and economic policies in order to isolate important messages and unveil potential patterns regarding strategy (Wenzelburger 2011). From a different angle, Boeri et al. (2001) used cross-country surveys focusing on unemployment policies and pensions, and interviews, to assess attitudes towards the welfare state. Elmelund-Præstekær and Klitgaard (2012; Elmelund-Præstekær et al. 2015), in a within-country comparative analysis of Denmark, distinguish between policy retrenchment and institutional retrenchment and used quantitative data1 that allowed for multiple controls, but also data that could qualitatively separate the two types of retrenchment. Emphasizing what many other scholars have also indicated, Elmelund-Præstekær and Klitgaard (2012) claim that ‘qualitative case studies are also needed to develop a deeper understanding of specific reform strategies chosen by particular governments within particular areas’ (1104). It provides invaluable context that can prevent flawed claims when presenting findings on retrenchment in welfare states, and several other settings and movements. Acknowledging that measuring retrenchment remains a problematic task, Pierson (1994, 1996) also advised against using solely quantitative indicators as they can be inadequate. ‘Pure spending estimates will fail to capture the impact of reforms that are designed to introduce retrenchment only indirectly or over the long term … rather than emphasizing cuts in spending per se, the focus is on reforms that indicate structural shifts in the welfare state’ (Pierson 1996, 157). Pierson’s claims that retrenchment has been pursued cautiously and radical changes are rare have influenced much of the literature on retrenchment in welfare states. There is often a divide between the use of quantitative versus qualitative data in the study of welfare state retrenchment, however the emphasis on this distinction is misplaced in terms of measurement of welfare state growth and retrenchment. The combination of the two serve to counteract the deficiencies in each one, in a manner that is conducive to more accuracy in findings, as well as a larger conversation from a comparative perspective that is attuned to the complexities and significance of local context in variations amongst countries and/or sectors.

58  Handbook on austerity, populism and the welfare state

PROBLEMS WITH POLICY AND OUTCOME-BASED DEFINITIONS OF SOCIAL WELFARE EFFORT: DEFINING THE DEPENDENT VARIABLE Divergent conceptualizations of welfare state composition and activity limit the possibility of engaging in mainstream debates and methodological generalizability, amongst other drawbacks (Crawford and Ostrom 1995; Spicker 2018; Starke 2006). While no one definition is inherently superior or inferior to another, the focus of the study of changes in welfare efforts changes considerably depending on how one defines a welfare state. Typically, the definition of welfare states is split into policy- and outcome-based definitions (Green-Pedersen 2004; Armingeon et al. 2016), with the former centred on the range of benefits provided by the state and the latter upon their impact on, for example, poverty rates, family stability or standards of living. The problems with policy-based definitions lies in deciding what policies to include and how to operationalize them in such a fashion as to allow cross-national or cross-sectoral comparisons (Green-Pedersen 2004; Howlett and Cashore 2009; Jahn 2018). Outcome-based definitions, on the other hand, revolve around regime-level changes linked to certain outcomes such as the development of market or social democratic regimes (Esping-Andersen), and these can be problematic when determining what classifies as change in such regimes (Green-Pedersen 2004). As Pierson noted As the concept of the welfare state or welfare regime ‘stretches’, it becomes inevitable that quite distinct processes and outcomes will be joined together under the umbrella of a single master variable ... [But] In a context where actors have complex motives, and the dependent variable is so heterogeneous, attempts to reduce change to a single dimension will be counterproductive. (Pierson 2002, 377–8)

Awareness of the dependent variable problem when examining policy dynamics requires better taxonomies of policy elements to be developed in order to disentangle what exactly is being measured and described (Mortensen 2005; Robinson and Caver 2006; Robinson et al. 2007). ‘These challenges constrain both our theoretical knowledge and our ability to provide practical advice for designing better public policies’ (Siddiki et al. 2019, 3).

MOVING BEYOND THE DEPENDENT VARIABLE PROBLEM: REFINING TAXONOMIES OF POLICY COMPONENTS TO BETTER UNDERSTAND WELFARE STATE CHANGE Peter Hall’s (1993) work on classification of policy change is the most cited taxonomy of policy components in the literature and has often been applied in empirical studies. It challenged the dominant view in existing scholarship that tended to conflate all the elements of a ‘policy’ into a single dependent variable (Heclo 1976; Rose 1976). For years Hall’s (1993) three principal types of policy change – first, second and third order changes – characterized the study of policy dynamics. Hall emphasized there must be a distinction between ‘ends’ and ‘means’ in policy development, as well as between theoretical goals, specific programme objectives/content and operational setting. In this approach, first order changes consist of calibrations of policy instruments within existing institutional

Measuring retrenchment in welfare states  59 and instrument bounds; second order changes involve changes to dominant types of policy instruments used within an existing policy regime; third order changes entail shifts in overall abstract policy goals (Howlett and Cashore 2009). Influenced by Simon’s (1957) and Lindblom’s (1959) works on incremental policy change, Hall and generations of scholars argued that incremental change is associated with marginal change and often synonymous to stability, while paradigmatic change was assessed as atypical and unstable. Thus, according to Hall, first and second order changes were ‘generally incremental and typically the result of activities endogenous to a policy subsystem while third order changes were “paradigmatic” and occurred as anomalies that arose between expected and actual results of policy implementation’ (Howlett and Cashore 2009, 37). In doing so, Hall linked each change process to a different cause agent and to an overall pattern of ‘punctuated equilibrium’ (Baumgartner et al. 2009). While path-breaking at the time, since the publication of Hall’s work, studies have shown that the links between policy components and sources of change are more complex than originally proposed (Bannink and Hoogenboom 2007; Hogan and Howlett 2015; Ostrom and Basurto 2011) and that Hall’s original classifications were underdeveloped. In a reconceptualization of the model, Howlett and Cashore (2009) identified six components of policy (Figure 5.1). These policy elements were argued to exist in a ‘nested’ relationship ranging from the macro (institutions) to the meso (sectoral policy regimes) and the micro (policy actor behaviour) (Hall 1993; Howlett 2009).

Source: Modified from Cashore and Howlett (2007).

Figure 5.1

A modified taxonomy of policy components following Hall

‘The implication of this taxonomy is that every “policy” is in fact a more complex regime of ends and means-related goals (more abstract), objectives (less abstract), and settings (least abstract) than was suggested by the use of Hall’s original de-composition and definition of the elements of policy into three “orders”’ (Howlett and Cashore 2007, 54).

60  Handbook on austerity, populism and the welfare state Table 5.1

Ostrom’s seven kinds of policy criteria/rules

1. Boundary rules that specify how actors were to be chosen to enter or leave these positions; 2. Position rules that specify a set of positions and how many actors hold each one; 3. Choice rules that specify which actions are assigned to an actor in a position; 4. Information rules that specify channels of communication among actors and what information must, may, or must not be shared; 5. Scope rules that specify the outcomes that could be affected; 6. Aggregation rules (such as majority or unanimity rules) that specify how the decisions of actors at a node were to be mapped to intermediate or final outcomes; and 7. Payoff rules that specify how benefits and costs were to be distributed to actors in positions.

Source: Ostrom, Elinor, Nobel Prize Lecture 2009 ‘Beyond Markets and States: Polycentric Governance of Complex Economic Systems’.

This understanding of policy elements helps to better operationalize and measure the components of welfare states and how they change. Content analysis of policies as well as quantitative measures can then be used to examine patterns in a replicable and systematic manner. Combined with the understanding of policy levels and dynamics per Howlett and Cashore (2007, 2009), this approach can be a step towards a shared understanding and language for future works on the topic, if incorporated by more scholars. Qualitative analyses can fill in much of the picture with respect to higher level orders of elements such as changes in policy goals and instruments. However, the ‘strategies, norms, and rules that indicate permitted, required, or prohibited conduct within specified temporal, spatial, and/or procedural boundaries’ (Siddiki et al. 2019, 2) are key mechanisms that permit, propose or prevent the provision of public goods and services such as those provided by the welfare state (Fukuyama 2013). Many of these exist at the level of policy calibrations and are often more susceptible to quantitative treatment. As a result, there is a need for a rigorous and systematic basis upon which to classify and evaluate the calibration of policy tools in the Howlett and Cashore model. Originally developed by Sue Crawford and Elinor Ostrom in 1995 as part of a ‘Grammar of Institutions’, more widely known as an Institutional Grammar, one such framework focuses on the kinds of rules put in place in any policy system to determine what good and services are provided to whom and how. This kind of analysis is needed in providing a foundation for detecting specific, empirical elements as instances of more abstract concepts that come together in meaningful patterns, which then allow for more theoretically driven analysis (Basurto et al. 2010; Carter et al. 2016; Schlüter and Theesfeld 2010; Siddiki et al. 2019 on Chomsky 1957). The first step consists of dividing calibrations into their constitutive elements (Crawford and Ostrom 1995). Ostrom argues that most policies articulate at least seven different kinds of rules governing everything from who is included in a policy, to who can contest it and what kinds of payoffs and penalties are levied on targets (Table 5.1). Taken together, these provide a clear set of guidelines for the study of how kinds of settings required to put such policy goals and objectives in come about, and to assess an important aspect of policy regime growth and retrenchment, including in the social welfare domain.

Measuring retrenchment in welfare states  61

CONCLUSION Assessing welfare state retrenchment accurately is a complex matter. Policy dynamics are complex and require attention to ensure that policy elements are not incorrectly juxtaposed, which can lead to misidentification of processes and patterns of policy change. The ‘dependent variable problem’ that occurs in comparative analysis of welfare states is an obstacle that needs to be overcome in order to guide careful consideration of the impact of conceptualization on operationalization and measurement in studies of policy change. To do so requires eschewing a tendency to rely exclusively on quantitative measures of welfare state effort. This chapter outlines these challenges and emphasizes the value of mixed methods and refining taxonomies of policy components and calibrations as a way to circumvent traditional obstacles to comparative analysis and measurement in retrenchment studies. Disaggregated explorations of social policy domains and interactions combined with meticulous case study analyses can help researchers see both the proverbial forest and the trees in a manner that expands our knowledge of, and language for, retrenchment in welfare states, as well as in other institutional and policy domains. Howlett and Cashore’s (2007, 2009) reconceptualization of the dependent variable problem and revision of existing taxonomies regarding policy levels and change can help disaggregate the tremendous variations in approaches and policy activity in this area, and their dynamics over time.

NOTE 1. Macroeconomic indicators were utilized, specifically yearly unemployment rates and growth rates with a one-year lag (Elmelund-Præstekær and Klitgaard 2012). For robustness, they compiled a quantitative dataset dating back to 1953 and conducted bivariate analyses.

REFERENCES Allan, James P., and Lyle Scruggs. 2004. ‘Political partisanship and welfare state reform in advanced industrial societies.’ American Journal of Political Science 48(3): 496–512. Armingeon, Klaus, Kai Guthmann, and David Weisstanner. 2016. ‘Choosing the path of austerity: how parties and policy coalitions influence welfare state retrenchment in periods of fiscal consolidation.’ West European Politics 39(4): 628–47. Bannink, Duco, and Marcel Hoogenboom. 2007. ‘Hidden change: disaggregation of welfare state regimes for greater insight into welfare state change.’ Journal of European Social Policy 17(1): 19–32. Basurto, Xavier, Gordon Kingsley, Kelly McQueen, Mshadoni Smith, and Christopher M. Weible. 2010. ‘A systematic approach to institutional analysis: applying Crawford and Ostrom’s grammar.’ Political Research Quarterly 63(3): 523–37. Bauer, Michael W., and Christoph Knill. 2014. ‘A conceptual framework for the comparative analysis of policy change: measurement, explanation and strategies of policy dismantling.’ Journal of Comparative Policy Analysis: Research and Practice 16(1): 28–44. Baumgartner, Frank R., Christian Breunig, Christoffer Green‐Pedersen et al. 2009. ‘Punctuated equilibrium in comparative perspective.’ American Journal of Political Science 53(3): 603–20. Boeri, Tito, Axel Börsch-Supan, Guido Tabellini, Karl Ove Moene, and Ben Lockwood. 2001. ‘Would you like to shrink the welfare state? A survey of European citizens.’ Economic Policy 16(32): 9–50.  Bridgen, Paul. 2018. ‘The retrenchment of public pension provision in the liberal world of welfare during the age of austerity – and its unexpected reversal, 1980–2017.’ Social Policy & Administration 53(1): 16–33.

62  Handbook on austerity, populism and the welfare state Carter, David P., Christopher M. Weible, Saba N. Siddiki, and Xavier Basurto. 2016. ‘Integrating core concepts from the institutional analysis and development framework for the systematic analysis of policy designs: an illustration from the US National Organic Program regulation.’ Journal of Theoretical Politics 28(1): 159–85. Cashore, Benjamin, and Michael Howlett. 2007. ‘Punctuating which equilibrium? Understanding thermostatic policy dynamics in Pacific Northwest forestry.’ American Journal of Political Science 51(3): 532–51. Castles, Francis G. 1998. Comparative Public Policy: Patterns of Post-war Transformation. Cheltenham, UK and Lyme, NH, USA: Edward Elgar Publishing. Chomsky, Noam. 1957. Syntactic Structures. Berlin: Walter de Gruyter. Clasen, Jochen, and Nico A. Siegel, eds. 2007. Investigating Welfare State Change: The ‘Dependent Variable Problem’ in Comparative Analysis. Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Crawford, Sue E.S., and Elinor Ostrom. 1995. ‘A grammar of institutions.’ American Political Science Review 89(3): 582–600. Del Pino, Eloisa, and Juan A. Ramos. 2018. ‘Is welfare retrenchment inevitable? Scope and drivers of healthcare reforms in five Spanish regions during the crisis.’ Journal of Social Policy 47(4): 701–20. Elmelund-Præstekær, Christian, and Michael Baggesen Klitgaard. 2012. ‘Policy or institution? The political choice of retrenchment strategy.’ Journal of European Public Policy 19(7): 1089–107. Elmelund-Præstekær, Christian, Michael Baggesen Klitgaard, and Gijs Schumacher. 2015. ‘What wins public support? Communicating or obfuscating welfare state retrenchment.’ European Political Science Review 7(3): 427–50. Esping-Andersen, Gosta. 1990. The Three Worlds of Welfare Capitalism. Princeton, NJ: Princeton University Press. Esping-Andersen, Gosta, Duncan Gallie, Anton Hemerijck, and John Myles. 2002. Why We Need a New Welfare State. Oxford: Oxford University Press. Fukuyama, Francis. 2013. ‘What is governance?’ Governance 26(3): 347–68. Green-Pedersen, Christoffer. 2004. ‘The dependent variable problem within the study of welfare state retrenchment: defining the problem and looking for solutions.’ Journal of Comparative Policy Analysis: Research and Practice 6(1): 3–14. Green-Pedersen, Christoffer. 2007. ‘More than data questions and methodological issues: theoretical conceptualization and the dependent variable “problem” in the study of welfare reform.’ In Jochen Clasen and Nico A. Siegel (eds), Investigating Welfare State Change. The ‘Dependent Variable Problem’ in Comparative Analysis. Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing, pp. 13–23. Green-Pedersen, Christoffer, and Markus Haverland. 2002. ‘Review essay: the new politics and scholarship of the welfare state.’ Journal of European Social Policy 12(1): 43–51. Greve, Bent. 2020. Austerity, Retrenchment and the Welfare State. Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Hall, Peter A. 1993. ‘Policy paradigms, social learning, and the state: the case of economic policymaking in Britain.’ Comparative Politics 25(3): 275–96. Heclo, Hugh. 1976. ‘Conclusion: policy dynamics.’ In The Dynamics of Public Policy: A Comparative Analysis. London: Sage, pp. 237–66. Hogan, John, and Michael Howlett, eds. 2015. Policy Paradigms in Theory and Practice: Discourses, Ideas and Anomalies in Public Policy Dynamics. Basingstoke: Palgrave Macmillan. Howlett, Michael. 2009. ‘Governance modes, policy regimes and operational plans: a multi-level nested model of policy instrument choice and policy design.’ Policy Sciences 42(1): 73–89. Howlett, Michael, and Benjamin Cashore. 2007. ‘Re-visiting the new orthodoxy of policy dynamics: the dependent variable and re-aggregation problems in the study of policy change.’ Canadian Political Science Review 1(2): 50–62. Howlett, Michael, and Benjamin Cashore. 2009. ‘The dependent variable problem in the study of policy change: understanding policy change as a methodological problem.’ Journal of Comparative Policy Analysis 11(1): 33–46.

Measuring retrenchment in welfare states  63 Jahn, Detlef. 2018. ‘Distribution regimes and redistribution effects during retrenchment and crisis: a cui bono analysis of unemployment replacement rates of various income categories in 31 welfare states.’ Journal of European Social Policy 28(5): 433–51. Korpi, W., and J. Palme. 2007. The social citizenship indicator program (SCIP). Swedish Institute for Social Research, Stockholm University. Kühner, Stefan. 2007. ‘ESPA net/JESP Doctoral Researcher Prize Essay: Country-level comparisons of welfare state change measures: another facet of the dependent variable problem within the comparative analysis of the welfare state?’ Journal of European Social Policy 17(1): 5–18. Lindblom, Charles E. 1959. ‘The science of muddling through.’ Public Administration Review 19(2): 79–88. Mortensen, Peter B. 2005. ‘Policy punctuations in Danish local budgeting.’ Public Administration 83(4): 931–50. OECD. 2020. How’s Life? 2020: Measuring Well-being. Paris: OECD Publishing. Ostrom, Elinor, and Xavier Basurto. 2011. ‘Crafting analytical tools to study institutional change.’ Journal of Institutional Economics 7(3): 317–43. Pierson, Paul. 1994. Dismantling the Welfare State?: Reagan, Thatcher and the Politics of Retrenchment. Cambridge: Cambridge University Press. Pierson, Paul. 1996. ‘The new politics of the welfare state.’ World Politics 48(2): 143–79. Pierson, Paul. 2000. ‘Increasing returns, path dependence, and the study of politics.’ American Political Science Review 94(2): 251–67. Pierson, Paul. 2002. ‘Coping with permanent austerity: welfare state restructuring in affluent democracies.’ Revue française de sociologie 43(2): 369–406. Robinson, Scott E., and Floun’say R. Caver. 2006. ‘Punctuated equilibrium and congressional budgeting.’ Political Research Quarterly 59(1): 161–6. Robinson, Scott E., Floun’say Caver, Kenneth J. Meier, and Laurence J. O’Toole Jr. 2007. ‘Explaining policy punctuations: bureaucratization and budget change.’ American Journal of Political Science 51(1): 140–50. Rose, R. 1976. The Dynamics of Public Policy: A Comparative Analysis, London: Sage. Schlüter, Achim, and Insa Theesfeld. 2010. ‘The grammar of institutions: the challenge of distinguishing between strategies, norms, and rules.’ Rationality and Society 22(4): 445–75. Scruggs, L. 2004. Welfare State Entitlements Data Set: A Comparative Institutional Analysis of Eighteen Welfare States, Version 2.0. Storrs: University of Connecticut. Scruggs, Lyle. 2013. ‘Measuring and validating social program replacement rates.’ Journal of European Public Policy 20(9): 1267–84. Siddiki, Saba, Tanya Heikkila, Christopher M. Weible et al. 2019. ‘Institutional analysis with the institutional grammar.’ Policy Studies Journal. https://​doi​.org/​10​.1111/​psj​.12361. Simon, Herbert A. 1957. Administrative Behavior: A Study of Decision-making Processes in Administrative Organizations. New York: The Macmillan Company. Spicker, Paul. 2018. ‘The real dependent variable problem: the limitations of quantitative analysis in comparative policy studies.’ Social Policy & Administration 52(1): 216–28. Starke, Peter. 2006. ‘The politics of welfare state retrenchment: a literature review.’ Social Policy & Administration 40(1): 104–20. United Nations. 2014. Prototype Global Sustainable Development Report. Division for Sustainable Development, United Nations Department of Economic and Social Affairs. Wenzelburger, Georg. 2011. ‘Political strategies and fiscal retrenchment: evidence from four countries.’ West European Politics 34(6): 1151–84. Wenzelburger, Georg, Reimut Zohlnhöfer, and Frieder Wolf. 2013. ‘Implications of dataset choice in comparative welfare state research.’ Journal of European Public Policy 20(9): 1229–50.

6. The dependent variable problem revisited: methods, concepts, and scope in the welfare retrenchment literature Mehmet Fuat Kına and Erdem Yörük

INTRODUCTION “What is welfare state retrenchment, and how can it be measured?” (Green-Pedersen 2004, 3). This question refers to the well-acknowledged “dependent variable problem” (DVP hereafter). The DVP has triggered a very productive domain of empirical and theoretical discussions among welfare state scholars, particularly since the seminal work of Paul Pierson on the new politics of welfare (Pierson 1994, 1996; Green-Pedersen 2004; Clasen and Siegel 2007; Wenzelburger et al. 2013; Kühner 2015; Spicker 2018; Otto and van Oorschot 2019; see also Howlett and Cashore 2009 for public policy). The basic definition of the problem in comparative social policy literature is to find the best indicator of the welfare state change. Despite a silent theoretical consensus on the use of social rights as the best choice over expenditures (Kühner 2015; Bölükbaşı and Öktem 2018), our analysis points out that expenditure is still the most commonly used indicator to represent and analyse welfare retrenchment. Yet, a binary answer is not enough to cover the complexity of the welfare state and the welfare state change has been designated with different terms, such as reform, retrenchment, restructuring, recalibration, recommodification (Clasen and Siegel 2007, 10). It has also turned out that the choice of the dependent variable has influenced whether scholars conclude that Western welfare states retrenched, remained unchanged, or expanded during the neoliberal period. In addition, the choice of the dependent variable has affected not only the characterization of the change but also the identification of the impact of specific explanatory variables in question. In this chapter, we are following in the footsteps of Green-Pedersen (2004), who was the first scholar to systematically study the DVP. Green-Pedersen argued that DVP is more than an empirical measurement problem but an issue of conceptualization – whether welfare state retrenchment should be theorized as an institutional change or entitlement cutbacks. Retrenchment is, accordingly, an inseparable part of how the welfare state has itself been theorized. Whether it is a governmental intension to protect the social well-being of citizens, or how governments successfully achieve that. Is the welfare state a policy or an outcome? Similarly, is the retrenchment of welfare state a legislative transformation in the duration of unemployment benefits or smaller life expectancy of unemployed people? Starting from this point, Green-Pedersen claimed that the operationalization of retrenchment should be in line with the specific conceptualization that is undertaken, since divergences in the conceptualization lead to divergences in the measurement problem and hence to the DVP. However, in addition to Green-Pedersen’s argument, our analysis of the literature reveals that DVP may not only be a conceptualization problem but also a measurement problem to the extent that scholars suffer from the availability and accessibility of data. It is a measurement problem 64

The dependent variable problem revisited  65 especially when scholars look for the impact of a specific independent variable of interest in their statistical models. They may likely be more interested in the theoretical literature on this variable, instead of the welfare state as the dependent variable. This chapter is based on a systematic literature review of welfare retrenchment studies that have been published after those reviewed by Green-Pedersen (2004). The systematic review, in which previous studies are classified in a reproducible and analytical manner, enables us to go beyond the most cited or most influential publications and to account for the dependent variable choices of a wide piece of literature. We can understand the extent to which the theoretical consensus on the use of social rights as the primary dependent variable has materialized. This analysis allows us to determine to what extent the DVP has been resolved.

THE DEPENDENT VARIABLE PROBLEM The new politics thesis of Pierson (1994) states that the mechanisms of welfare state expansion are not replicated during the austerity period. In times of “austerity,” established institutions of the welfare state and vested interests of citizens can contribute to welfare state resilience. Partisan politics and power-resource mobilization arguments (Esping-Andersen 1985; Korpi 1989) do not have enough explanatory capacity to understand the welfare state preferences of policy-makers during the neoliberal period since politicians from both left or right have become interested in maintaining the welfare state because of electoral competition. By looking at the social expenditures of governments, Pierson preferred the terms “restructuring” and “recalibration” (2001) rather than retrenchment to characterize welfare reform, due to this “blame avoidance” mechanism (1996). Since then, many scholars have tested the argument of Pierson against the impact of partisan politics, concluding with confirmations (Castles 2001) and sharp criticisms regarding methodology and results (Korpi and Palme 2003; Allan and Scruggs 2004). While scholars have agreed on the existence of welfare reform in the Western world, they have sharply disagreed on the measurement and direction of the welfare change, and the effect of partisan politics. Pierson’s argument has triggered the debate on DVP as it was a frontal rejection of Esping-Andersen’s ground-breaking emphasis on social rights and the concept of decommodification instead of expenditures (1990). However, as much as the new politics thesis has enjoyed a good reputation, it has also attracted criticisms regarding its methodology. These criticisms have explicitly emphasized the validity of the retrenchment argument and proposed new datasets on different dimensions of social rights. The most severe challenge for social policy scholars following the social rights approach was that the only available data on the comparative welfare state (Organisation for Economic Co-operation and Development (OECD) and EUROSTAT data) was related to social expenditure. As one of the most utilized right-based datasets, Social Citizenship Indicator Program (SCIP) was developed by Esping-Andersen and Korpi, highlighting the Marshallian “social citizenship” formulation, and updated up until the 2010s by various scholars, without being made publicly available. Then, another dataset on social rights, developed by Lyle Scruggs, Comparative Welfare Entitlements Dataset (CWED) (Scruggs 2004), replicated right-based indicators of the SCIP. These empirical emphases on right-based indicators depend on a similar welfare state conceptualization that takes providing social rights to the citizens as the principal duty of the welfare

66  Handbook on austerity, populism and the welfare state state, just as the state is responsible for providing political and civil rights such as rights of voting or assembly (Stephens 2010). Since social expenditure levels could be affected by a wide range of societal needs like aging or fluctuations in unemployment rates, the DVP discussions tended to end up with a preference on social rights (Clasen and Siegel 2007). While expenditure is an overall output measure at the governmental level, even in disaggregated forms, right-based (entitlement) indicators measure how these macro-level factors change the real lives of individuals. More explicitly, expenditure variables stand for the amount of budgets allocated, without telling much about how it is spent. For example, the number of unemployed or aged people or variations of tax systems are often ignored by the providers of expenditure data (Kühner 2015). Aggregate social expenditure data generally include the following areas: old age (e.g., pension, home-help and residential services), family (e.g., family allowances, parental leave), unemployment (e.g., unemployment benefit), health (e.g., spending on patient care, medical goods), active labor market programs (e.g., training, direct job creation), housing (e.g., housing assistance, rent subsidies), and social assistance (e.g., conditional cash benefits, food stamps), and they mostly exclude education spending (Adema and Fron 2019). However, moving beyond the declared amount of budgets, social rights have been regarded as more direct measures of how people are successfully protected from financial and social risks such as unemployment, sickness, and old age (Scruggs and Allan 2006). The quality of social rights can be understood under these three conditions by asking the following questions: how long do governments provide the benefits; how universally are these benefits distributed; how accessible are they; what are the conditions that one should satisfy to become eligible; and how much of the actual wage is compensated? Since scholars favoring social rights assert that their methodology covers more complicated dimensions of the welfare state, they conceptualized their measurement as “welfare generosity” (Allan and Scruggs 2004; Scruggs 2008), while earlier generations of scholars had defined social expenditure as “welfare effort” (Amenta 2003). In this sense, effort and generosity respectively correspond to traditional conceptual equivalents of expenditure and rights. More specifically, unlike expenditure levels, the right-based approach often uses replacement rates as dependent variables, in addition to eligibility rules and durations. Replacement rates were mainly developed as institutional indicators for welfare generosity (Scruggs 2008), and they are based on a mathematical calculation on how much the payment compensates the wage level of the average production worker under the conditions of unemployment, sickness, or work accident. The eminent challenges against the new politics thesis from right-based datasets were in utilizing replacement rates for different welfare programs (unemployment and sickness insurance) as dependent variables (Korpi and Palme 2003; Allan and Scruggs 2004). Unlike Pierson, scholars found a retrenchment in social rights since the 1980s, and that the formally organized power of the right had played a significant role. Thus, both the direction of welfare state change and its main triggers are largely determined by the dependent variable choice of scholars. Kühner (2015) pointed out how expenditure and right-based indicators could lead to different conclusions, as one of retrenchment, resilience, or expansion. Consistent with the previous bifurcation on the DVP, the expenditure-based calculation of the welfare state change for the 21 OECD countries seems to bring more positive results on average. From 1980 to 2013, the average ratio of welfare state expenditure over gross domestic product (GDP) in these countries had expanded by 7.7 percent. The change has a negative sign only for the Netherlands

The dependent variable problem revisited  67 with –0.5, and Portugal has the highest growth rate with 16 percent (see table 2 in Kühner 2015). However, for the change of right-based measures in the same period, Sweden has the worst performance with –8.9 percent, while Ireland has the most substantial expansion with 9.5 percent (see table 3 in Kühner 2015). The critical points are that the amount of change of the right-based indicators has a larger standard deviation, and its confidence interval includes zero, which makes it difficult to assume a general trend. As shown in this example, right-based measures frequently present more mixed results, while scholars favoring social rights argue for a more complex picture of the welfare state. That is also why the use of social expenditure has been criticized in the DVP debates as overly optimistic about the change of welfare state, and its detrimental effects on individual lives. On the one hand, covering more qualitative dimensions of welfare needs, the right-based approach provided more realistic numbers for the “real change.” On the other hand, it was not safe from similar harsh criticisms, especially concerning the replacement rates. Some other scholars argued that the denominator of replacement rates, the real wage, could also be biased with uncontrollable effects of macro-economic trends. Therefore, the replacement rate may be determined by factors other than government decisions (Jensen 2011). Another problem other than omitted factors is the time-lag between government intentions and the actual change in the conditions of welfare state beneficiaries. It is almost impossible to calculate the real time-lag because institutional frameworks can impose stability to varying, unmeasurable degrees (Green-Pedersen 2004). Similarly, replacement rates are more applicable to some areas of the welfare state, such as unemployment. This is because of the reduced time-lag expected between the work and the benefit. However, for pension benefits, it gets harder to provide a standardized measure since it is more likely to be affected by various policy reforms during the lagged time (Wenzelburger et al. 2013). Given that scholars, who preferred to use the expenditure data, often use it as a ratio of the GDP, the risk of being affected by omitted macro-trends and undefined time-lags are similar problems for both expenditure and replacement rate variables. Notwithstanding the theoretical consensus on the use of right-based variables, the DVP has escalated as a debate between the proponents of right-based variables and those of new alternatives. An essential part of this debate is the issue of consistency among different right-based indicators. Wenzelburger et al. (2013) argued that replacement rates in SCIP and CWED do not measure the same dependent variable. They found “substantial divergences” in replacement rates among SCIP and CWED, which “have serious consequences for the explanation of retrenchment dynamics” (2013, 1230). They did not explain why divergences exist but questioned the ambiguity of the replacement rate itself. Scholars who developed SCIP and CWED have responded differently to this critique of data mismatch. While Scruggs (2013) from CWED argued that non-corresponding data points might be incidental and related to inadvertent mistakes, Ferrarini et al. (2013) from SCIP contended that there are theoretical divergences behind the two data measurement processes. By illustrating that “replacement rates do not exist readymade ‘out there,’ just waiting to be discovered,” the latter argued these two datasets intendedly measure two different dimensions of the welfare state (Ferrarini et al. 2013, 1251). Although replacement rates in the following editions of SCIP and CWED demonstrated less divergence, Bölükbaşı and Öktem (2018) found another significant divergence in their analysis on non-replacement rate variables such as benefit durations, qualification periods, waiting days, and coverages. Therefore, they concluded that the choice of dataset would still lead to different interpretations of welfare state

68  Handbook on austerity, populism and the welfare state changes, even looking at the same factors. This extension of the debate into non-replacement rate indicators showed the DVP remains, yet again, even with high quality datasets. Against the backdrop of the shortcomings of both expenditure and right-based variables, coding legislative reforms on specific types of programs have recently appeared as a third option in the welfare state literature (Clasen and Clegg 2007; Däubler 2008; Klitgaard et al. 2013; Immergut and Abou-Chadi 2014; Riekhoff 2017). This effort has led to a comparative welfare state database on employment protection legislation: fRDB-IZA Social Reforms Database (Debenedetti 2012). The primary contribution of looking at legislative processes is to have a better-directed idea about the governmental intention, looking for input instead of output measures. Nonetheless, since this approach is not very common in the quantitative welfare state literature, as well as for more structural reasons, coding preferences of scholars have less standardized designs. How to codify a reform in a binary or multinomial format is the central question. What is more, while studying the input variables, it becomes more difficult for comparative policy researchers to develop universal measures such as expenditure levels or benefit durations. From the expansion of specific pension programs to the distribution of benefits, there are various dimensions of policy reforms that reflect the multidimensional character of the welfare state. Nevertheless, easy availability of expenditure data from widely known international sources (mostly by OECD or EUROSTAT), more recent availability of replacement rates after the release of CWED, and the time-consuming character of collecting and coding information about legislations might infer that the interest in input indicators like legislative indices remains limited. Scholars might not prefer policy reforms, as input values, if they are more likely to focus on the actual change in people’s well-being at the end of the day. Quantitative macro measures might fail to picture institutional transformation in detail, and scholars need “qualitative leitmotifs of welfare state change” (Clasen and Clegg 2007). Using fuzzy sets has also been perceived as an alternative for this reason (Kvist 2007; Vis 2009).

THE SYSTEMATIC REVIEW OF THE STATE-OF-THE-ART RETRENCHMENT LITERATURE In order to understand the trends in the most recent literature on retrenchment, we conducted a systematic review of the empirical studies, utilizing the advanced search option of a widely used online scholar search engine (Scopus). We searched for a reproducible string, including “welfare state” and “retrenchment” words, in the titles, abstracts, and keywords of previous publications. We extracted the abstracts of 506 papers to carry out an abstract-level classification with Rayyan, which is standard software used in a systematic literature review. All quantitative analyses that took the welfare state as the dependent variable are included in this classification. Studies belonging to the excluded majority are those that focused on the impacts of retrenchment or did not employ a quantitative relationship. After we reviewed full texts, one more time, we excluded the articles that do not try to explain the welfare state. Our final dataset consists of 35 papers. We coded each of them concerning their dependent variable, the dataset in charge, the scope of the dependent variable, the independent variable of interest, their basic findings, and the countries and the time range of analysis.

The dependent variable problem revisited  69 Type of Dependent Variable We observed three types of dependent variables in the welfare state literature. These are expenditure, right-based, and legislative variables. While more than half of these studies (20) use expenditure variables, we see right-based dependent variables in 15 papers. However, legislative variables fall far behind, with only four papers.1 The timely transformation of the dependent variable choices of scholars is presented in Figure 6.1. As can be seen, even after the release of open-access databases on the right-based variables and severe criticisms against the expenditure variables, scholars still largely continue to rely on expenditure levels. On the one hand, the use of right-based variables seems to close the gap during the 2010s. For the most recent time range (2015–19), the number of papers employing right-based and expenditure variables are equal. On the other hand, until 2015, the expenditure variable had retained its position as the most common dependent variable.2

Figure 6.1

Dependent variable type, by year

Data Sources Overall, the OECD and CWED are very much dominant in the literature.3 While 14 of 20 expenditure variables come from the OECD, 12 of 15 right-based variables belong to CWED (Figure 6.2 shows the comparison by year). The increase in the interest in CWED (and right-based variables in Figure 6.1) after the 2010s might reflect the escalation of debate regarding the DVP. While the CWED variables constituted the most popular alternative for widely criticized expenditure variables, the interest in OECD expenditure data surprisingly remained intact during the 2000s.

70  Handbook on austerity, populism and the welfare state

Figure 6.2

Comparison of the utilization of OECD and CWED datasets by year

Level of Data Aggregation Considering the emphasis of DVP debates on the importance of using disaggregated measures, we also present the distribution of the dependent variables according to whether they are in aggregated or disaggregated form (i.e., program-level data), independently of expenditure or right-based variable choice (Figure 6.3). There has been a remarkable convergence in the literature as the use of aggregated measures significantly decreased after the late 2000s. Discipline of the Journal We also asked how these distributions of dependent variable characteristics vary according to the discipline of the journal. We assumed that social policy journals represent a theoretically more engaged approach to the DVP. By doing so, we tested whether or not this theoretical engagement would crystallize an empirical consensus on the use of right-based variables. Studies published in the social policy journals analysed right-based variables more than the other studies. However, contrary to our expectations, being published in a social policy journal does not result in any additional variation in terms of whether expenditure variables, social rights, OECD datasets, or CWED are used. Therefore, we fail to assert that the DVP has been more likely resolved in the core social policy circles. However, the ratio of publications that use disaggregated dependent variables is much higher in social policy journals. In eleven papers published in social policy journals, only three use aggregated measures, while the ratio of aggregated articles is 11/24 in non-social policy journals. Higher usage of disaggregated variables as both expenditure- and right-based suggests that the more theoretical engagement to the problem, the more detailed approach for the scope of the dependent variable. In compar-

The dependent variable problem revisited  71

Figure 6.3

The scope of dependent variables, by year

ison with publications in non-social policy journals, social policy journals are investigating the welfare state change with more in-depth analyses. Temporality We have observed that scholars usually trace their analyses of retrenchment back to the 1980s; 8 of 35 analyses start from 1980 as the first year of retrenchment, and most others the mid-1980s. This is because the beginning of the welfare state reform has mostly been identified with the financial crises of the 1980s. Case Selection There has been remarkable stability among studies published in the 2000s. Most studies analyse OECD-18 countries. The only exception until 2017 analysing more than 20 countries is Dreher et al. (2008), as an outlier with 60 countries, who utilized OECD and World Bank expenditure data in aggregated form. Nonetheless, after 2017 there seems to be a tiny increase in the number of countries included. The widening scope of the latest versions of datasets (see Scruggs et al. 2014) again brings us to the data availability question.

THE ONTOLOGY AND DYNAMICS OF RETRENCHMENT As mentioned in the introduction, there were two main questions regarding the DVP. First, whether or not retrenchment has taken place, and second, whether or not partisan politics still matters. Yet, more than half of the scholars did not even ask the first retrenchment question in their analyses. On the contrary, retrenchment has more often been taken for granted. While

72  Handbook on austerity, populism and the welfare state before 2010 more than half of the studies (7/11) tried to understand whether the welfare state retrenched, expanded, or remained, only one-third of analyses in the 2010s (8/24) make any argument on the ontology of the retrenchment. Retrenchment has increasingly become a descriptive term characterizing a period, rather than a research question (see Dahlström 2011; Brady and Lee 2014; Tromborg 2014; Soroka et al. 2016; Baek et al. 2017). Our analysis of scholarly responses to the first question is very telling. Among 15 papers that asked the first retrenchment question, “there-is-no-retrenchment” argument of Pierson is only supported by two papers. Only one analysis in the early 2000s concluded that there is no retrenchment at all (Castles 2001), and another one found a statistically significant retrenchment effect, which then vanished in robustness checks (Armingeon et al. 2016). The critical point here is that both papers utilized expenditure levels in aggregated forms as dependent variables, with the data from the OECD (Table 6.1). On the other hand, there is no consensus in the “there-is-retrenchment” camp either. Following the work of Korpi and Palme, and Allan and Scruggs, there are only two papers that found total retrenchment (Tepe and Vanhuysse 2009; Jordan 2011). More importantly, nine analyses using expenditure, right-based or legislative dependent variables showed that there is no universal pattern anymore, but only partial retrenchment, or a mix of expansion and retrenchment, which vary across countries, years, and even programs (Hicks and Zorn 2005; Immergut and Abou-Chadi 2014; Wolf et al. 2014; Bruch and White 2018; Jahn 2018). When it comes to the second question, all but three papers are either controlling for (12) or directly analysing (20) the partisan impact in their empirical analyses. Among scholars who hypothesized the impact of partisan politics, we observed four different clusters of findings: significant effect, partial-weak significance, no significant result, and significant effect in the opposite direction of the power-resources argument. The first and second clusters are more dominant in the literature – 15 of these 20 papers fall under these two clusters, with almost similar weights. More importantly, there is an increasing trend to find a more complicated impact of partisan politics (Vis 2009), which is structured by or interacting with the use of political strategies (Jensen et al. 2014), party coalitions (Röth et al. 2018), electoral competitions (Abou-Chadi and Immergut 2019), veto player constellations (Becher 2010), or even the welfare state generosity itself (Starke et al. 2014). The third and fourth clusters of findings are consistent with the new politics thesis to varying degrees. While some of the scholars found insignificant relationships or argued that there is a declining impact of partisanship during the retrenchment period (Jordan 2011; Wolf et al. 2014; Jakobsson and Kumlin 2017), others did not find any impact (Däubler 2008; Kail and Dixon 2011). On the other hand, there are still significant findings that show that left-wing governments limit the retrenchment of the welfare state (Klitgaard and Elmelund-Præstekær 2013; Riekhoff 2017; Bruch and White 2018). Finally, a small group of scholars, following the argument of Ross (2000), have found pieces of evidence to reverse the partisan logic, by claiming that left-wing governments have more credit to retrench in economically hard times (Bojar 2018; Armingeon et al. 2016) (Table 6.1). As scholars are increasingly taking retrenchment as given, they are also less likely to describe their conceptual framework on the welfare state. This might be for two reasons. First, scholars might theoretically be more interested in one of their independent variables than the dependent variable (see as an example Xu 2017; the author theoretically focused on the impact of immigrant influx). Second, recent literature might have a conceptual consensus on

Number of papers:

Legislative

4

2

9

7

2013

Klitgaard & Elmelund-Præstekær

Riekhoff 2017,

Bruch & White 2018

2014

Starke et al. 2014 Immergut & Abou-Chadi

Bruch & White 2018,

Abou-Chadi &

& Scruggs 2004

2017, Jahn 2018

2009

Röth et al. 2018,

& Palme 2003, Allan

Jakobsson & Kumlin

8

al. 2014

Vis 2009, Starke et

Immergut 2019

Jensen et al. 2014,

Wolf et al. 2014,

2000

Becher 2010, Korpi

Jahn 2018

2016

Kittel & Obinger 2003, Jordan 2011

Huber & Stephens

Significant

3

Däubler 2008

Kumlin 2017

Jakobsson &

Kail & Dixon 2011

Partisan Impact: 20 Partially Not Significant Significant

Wolf et al. 2014,

Kittel & Obinger 2003,

Hicks & Zorn 2005,

Retrenchment

Armingeon et al.

Castles 2001,

No Retrenchment

Mixed or Partial

Tape & Vanhuysse

Allan & Scruggs 2004

Korpi & Palme 2003,

Jordan 2011

Retrenchment

Retrenchment: 15

Analysis of existing studies concerning the dependent variable and retrenchment

Expenditure and Right-based

Right-based

Expenditure

Dependent Variable

Table 6.1

2

al. 2016

Bojar 2018, Armingeon et

direction

Significant in the opposite

The dependent variable problem revisited  73

74  Handbook on austerity, populism and the welfare state the right-based definition of the welfare state, as Bölükbaşı and Öktem (2018, 88) claimed. However, our findings show that this consensus has not empirically materialized.

DISCUSSION AND CONCLUSION Looking at the dependent variable choices of welfare retrenchment scholars has enabled us to see the details of the type of dependent variables, data sources, case selection, level of data aggregation, discipline of the journal that published the work, temporality, and the conclusions on the two prominent research questions: the existence of retrenchment and the impact of partisan politics on retrenchment. First, the prevailing case selection strategy has mostly been borrowed from earlier generations of studies. Except for one study, scholars did not include any country from the Global South. However, recent findings have shown that since the 2000s, there has been a great deal of welfare state changes that take place in the non-West, especially in those countries that are classified as emerging markets, such as Brazil, China, India, Turkey, South Africa, and so on (ILO 2014; World Bank 2014). Welfare state expansion and/or partial retrenchment in the Global South should have been reflected in the existing literature on the welfare reform, with a greater sample of countries being analysed. Scholars have considered retrenchment mostly as a Western notion, just like the very concept of “the welfare state” itself (Takegawa 2009; Ferragina and Seeleib-Kaiser 2011; Vrooman 2012; Kim 2015). While most countries in the Global South have expanded new social assistance and healthcare programs for the poor, there are significant retrenchments, especially in terms of social rights and generosity in traditional social security programs for formal sector employees. Scholars did not pay much attention to retrenchment in the Global South, and this is most likely related to a lack of comparative data on welfare state generosity. Quantitative and explanatory studies on retrenchment have investigated the impact of political factors by either controlling for or directly investigating the effect of partisan politics. This literature has mostly gone beyond the structuralist paradigm that in the main focuses on social needs and governmental capacities like aging, unemployment rates, Gini, GDP, trade openness to explain social policies (see the systematic review of the social assistance literature by Yörük et al. (under review), which illustrates that only a small minority of studies (less than 1 percent) have analysed contentious politics to explain social assistance programs). However, scholars often limited their political models to formal party politics, partisanship, and ideology. Except for Bailey (2015), who analysed the effect of disruptive protests, studies in our sample did not consider the effect of social movements. Furthermore, in recent years, retrenchment has generally been taken for granted without conducting any empirical analysis. Scholars constructed analytical models mostly to investigate the reasons for the welfare reform, “why and under what conditions welfare state expansion, and retrenchment takes place” (Ewalt and Jennings 2014, 308), without making arguments as to whether or not there is actual retrenchment or expansion. As such, retrenchment has been taken for granted as a name for the period, rather than as a government policy to cutback entitlements or decrease spending. In the earlier periods of the retrenchment literature, there was a correspondence between the choice of the dependent variable and the conclusion regarding the ontology and dynamics of the retrenchment. Scholars that favored expenditure variables were much more inclined to conclude that “there is no retrenchment and no significant partisan effect” than those

The dependent variable problem revisited  75 rights-based scholars who were to claim “there are an obvious retrenchment and a significant partisan effect.” We observed that this earlier correspondence between methodological preference and scientific conclusion has seemed to vanish. There is now much more diversity in conclusions from studies that use any type of dependent variable (Table 6.1). Yet, whether or not this diversification of conclusions is a positive scientific development is debatable. There is a need for more robustness checks, different model specifications, and even alternative types of dependent variables. However, it is possible that the recent increase in the number of partial results like the mixed impact of partisan politics or diversified retrenchments might also be reflecting the actual diversification in politics and social policy. One of the most significant findings of our systematic review is that expenditure as the dependent variable is still the most dominant preference in the literature, while the popularity of right-based variables is increasing in parallel to the available datasets that contain generosity variables (Figures 6.1 and 6.2). This contradicts the assumption of scholars discussing the DVP that there is a consensus starting in the late 2000s on the use of right-based dependent variables. The OECD is still significantly preferred, due to its considerable reputation as one of the most credible international and institutional data offices which sustains the popularity of expenditure variables. Moreover, while generosity, most of the time, corresponds to replacement rates in empirical models, the concept of welfare effort represents a theoretical equivalent of expenditure variables (when scholars explicitly refer to these conceptual equivalents). Therefore, another part of the problem is the conceptual distance between welfare generosity and replacement rate, which has been wider than the conceptual distance between welfare effort and expenditure. In short, expenditure variables are easier to theorize and access. However, the usage of expenditure variables is very problematic, at least more problematic than the right-based variables, due to their incapacity to account for the societal needs and categorical complications of the welfare state services. Older countries, for example, have to spend more on old-age pensions by nature. Likewise, fluctuations in unemployment levels would change the total budget used for unemployed people, independently of the eligibility criteria. Considering this wide interest in expenditure variables, we argue that there is a gap between the ongoing debate on the DVP and the actual DVP in the deeper contours of the welfare state (change) literature. As another example of this gap, we could not see any recent interest in the SCIP, at least by scholars in our sample, although the ongoing DVP discussion is focused on the comparison of the two most credible right-based datasets (CWED and SCIP) to determine whether certain variables are consistent in these two datasets. Most scholars work with CWED if they utilize right-based variables. The reason again could be availability and explanatory capacity. The SCIP dataset is only available for every five years until 1995 and for 18 OECD countries. CWED is available for 33 countries for each year until 2005. Furthermore, note that SCIP became publicly available only after the release of CWED. Implying a similar difference between welfare effort and generosity, scholars did not construct their analyses on the decommodification concept of Esping-Andersen (1990) anymore, although its conceptualization is very similar to generosity, although Scruggs (2007) himself appreciated the concept of decommodification as a good indicator of benefit generosity. Decommodification is a very loaded concept, referring to how much the working class has achieved historical gains. However, generosity is a much less ambitious concept and requires less explanation than decommodification.

76  Handbook on austerity, populism and the welfare state

Figure 6.4

Availability and explanatory capacity order of the three concepts (from lower to higher)

Thus, we observed a ranking of availability and explanatory capacity of the three concepts: from welfare effort to generosity, and from generosity to decommodification, as shown in Figure 6.4. The changes in the availability of both expenditure and right-based measures and the efficiency of the concepts might have shifted the earlier dichotomy between the use of datasets, which was SCIP versus OECD, and more recently has changed to CWED versus OECD. In our analysis, we found a (limited) consensus on the use of disaggregated dependent variables (Figure 6.3). Except for this, there is still large dissensus concerning other dimensions of the DVP. Moreover, the distribution of choices on dependent variable types and the datasets are not changed when we control for the discipline of the journal where studies are published. These findings lead us to think that the most recent DVP is not in conceptualization and theoretical interest, in contrast with the findings of Green-Pedersen (2004). It is also about the availability and explanatory capacity datasets. Therefore, DPV is both a problem of conceptualization and measurement. We came to this conclusion by widening the scope of DVP discussion beyond mostly cited papers and carrying out a systematic literature review analysis to make sure we did not miss any corner of the field. Focusing only on the most famous scholars and journals is a common problem for many other literature reviews. However, systematic collection and classification of previous literature made it easier for us to handle this problem and reveal the state of the ongoing debate on the welfare state change as the dependent variable.

The dependent variable problem revisited  77

NOTES 1. The total becomes 39, since both expenditure and rights are used in four papers, and we counted them in both categories. 2. In addition to the three major dependent variables (expenditure, right-based, and legislative), an emphasis on another alternative has more recently come into the picture – utilizing micro-level survey data (Otto and van Oorschot 2019). The primary motivation is to trace the reflection of policy reforms, based on cash benefit receipt. The suggestion can be justified as driving macro-level variables from micro-level data could mirror “(changing) access to and levels of cash benefits more directly than the hypothetical paper reality of social rights data or the pure cost of social expenditure” (2019, 309). The question appears once more as “how to measure the real change?” rather than being rooted in theoretical differentiation. Despite these ongoing debates, there are some common issues like disadvantages of aggregated data. Even if scholars are interested in macro-level transformations, they prioritize disaggregated data, because retrenchment, resilience, or expansion might co-exist for different types of programs (De Deken and Kittel 2007; Kühner 2015). 3. Though the OECD has some replacement rate data (OECD 2020), scholars do not employ it. We did not observe any analysis that used replacement rate data from the OECD. Scholars utilizing OECD data always use expenditure-based variables. Even if they do use both types of dependent variables, they take expenditure data from the OECD and replacement rates from CWED (Tepe and Vanhuysse 2009; Starke et al. 2014).

REFERENCES Abou‐Chadi, Tarik, and Ellen M. Immergut. 2019. “Recalibrating Social Protection: Electoral Competition and the New Partisan Politics of the Welfare State.” European Journal of Political Research 58 (2): 697–719. Adema, Willem, and Pauline Fron. 2019. The OECD SOCX Manual – 2019 Edition. Paris: OECD. Retrieved June 24, 2020 from http://​www​.oecd​.org/​social/​soc/​SOCX​_Manuel​_2019​.pdf. Allan, James P., and Lyle Scruggs. 2004. “Political Partisanship and Welfare State Reform in Advanced Industrial Societies.” American Journal of Political Science 48 (3): 496–512. Amenta, Edwin. 2003. “What We Know about the Development of Social Policy.” In Comparative Historical Analysis in the Social Sciences, edited by James Mahoney and Dietrich Rueschemeyer, 91–130. New York: Cambridge University Press. Armingeon, Klaus, Kai Guthmann, and David Weisstanner. 2016. “Choosing the Path of Austerity: How Parties and Policy Coalitions Influence Welfare State Retrenchment in Periods of Fiscal Consolidation.” West European Politics 39 (4): 628–47. Baek, Seung‐Ho, Yun‐Kyu Ryu, and Sophia Seung‐yoon Lee. 2017. “The Political Economy of Changes in Family and Old‐age Welfare Policy Spending: Analysis of OECD Countries, 1998–2011.” International Journal of Social Welfare 26 (2): 116–26. Bailey, David J. 2015. “Resistance is Futile? The Impact of Disruptive Protest in the ‘Silver Age of Permanent Austerity’.” Socio-Economic Review 13 (1): 5–32. Becher, Michael. 2010. “Constraining Ministerial Power: The Impact of Veto Players on Labor Market Reforms in Industrial Democracies, 1973–2000.” Comparative Political Studies 43 (1): 33–60. Bojar, Abel. 2018. “Biting the Hand That Feeds: Reconsidering the Partisan Determinants of Welfare Spending in Times of Austerity.” Government and Opposition 53 (4): 621–52. Bölükbaşı, H. Tolga, and Kerem Gabriel Öktem. 2018. “Conceptualizing and Operationalizing Social Rights: Towards Higher Convergent Validity in SCIP and CWED.” Journal of European Social Policy 28 (1): 86–100. Brady, David, and Hang Young Lee. 2014. “The Rise and Fall of Government Spending in Affluent Democracies, 1971–2008.” Journal of European Social Policy 24 (1): 56–79. Bruch, Sarah K., and KaLeigh K. White. 2018. “Politics, State Discretion and Retrenchment in Safety Net Provision: Evidence from the USA in the Post-welfare reform Era.” Cambridge Journal of Regions, Economy and Society 11 (3): 459–83.

78  Handbook on austerity, populism and the welfare state Castles, Francis G. 2001. “On the Political Economy of Recent Public Sector Development.” Journal of European Social Policy 11 (3): 195–211. Clasen, Jochen, and Daniel Clegg. 2007. “Levels and Levers of Conditionality: Measuring Change within Welfare States.” In Investigating Welfare State Change: The “Dependent Variable Problem” in Comparative Analysis, edited by Jochen Clasen and Nico A. Siegel, 166–97. Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Clasen, Jochen, and Nico A. Siegel. 2007. Investigating Welfare State Change: The “Dependent Variable Problem” in Comparative Analysis. Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Dahlström, Carl. 2011. “Who Takes the Hit? Ministerial Advisers and the Distribution of Welfare State Cuts.” Journal of European Public Policy 18 (2): 294–310. Däubler, Thomas. 2008. “Veto Players and Welfare State Change: What Delays Social Entitlement Bills?” Journal of Social Policy 37 (4): 683–706. De Deken, Johan, and Bernhard Kittel. 2007. “Social Expenditure under Scrutiny: The Problems of Using Aggregate Spending Data for Assessing Welfare State Dynamics.” In Investigating Welfare State Change: The “Dependent Variable Problem” in Comparative Analysis, edited by Jochen Clasen and Nico A. Siegel, 72–104. Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Debenedetti, Fondazione Rodolfo. 2012. “fRDB-IZA Social Reforms Database.” Retrieved January 18, 2021 from http://​www​.frdb​.org/​page/​data/​scheda/​frdb​-iza​-social​-reforms​-database/​doc​_pk/​9027. Dreher, Axel, Jan-Egbert Sturm, and Heinrich W. Ursprung. 2008. “The Impact of Globalization on the Composition of Government Expenditures: Evidence from Panel Data.” Public Choice 134 (3–4): 263–92. Esping-Andersen, Gosta. 1985. Politics Against Markets. Princeton, NJ: Princeton University Press. Esping-Andersen, Gosta. 1990. The Three Worlds of Welfare Capitalism. Princeton, NJ: Princeton University Press. Ewalt, J.A.G., and E.T. Jennings Jr. 2014. “The Great Recession and Social Welfare Spending in the American States.” International Review of Public Administration 19 (3): 308–23. Ferragina, Emanuele, and Martin Seeleib-Kaiser. 2011. “Thematic Review: Welfare Regime Debate: Past, Present, Futures?” Policy & Politics 39 (4): 583–611. Ferrarini, Tommy, Kenneth Nelson, Walter Korpi, and Joakim Palme. 2013. “Social Citizenship Rights and Social Insurance Replacement Rate Validity: Pitfalls and Possibilities.” Journal of European Public Policy 20 (9): 1251–66. Green-Pedersen, Christoffer. 2004. “The Dependent Variable Problem within the Study of Welfare State Retrenchment: Defining the Problem and Looking for Solutions.” Journal of Comparative Policy Analysis: Research and Practice 6 (1): 3–14. Hicks, Alexander, and Christopher Zorn. 2005. “Economic Globalization, the Macro Economy, and Reversals of Welfare: Expansion in Affluent Democracies, 1978–94.” International Organization 59 (3): 631–62. Howlett, Michael, and Benjamin Cashore. 2009. “The Dependent Variable Problem in the Study of Policy Change: Understanding Policy Change as a Methodological Problem.” Journal of Comparative Policy Analysis 11 (1): 33–46. Huber, Evelyne, and John D. Stephens. 2000. “Partisan Governance, Women’s Employment, and the Social Democratic Service State.” American Sociological Review 65: 323–42. Immergut, Ellen M., and Tarik Abou‐Chadi. 2014. “How Electoral Vulnerability Affects Pension Politics: Introducing a Concept, Measure and Empirical Application.” European Journal of Political Research 53 (2): 269–87. ILO (International Labour Organization). 2014. “World Social Protection Report 2014/15.” Geneva: ILO. Jahn, Detlef. 2018. “Distribution Regimes and Redistribution Effects during Retrenchment and Crisis: A Cui Bono Analysis of Unemployment Replacement Rates of Various Income Categories in 31 Welfare States.” Journal of European Social Policy 28 (5): 433–51. Jakobsson, Niklas, and Staffan Kumlin. 2017. “Election Campaign Agendas, Government Partisanship, and the Welfare State.” European Political Science Review 9 (2): 183–208. Jensen, Carsten. 2011. “Less Bad Than Its Reputation: Social Spending as a Proxy for Welfare Effort in Cross-national Studies.” Journal of Comparative Policy Analysis: Research and Practice 13 (3): 327–40.

The dependent variable problem revisited  79 Jensen, Carsten, Christoph Knill, Kai Schulze, and Jale Tosun. 2014. “Giving Less by Doing More? Dynamics of Social Policy Expansion and Dismantling in 18 OECD Countries.” Journal of European Public Policy 21 (4): 528–48. Jordan, Jason. 2011. “Health Care Politics in the Age of Retrenchment.” Journal of Social Policy 40 (1): 113–34. Kail, Ben Lennox, and Marc Dixon. 2011. “The Uneven Patterning of Welfare Benefits at the Twilight of AFDC: Assessing the Influence of Institutions, Race, and Citizen Preferences.” The Sociological Quarterly 52 (3): 376–99. Kim, Ki-tae. 2015. “From Worlds to Cases: Case Selection and ‘Other Worlds’ in the Welfare Modelling Business.” Social Policy & Society 14 (2): 309–21. Kittel, Bernhard, and Herbert Obinger. 2003. “Political Parties, Institutions, and the Dynamics of Social Expenditure in Times of Austerity.” Journal of European Public Policy 10 (1): 20–45. Klitgaard, Michael Baggesen, and Christian Elmelund‐Præstekær. 2013. “Partisan Effects on Welfare State Retrenchment: Empirical Evidence from a Measurement of Government Intentions.” Social Policy & Administration 47 (1): 50–71. Korpi, Walter. 1989. “Power, Politics, and State Autonomy in the Development of Social Citizenship: Social Rights during Sickness in Eighteen OECD Countries Since 1930.” American Sociological Review 54 (3): 309–28. Korpi, Walter, and Joakim Palme. 2003. “New Politics and Class Politics in the Context of Austerity and Globalization: Welfare State Regress in 18 Countries, 1975–95.” American Political Science Review 97 (3): 425–46. Kühner, Stefan. 2015. “What If We Waited a Little Longer? The Dependent Variable Problem Within the Comparative Analysis of the Welfare State Revisited.” Social Policy Review 27: 199–224. Kvist, Jon. 2007. “Exploring Diversity: Measuring Welfare State Change with Fuzzy-set Methodology.” In Investigating Welfare State Change: The “Dependent Variable Problem” in Comparative Analysis, edited by Jochen Clasen and Nico A. Siegel, 198–214. Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. OECD (Organisation for Economic Co-operation and Development). 2020. “Social Protection and Well-being.” Retrieved January 18, 2021 from https://​stats​.oecd​.org/​Index​.aspx​?DataSetCode​=​NRR. Otto, Adeline, and Wim van Oorschot. 2019. “Welfare Reform by Stealth? Cash Benefit Recipiency Data and Its Additional Value to the Understanding of Welfare State Change in Europe.” Journal of European Social Policy: 29 (3): 307–25. Pierson, Paul. 1994. Dismantling the Welfare State? Reagan, Thatcher and the Politics of Retrenchment. Cambridge: Cambridge University Press. Pierson, Paul. 1996. “The New Politics of the Welfare State.” World Politics 48 (2): 143–79. Pierson, Paul. 2001. The New Politics of the Welfare State. Oxford: Oxford University Press. Riekhoff, Aart‐Jan. 2017. “The Divergent Logics of Labour Market Reforms in Times of Permanent Austerity.” International Journal of Social Welfare 26 (1): 5–20. Ross, Fiona. 2000. “‘Beyond Left and Right’: The New Partisan Politics of Welfare.” Governance 13 (2): 155–83. Röth, Leonce, Alexandre Afonso, and Dennis C. Spies. 2018. “The Impact of Populist Radical Right Parties on Socio-economic Policies.” European Political Science Review 10 (3): 325–50. Scruggs, Lyle. 2004. Welfare State Entitlements Data Set: A Comparative Institutional Analysis of Eighteen Welfare States, Version 1.1. Storrs: University of Connecticut. Scruggs, Lyle. 2007. “Welfare State Generosity across Space and Time.” In Investigating Welfare State Change: The “Dependent Variable Problem” in Comparative Analysis, edited by Jochen Clasen and Nico A. Siegel, 133–64. Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Scruggs, Lyle. 2008. “Social Rights, Welfare Generosity, and Inequality.” In Democracy, Inequality, and Representation: A Comparative Perspective, edited by C. Anderson and P. Baramendi, 62–90. New York: Russell Sage Foundation. Scruggs, Lyle. 2013. “Measuring and Validating Social Program Replacement Rates.” Journal of European Public Policy 20 (9): 1267–84. Scruggs, Lyle, and James P. Allan. 2006. “The Material Consequences of Welfare States: Benefit Generosity and Absolute Poverty in 16 OECD Countries.” Comparative Political Studies 39 (7): 880–904.

80  Handbook on austerity, populism and the welfare state Scruggs, Lyle, Detlef Jahn, and Kati Kuitto. 2014. Comparative Welfare Entitlements Dataset 2: Version 2014-03. University of Connecticut and University of Greifswald. Soroka, Stuart N., Richard Johnston, Anthony Kevins, Keith Banting, and Will Kymlicka. 2016. “Migration and Welfare State Spending.” European Political Science Review 8 (2): 173–94. Spicker, Paul. 2018. “The Real Dependent Variable Problem: The Limitations of Quantitative Analysis in Comparative Policy Studies.” Social Policy & Administration 52 (1): 216–28. Starke, Peter, Alexandra Kaasch, and Franca Van Hooren. 2014. “Political Parties and Social Policy Responses to Global Economic Crises: Constrained Partisanship in Mature Welfare States.” Journal of Social Policy 43 (2): 225–46. Stephens, John D. 2010. “The Social Rights of Citizenship.” In The Oxford Handbook of the Welfare State, edited by F.G. Castles, S. Leibfried, J. Lewis, H. Obinger and C. Pierson, 511–25. Oxford: Oxford University Press. Takegawa, Shogo. 2009. “International Circumstances as Factors in Building a Welfare State: Welfare Regimes in Europe, Japan and Korea.” International Journal of Japanese Sociology 18 (1): 79–96. Tepe, Markus, and Pieter Vanhuysse. 2009. “Are Aging OECD Welfare States on the Path to Gerontocracy? Evidence from 18 Democracies, 1980–2002.” Journal of Public Policy 29 (1): 1–28. Tromborg, Mathias Wessel. 2014. “Bringing the Median Voter Back in: The Relationship Between Government Debt, Median Voter Preferences, and Welfare State Spending.” Journal of European Social Policy 24 (2): 107–21. Vis, Barbara. 2009. “The Importance of Socio-economic and Political Losses and Gains in Welfare State Reform.” Journal of European Social Policy 19 (5): 395–407. Vrooman, J. Cok. 2012. “Regimes and Cultures of Social Security: Comparing Institutional Models Through Nonlinear PCA.” International Journal of Comparative Sociology 53 (5–6): 444–77. Wenzelburger, Georg, Reimut Zohlnhöfer, and Frieder Wolf. 2013. “Implications of Dataset Choice in Comparative Welfare State Research.” Journal of European Public Policy 20 (9): 1229–50. Wolf, Frieder, Reimut Zohlnhöfer, and Georg Wenzelburger. 2014. “The Politics of Public and Private Pension Generosity in Advanced Democracies.” Social Policy & Administration 48 (1): 86–106. World Bank. 2014. “The State of Social Safety Nets.” Retrieved June 15, 2020 from http://​documents​ .worldbank​.org/​curated/​en/​302571468320707386/​850097536​_201405175114103/​additional/​87984​ 0WP0FINAL0​0Box385208​B00PUBLIC0​.pdf. Xu, Ping. 2017. “Compensation or Retrenchment? The Paradox of Immigration and Public Welfare Spending in the American States.” State Politics & Policy Quarterly 17 (1): 76–104. Yörük, Erdem, Mehmet Fuat Kına, and Ali Bargu. “Political Determinants of Social Assistance: A Critical Systematic Review.” (Manuscript under review).

7. Understanding the ‘welfare state’ in the context of austerity and populism Sonja Blum and Johanna Kuhlmann1

INTRODUCTION While the historical origins of the Western welfare state can be traced back to the late 19th century and industrialisation, its rise is particularly connected to the 20th century, when the provision of welfare became a genuine function of governmental activity in Western countries. Whether in ‘Bismarckian’ or ‘Beveridgean’ stance; whether geographically delineated (and thereby implicitly characterised) as Nordic, Mediterranean, or Antipodean; or whether explicitly characterised (and benchmarked) as social-democratic, conservative-corporatist, or liberal – it is this 20th century and Western type of ‘welfare state’ that definitional struggles and delimitations have been directed at. It is the ‘welfare state’ protecting its clientele from ‘old social risks’, the perils of life – in some places, and then increasingly after the turn to the 21st century, also the ‘new social risks’ of post-industrial labour markets and family structures. Around the turn to the 21st century, significant changes and restructurings of traditional welfare state features were identified. There is a long-standing scholarly debate on a welfare state under ‘permanent austerity’ and retrenchment (see Farnsworth and Irving; Starke, in this volume), but also ‘new’ types of welfare state have been described or propagated, including expansionary movements towards a ‘social investment state’. More recently, it is difficult to name a country where (especially right-wing) populist parties do not influence the welfare agendas, either through direct participation in (coalition) governments or indirectly through affecting other parties’ social policy agendas (Fenger, 2018; Greve, 2019b); wherefore their – still opaque, but potentially fundamental – consequences have gained scholarly attention as well (see Rovira Kaltwasser and Zanotti, in this volume). Moreover, the two phenomena which are at the centre of this Handbook – namely, those of austerity and populism – may touch upon the fundamental pillars of what is conceptualised to be ‘a welfare state’. Against this backdrop, this chapter reflects upon the broader question ‘what is a welfare state’ under the conditions of austerity and populism. Austerity and populism are important factors for explaining social policy reform and, therefore, need to be taken seriously when thinking about the contemporary welfare state, and (dis-)continuities regarding its classical elements. However, the nature of their relationship and combined effect needs to be considered. What are the causal relationships between austerity and populism, and when can they be understood as attacks on the modern welfare state? When does populism become a driver for austerity and when does, in contrast, austerity give rise to specific forms of populism directed at fighting austerity? How can we illuminate this complex austerity–populism nexus in the welfare state of the 21st century, given that not only the role of austerity in contemporary social policymaking, but also the social policy profile of the populist radical right (PRR) are still somewhat unclear and PRR parties have been described to ‘act like chameleons’ (Schumacher and Van Kersbergen, 2016, p. 300) in social policy terms? These are highly 81

82  Handbook on austerity, populism and the welfare state complex questions, and by no means can this chapter provide definitive answers. Rather, based on extant scholarship, it aims to shed light on these questions by highlighting three different relationships between austerity and populism in the welfare state: a welfare-hostile, a welfare-friendly, as well as a welfare-ambiguous face. We start by discussing definitions and common denominators of the ‘traditional’ welfare state of the 20th century, including its different manifestations and typologies. The third section then spotlights the welfare state around the turn of the century, including new policy priorities and the ambiguous role of austerity. In the fourth section, the three different faces of the austerity–populism nexus in the modern welfare state are highlighted and empirically illustrated, before the final section closes with a more general outlook.

WHAT IS THE ‘WELFARE STATE’? Conceptualisations Conceptualisations of the welfare state always reflect the time and context within which they are developed. The difficulties of defining the term ‘welfare state’ have frequently been dealt with in the comparative welfare state literature (for an overview see Kuhlmann, 2019), with Garland even provocatively stating that the whole concept was actually a ‘misnomer’ (Garland, 2016, p. 3). Notwithstanding these difficulties (which the ‘welfare state’ of course shares with many other concepts in the social sciences), definitions of the welfare state abound in the literature. Esping-Andersen (1990) distinguished narrow and broad definitions: while rather narrow definitions focus mostly on benefits in cash and kind, broader ones also consider the complex relationship between welfare benefits and the political economy. A seminal definition that falls into this latter category is given by Briggs who delineates a welfare state as: a state in which organized power is deliberately used (through politics and administration) in an effort to modify the play of market forces in at least three directions – first, by guaranteeing individuals and families a minimum income irrespective of the market value of their work or their property; second, by narrowing the extent of insecurity by enabling individuals and families to meet certain ‘social contingencies’ (for example, sickness, old age and unemployment) which lead otherwise to individual and family crises; and third, by ensuring that all citizens without distinction of status or class are offered the best standards available in relation to a certain agreed range of social services. (Briggs, 1961, p. 228)

Briggs developed this definition against the background of the British discussion on social policy. The third direction that he mentions includes a strong normative assumption that is oriented towards a rather encompassing provision of welfare. Yet Briggs’s definition remains silent about the policy sectors that constitute parts of the described welfare state activities. In fact, what counts as part of the welfare state and what does not is contested and differs depending on different countries’ developments and not least also the researcher’s theoretical perspective (Kuhlmann, 2019, p. 16). A consideration of recent handbooks on the welfare state reveals that the policy sectors of old-age policy, labour market policy, healthcare, long-term care, disability, social assistance, family policy, housing, and (especially in recent years) education can generally be regarded as key areas in most welfare states (Blum et al., 2020; Greve, 2019a).

The ‘welfare state’ in the context of austerity and populism  83 Debates on welfare state conceptualisations are strongly and not only implicitly linked to ‘established’ welfare states, which are mostly European and/or OECD (Organisation for Economic Co-operation and Development) welfare states. However, comparative welfare state research increasingly also focuses on ‘emerging’ welfare states, especially in the Global South. While we would argue that Briggs’s definition can at least partly also be applied to the study of emerging welfare states, it is important to keep in mind that ‘emerging’ welfare states, while in some ways adopting concepts from ‘established’ welfare states (such as e.g. social insurance systems), are also following their own paths when it comes to ‘modify[ing] the play of market forces’ (Briggs, 1961, p. 228). Sequences and Typologies When it comes to analysing welfare state development, different sequences have been distinguished (Obinger and Petersen, 2019, p. 11) – keeping in mind that sequences are not ‘objective’ realities, but the product of interpretation. In a nutshell, the first phase usually comprises the formation of the welfare state from the 1880s (e.g. in the context of industrialisation) to the First World War, while the second phase covers the interwar period. The third phase comprises the post-war welfare state and is often characterised as the ‘golden age’ of the welfare state, and the fourth phase follows the oil crisis in 1973, when the previous expansion of the welfare state was superseded by a restructuring of the welfare state, often including significant retrenchment. In fact, the ‘golden age’ of the welfare state also served as a main reference point for comparative welfare state research at least until the 1990s (Castles et al., 2010, p. 5). The heyday of the (Western European) welfare state saw the maturation of a complex institutional arrangement that was designed in order to ‘protect the income of the male breadwinner’ (Bonoli, 2005, p. 432), which basically meant material compensation in case the male breadwinner could not participate in the labour market because of the ‘old social risks’ of an ‘industrial accident, invalidity, sickness, or unemployment’ (Bonoli, 2005, p. 432), as well as old age. In line with research on the post-war welfare state, prominent analytical distinctions that have been used to analyse these welfare states have included the distinction between an employment-centred social insurance system financed by contributions from both employers and employees with the ultimate aim of status maintenance (the so-called Bismarckian approach that has its roots in the German Empire), and a universal and tax-financed system directed at poverty alleviation (the so-called Beveridgean approach that goes back to the British Beveridge plan from 1942). Besides that, Esping-Andersen’s The Three Worlds of Welfare Capitalism (1990) became a benchmark of comparative welfare state research. In their appreciation of the book, Emmenegger et al. (2015) argue that the Three Worlds stands out because it introduces three types of welfare capitalism – a liberal, a conservative, and a social-democratic model – which can be distinguished with regard to their level of stratification, de-commodification, and public–private welfare mix, thereby reflecting different historical and political developments. Although (or maybe also: because) comparative welfare state research has constantly added further ‘worlds of welfare capitalism’ (such as the Mediterranean model (Ferrera, 1996) or the productivist model (Holliday, 2000)), Esping-Andersen’s work remains an important point of reference.

84  Handbook on austerity, populism and the welfare state The Welfare State around the Turn of the Century: New Policy Priorities and the Ambiguous Role of Austerity The context of ‘permanent austerity’ (Pierson, 2001) has become relevant for scholars of the welfare state at least since the 1990s, and it has received even more attention following the financial and economic crises (Farnsworth and Irving, 2015; Nyby et al., 2017; Taylor-Gooby et al., 2017). Yet, despite austerity’s indisputable presence within both academic debate and public discourse, what austerity actually refers to remains contested, it is a ‘slippery idea’ (see Farnsworth and Irving, in this volume). In his highly cited book Austerity. The History of a Dangerous Idea, Blyth (2013) defines austerity as ‘a form of voluntary deflation in which the economy adjusts through the reduction of wages, prices, and public spending to restore competitiveness, which is (supposedly) best achieved by cutting the state’s budget, debts, and deficits’ (p. 2), or, in a nutshell, ‘the policy of cutting the state’s budget to promote growth’. Although austerity is clearly concerned with economic developments, it is as much an economic as a political project (see Farnsworth and Irving, in this volume): Budget cuts neither fall from the sky nor are they – as we will also show later – the only available option to policymakers in a context of bigger transformations. Rather, policymakers have options – and if they opt for ‘austerity measures’, this can often be better understood in terms of ideological considerations than of apparent ‘necessities’ (see also the fourth section on the welfare state and the austerity–populism nexus). When austerity becomes a core government priority, existing welfare state arrangements can also be at stake. This requires rethinking about conceptual foundations of the welfare state as defined by Briggs (1961), that is, granting a minimum income, narrowing insecurity, and offering a certain standard of living irrespective of status or class, because budget cuts lower the welfare state’s means to achieve these goals, or at least request that existing means of achieving these goals, or even the goals themselves, are questioned. But how does the idea of austerity play out in concrete social policy measures? Has the context of ‘permanent austerity’ paved the way for a straightforward dismantling of the welfare state? From Pierson we know that the story is much more complex. For policymakers, cutting back the welfare state is tricky; in fact, the very logic of welfare state expansion creates powerful welfare state supporters (Pierson, 1994; see also Starke, in this volume) that advocates of austerity have to face. Pierson (1994) distinguishes between programmatic and systemic retrenchment: while programmatic retrenchment implies cutbacks in welfare benefits and services in the short run, systemic retrenchment is ‘directed at undermining the very foundations of the welfare state’ (Jensen et al., 2019, p. 684), for example, by weakening the financial basis of the welfare state, or by weakening strategically important support groups such as trade unions. Moreover, existing actor constellations make a restructuring of the welfare state more likely than welfare state dismantling (Pierson, 2001). Consequently, Pierson has argued for a more nuanced assessment of welfare state restructuring in the context of austerity, focusing on the dimensions of re-commodification, cost containment, and recalibration (Pierson, 2001, pp. 419–27). Cost containment deserves special attention here: Esping-Andersen coined the famous sentence that ‘it is difficult to imagine that anyone struggled for spending per se’ (Esping-Andersen, 1990, p. 21). In contrast, Pierson argues that in the era of austerity, actually fighting the spending per se becomes a key benchmark of policymaking (Pierson, 2001, pp. 423–4), and in fact policy measures aimed at reducing costs have been a crucial driver of welfare state reform in recent decades (see

The ‘welfare state’ in the context of austerity and populism  85 Farnsworth and Irving, in this volume; Blum and Kuhlmann, 2020; Borosch et al., 2016). It is important to note that while austerity and retrenchment often refer to similar developments, the two concepts are distinct. Pierson (1994, 2001) talked about a climate of austerity in which retrenchment is conducted, which, as the distinction between programmatic and systemic retrenchment already reveals, is sometimes used to describe welfare state cutbacks, and sometimes to denote welfare state restructuring more broadly. Austerity clearly depicts a larger economic or ideological programme (see Farnsworth and Irving, in this volume), while at the same time ‘austerity measures’ are clearly concerned about cutbacks. It is beyond the realm of this chapter to fully disentangle similarities and differences between the two concepts. For this chapter, we will focus primarily on ‘austerity measures’ in the stricter sense – that is, in the sense of cutting back benefits and services – to distil the different faces of the austerity– populism nexus in the welfare state. Also in a context of ‘permanent austerity’, budget cuts are by no means ‘the only game left in town’ (Van Kersbergen et al., 2014). From around the turn to the 21st century, comparative welfare state research starts to diagnose something fundamentally new to the welfare state in several respects. ‘New social risks’ have emerged, or rather been reinforced through socio-economic transformations, post-industrial labour markets, and family structures (Bonoli, 2005). Typical examples are work–family reconciliation, single parenthood, having a frail relative, being long-term unemployed or ‘working poor’, as well as insufficient coverage by social security (Bonoli, 2005, pp. 433–5). New social risks have challenged the ‘traditional welfare state’, in which they were not sufficiently covered; particularly the conservative-corporatist welfare state type (Esping-Andersen, 1999). They cut across the post-war welfare state programmes directed at the ‘male breadwinner’ in standard employment; and they deprive them of their necessary counterpart, the ‘female caregiver’ (see e.g. Esping-Andersen, 1999; Lewis, 1992). Along with the transformation of risk structures came also a description of ‘new politics’ of the welfare state, as mentioned above first for the welfare state’s retrenchment, then also for its recalibration and the expansion of new social risk policies. New social risks tend to be concentrated among groups with little mobilising capacity, in particular ‘women, the young and the low skilled’ (Bonoli, 2005, p. 431), but new alliances for new social risk policies were forming in the face of forthcoming labour shortages as, for instance, policies to increase women’s labour market participation aligned interests of such different groups as employers, feminists, and the left (Häusermann, 2006, p. 7). Against the backdrop of risk transformation, the notion of a ‘new welfare state’ (Taylor-Gooby, 2004) emerged, of a ‘social investment welfare state’ in particular (Hemerijck, 2017; Morel et al., 2012), which until today keeps a somewhat unclear status between a more normative political strategy and analytical concept (Hemerijck, 2017). A social investment welfare state is understood as following policies ‘that both invest in human capital development (early childhood education and care, education and lifelong training) and that help to make efficient use of human capital …, while fostering greater social inclusion’ (Morel et al., 2012, p. 2). Determining the role of austerity in the social investment state is challenging. On the one hand, the social investment state is clearly aimed at being economically efficient and promoting growth (especially the former being clearly in line with austerity policies). On the other hand, social investment targeted on human capital development requires considerable amounts of social spending. The simultaneity of austerity policies and social investment policies in European welfare states is in fact remarkable (Hemerijck, 2017, p. 17). Yet, evidence

86  Handbook on austerity, populism and the welfare state from Blum and Kuhlmann (2020) suggests that most countries stick to social investment as a feasible paradigm only as long as they can afford it. What is more, also on a conceptual level, the notion of a ‘new welfare state’ often explicitly contains the idea of reallocating funds, that is, for instance, to cut back cost-intensive pensions in favour of investing in children.

THE WELFARE STATE AND THE AUSTERITY–POPULISM NEXUS Put emphatically, it may be asked whether there is something inherently populist to the welfare state as such. Starting from Bismarck’s social insurance programmes of the late 19th century, which were not least political means to undermine support for socialist movements, social policy remains of outstanding electoral importance until today, and social policies are often reflective of the corresponding political motivations. Yet the interest of this section is not primarily in populism in such a broad or colloquial sense of policies ‘for the people’. Rather, in the face of the ‘stronger support for what are labelled populist parties in many countries’ (Greve, 2019a, p. 1), a narrow definition is followed to think about the austerity–populism nexus. Inclusionary and Exclusionary Populism Cas Mudde (2004, p. 543) has provided us with a narrow definition of populism, of which he conceives as ‘a thin-centred ideology that considers society to be ultimately separated into two homogenous and antagonistic groups, “the pure people” versus “the corrupt elite”, and which argues that politics should be an expression of the volonté générale (general will) of the people’. While this definition implies a certain style of populist policymaking, the ideological contours and their implications for distinct social policy positions take more shape when different types of populism are distinguished. Mudde and Rovira Kaltwasser (2013, p. 158) argue that the ‘issue of inclusion versus exclusion is probably the most important question discussed in the scholarly debate’ around populism, and against this backdrop distinguish an ‘inclusionary populism’ from an ‘exclusionary populism’, that is, paying attention to whether focus is mostly on the inclusion of (parts of) the ingroup, or on the exclusion of outgroups (Mudde and Rovira Kaltwasser, 2013, p. 160). Their distinction is based on three dimensions of inclusion or exclusion (namely: material, symbolic, and political). With regard to social policy – which is essentially about the allocation and design of social rights (see e.g. Blank, 2011) – the material dimension comes to the fore.2 On this material dimension, exclusion and inclusion ‘refer to the distribution of state resources, both monetary and non-monetary, to specific groups in society’ (Mudde and Rovira Kaltwasser, 2013, p. 158; see also Rovira Kaltwasser and Zanotti, in this volume). It is important to note that, in a certain sense, populism always entails both inclusionary and exclusionary features (Mudde and Rovira Kaltwasser, 2013). Looking at the empirical prevalence of the different types of populism, Mudde and Rovira Kaltwasser (2013) have argued that inclusive populism is primarily to be found in Latin America, while European populism is predominantly exclusive. They have explained this observation with the (comparative) generosity of European welfare states, which (compared to Latin America) grant a certain standard of social and political rights also to the weakest groups of their populations. Hence, European populists’ ‘prime focus is on the exclusion of the outgroups rather than on the inclusion of (parts of) the ingroup’ (Mudde and Rovira Kaltwasser,

The ‘welfare state’ in the context of austerity and populism  87 2013, p. 160). Distinctive patterns are arising within Europe, though, with a view to Southern Europe and the recent rise of left-wing populist parties in that area (see below). Manow (2018) takes these patterns (i.e. left-wing populism mainly in Southern European, and right-wing populism mainly in Continental and Northern European countries) as a starting point in his enlightening analysis of variants of populism. With the generous and more universal welfare states in the latter regions, the free movement of people through migration induces political conflict (Manow, 2018, p. 19). Seen more generally, when formerly weaker groups are granted new entitlements (e.g. migrants), this can open the way for other groups’ ‘feeling of being pushed to the back of the queue by these groups’ (Greve, 2019b, p. 155), and blaming them for deterioration in their standard of living. In Central and Eastern European countries, the ‘high costs of economic transition and the low level of political and administrative performance’ (Kriesi and Pappas, 2015, p. 3) have provided fertile ground for anti-elitist sentiments and populist parties. More recently, the welfare–populism nexus has received growing attention (Afonso, 2015; Fenger, 2018; Greve, 2019b; Manow, 2018; Rathgeb, 2020; Röth et al., 2018; Schumacher and Van Kersbergen, 2016), thereby mostly focusing on populism from the right. In the following, we will briefly shed light on three different relationships between austerity and populism that have been theorised: one in which populist parties promote austerity (what we name the welfare-hostile perspective), one in which populist parties oppose austerity (the welfare-friendly perspective), and one in which the relationship between austerity and populism is more complex (the welfare-ambiguous perspective). It is important to note that austerity and populism can both act as either cause or effect within the relationship. Of course, these three relationships are rather stylised, and it is not always easy to draw the line between them. Three Faces of a Complex Relationship The welfare-hostile face The first line of research on the relationship between austerity and populism focuses on PRR parties as advocates of austerity. Kitschelt and McGann (1995) have famously argued that the ‘winning formula’ of radical right parties consists of ‘a resolutely market-liberal stance on economic issues and an authoritarian and particularist stance on political questions of participatory democracy, of individual autonomy of lifestyles and cultural expressions, and of citizenship status’ (see also Betz, 1993; Kitschelt and McGann, 1995, p. 275). This positioning can be read in terms of radical right (PRR) parties’ voter groups, namely the state-critical petite bourgeoisie and blue-collar workers that appeal to right authoritarian positions. In the context of affluent welfare states, it can be argued that the main focus here thus lies on exclusionary populism. A classic example of this welfare-hostile face is the Front National in France in the 1980s and 1990s (Kitschelt and McGann, 1995). Also the Austrian Freedom Party (FPÖ) especially under the first ÖVP-FPÖ government (2000–06) has been connected with a welfare-hostile profile (but see also Rathgeb, 2020, summarised below): As Tálos (2013, p. 341) states, both parties shared the view that a paradigmatic, neo-liberal change was necessary in economic policy and social policy. Their government programme bore a tone of cost containment, deregulation, and risk privatisation (as more programmatic forms of retrenchment), but the attacks on the Austrian social partnership and the trade unions in particular are also witness of systemic retrenchment. Overall, however, populist parties increasingly perceived that showing

88  Handbook on austerity, populism and the welfare state the welfare-hostile face was not electorally successful. Populist parties have therefore been found to subsequently adopt less welfare-hostile positions (Röth et al., 2018). The welfare-friendly face The 1990s and 2000s experienced a strengthening of the climate of permanent austerity (Pierson, 2001), which materialised as welfare cutbacks in countries with governments across the political spectrum – and, notably, also the political left. Especially in the context of the financial crisis from 2008, socio-economic issues and the question how to meet social hardship, for example due to rising unemployment, have become more prominent. Against this backdrop, both radical left and radical right populist parties have been found to oppose austerity measures. First, populist radical left (PRL) parties rose as opponents to the harsh austerity measures implemented to fight the debt crisis, and against the neo-liberal economic system, that is, the free movement of goods and capital (see also Kriesi and Pappas, 2015; Manow, 2018).3 These include the Podemos party in Spain, and the Five Star Movement in Italy, possibly also Greek Syriza. As ‘left-wing’, those parties follow an anti-austerity course, and as ‘populist’ parties, they advocate anti-establishment positions, direct democracy, and Euroscepticism. ‘The people’ in PRL parties’ understanding, who are deserving of support, are typified in ‘the common man’, ‘the poor’, or ‘the 99 per cent’ (Kriesi and Pappas, 2015, p. 5). These figures are indicative of an inclusionary populism, as they focus on mass welfare programmes and empowering disregarded groups (Caiani and Graziano, 2019, p. 1144). Second, not only populist left parties, but also PRR parties have been found to be strong opponents of austerity measures, indicating a shift from the welfare-hostile face adopted earlier. The Front National throughout the 2010s is – again – a case in point: under the leadership of Marine Le Pen (since 2011), the Front National strongly opposed austerity measures and promoted redistributive policies (Ivaldi, 2013). The theoretical argument behind these empirical observations is that, following the austerity reforms also spearheaded by leftist parties, working-class voters form an increasingly crucial part of PRR parties’ electorate. Therefore, these parties give up on their neo-liberal positions and adopt a welfare-friendly position in line with their core voters’ preferences (Afonso and Rennwald, 2018). The welfare-friendly perspective could be understood as portraying PRR parties as rather inclusionary, but only in the sense of safeguarding protection for those who are already labour market insiders; their exclusionary populist stances remain somewhat sidelined in this argumentation. The welfare-ambiguous face Finally, a third strand of literature, which – as the welfare-hostile face – again especially focuses on populism from the right, argues that the relationship between austerity and populism is actually not that straightforward. As in the welfare-friendly variant, the expansionary agendas for the parties’ ingroups are stressed, but so is the – inextricably linked – exclusionary counterpart. Namely, one important manifestation of PRR parties’ position on social policy is welfare chauvinism, depicting a set of ideas which hold that welfare services and expenditure ‘should be restricted to “our own”’ (Andersen and Bjørklund, 1990, p. 212), indicating the perception that immigrants are less entitled to social rights than the native population (Van der Waal et al., 2013). Such welfare-chauvinist strategies have, for instance, been described as particularly successful for the PRR parties in Denmark and the Netherlands (Schumacher and Van Kersbergen, 2016). Rathgeb (2020) has argued that while the Austrian FPÖ supports

The ‘welfare state’ in the context of austerity and populism  89 (social) policies for deserving ‘makers’, that is, employers and employees as well as the traditional especially male workforce, they oppose (social) policies for undeserving ‘takers’, who are particularly immigrants but also the supposedly ‘corrupt elite’. Some work has also found that, given the rather different electoral constituencies, PRR parties will engage in ‘position blurring’ of their social policy programme to attract voters (Rovny, 2013). To give a present-day example, the social policy orientation of Germany’s Alternative für Deutschland (AfD) is still unclear. While the party was founded in 2013 with a strong neo-liberal and anti-Euro profile, welfare-chauvinist positions gained importance alongside its outreach to other voter groups; yet both streams continue to exist within the party. Going beyond the phase of electoral competition, it has also been shown that PRR parties, once in office, are likely to change their stance towards social policymaking (Afonso, 2015). Scholars have also argued that the social policy preferences of PRR parties depend on concrete policies. As Röth et al. (2018) have shown, those parties rather reject market-liberal reforms of general retrenchment and cutting back welfare generosity, but they support to a certain extent reforms of deregulation and privatisation. These positions in a way transverse left–right divides, but can be better grasped by the categories of inclusionary and exclusionary populism, and by whom PRR parties construct as ‘deserving’ or ‘undeserving’ of support (Blum and Kuhlmann, 2019; see also Rathgeb’s 2020 distinction of ‘makers’ and ‘takers’). To conclude, the complex relationship means that PRR parties do not take an overall stance on the issue of austerity, they may in a way support and oppose it. In fact, we can observe growing selectivity and exclusionary tendencies, through which social rights are undermined for groups constructed as ‘undeserving’ (such as migrants), while fairly generous social rights are protected or strengthened for groups constructed as ‘deserving’ – in particular labour market insiders, or the ‘native, obedient, common citizens’ (see Ennser-Jedenastik, 2016; Fenger, 2018, p. 192). Such a development can also be observed in Hungary, with Premier Orbán aiming to ‘create a society based on work instead of the uncompetitive Western type of welfare state’ (Tausz, 2020, p. 257), which also comes with a strong welfare-chauvinist orientation, while at the same time affording generous family benefits, especially to traditional two-parent families with middle and higher incomes (Herke, 2020).

CONCLUSIONS ‘Permanent austerity’ (Pierson, 2001) has been around for 30 years now, but clearly it does not determine the scene: social policymaking continues to take different (sometimes also contradictory) paths, that include various forms of recalibration, and also significant expansion. Yet it is safe to say that austerity is shaping ‘the welfare state’ of today, and that various crises – some national, some cross-national, some even global – have contributed to this. The notion of ‘crisis’ is also essential to both PRL and PRR parties in the welfare state context: both PRL and PRR parties have been reinforced through crisis, but especially for PRR parties it has been described how they also actively and strategically ‘perform and perpetuate a sense of crisis’ in their engagement with social policy (Ketola and Nordensvard, 2018; Moffitt, 2015, p. 195). Although it has been researched for the past 30 years now, how the answer to the question ‘what is a welfare state’ changes in a context of austerity is far from clear – with directions including cutbacks, privatisation of risks provision (e.g. in pensions), and targeting of benefits and services, for example, through stronger means-testing, while at the same time also leaving

90  Handbook on austerity, populism and the welfare state room for manoeuvre for policies directed at ‘new social risks’. How populism affects our understanding of ‘the welfare state’ is even more opaque. This also has to do with the definition of populism as such. We may conceive of populism as a strategy, and only ‘thin-centred ideology’, as Mudde (2004) put it. Only if specific types of populism are distinguished and their relationship to austerity is clarified can conclusions on social policy agendas be drawn. This chapter has discussed three current faces of the austerity–populism nexus: the welfare-hostile face, the welfare-friendly face, and the welfare-ambiguous face. Reviewing extant scholarship, we can conclude that while populist parties today have largely disbanded a welfare-hostile austerity face, neither are they welfare-friendly per se: their dominating welfare-ambiguous face – including welfare chauvinism and selective expansion – touches upon fundamental pillars of the welfare states’ architecture. So indeed, populism puts the classical ‘welfare state’ under pressure, and can change in which ways (and for whom!) social contingencies are offset and social services provided (Briggs, 1961), which risks are covered, and which degree of inequality is accepted. In a welfare state in the context of populism, the antagonism of ‘the pure people’ versus ‘the elite’, and social policies become an important reference point for discussing, negotiating, and reforming the foundations of the welfare state, including an adequate minimum income level, level of insecurity, and standard of living irrespective of class (Briggs, 1961). Yet these distinctions are not straightforward and can change, as the example of PRR parties shows: while many used to follow a neo-liberal agenda, promoting welfare cutbacks overall, welfare chauvinism has now become dominant, where welfare benefits are generally welcome, but should be restricted to ‘our own’. ‘What is a welfare state’ under such conditions of welfare chauvinism involves changes towards stronger selectivity and exclusionary tendencies. Crucial for ‘the welfare state’ under conditions of ‘permanent austerity’ are the ‘new politics’ (Pierson, 2001); and later new alliances have also been identified for the extension of new social risk policies (Bonoli, 2005; Häusermann, 2006). A crucial question for whether ‘austerity’ and ‘populism’ will attack the welfare state will be the extent and ways in which new alliances between respective proponents will form, and shape ‘what is a welfare state’ in the future.

NOTES 1.

For valuable comments on previous versions of this chapter (and particularly helping us to disentangle the complex relationship between austerity, populism, and the welfare state), our thanks go to Bent Greve, Kevin Farnsworth and Zoë Irving, Mikael Nygård and Mikko Kuisma, Friederike Römer, and Peter Starke. Johanna Kuhlmann acknowledges funding by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – Projektnummer 374666841 – SFB 1342. 2. Of course, also the political dimension (i.e. political participation and representation) and especially the symbolic dimension become specifically relevant for social policy. The symbolic dimension denotes how the boundaries of ‘the people’ and ‘the elite’ are set, and how ‘(characteristics and values of) certain groups’ are included or excluded in ‘rhetoric and symbols’ (Mudde and Rovira Kaltwasser, 2013, p. 164). 3. As Manow (2018) further elaborates, the Southern European welfare states also offer more or less generous benefits, but benefit access in those systems is typically clientelist and focused on labour market insiders. As migrants therefore do not have access to the main welfare state schemes, the free movement of people does not implicate the same social battles as in Northern European countries. And further, the welfare states and labour markets also do not attract the same migration as in Northern Europe.

The ‘welfare state’ in the context of austerity and populism  91

REFERENCES Afonso, A. (2015). Choosing Whom to Betray: Populist Right-wing Parties, Welfare State Reforms and the Trade-off between Office and Votes. European Political Science Review, 7(2), 271–92. Afonso, A., and Rennwald, L. (2018). Social Class and the Changing Welfare State Agenda of Radical Right Parties in Europe. In P. Manow, B. Palier, and H. Schwander (eds), Welfare Democracies and Party Politics. Explaining Electoral Dynamics in Times of Changing Welfare Capitalism (pp. 171–94). Oxford: Oxford University Press. Andersen, J.G., and Bjørklund, T. (1990). Structural Changes and New Cleavages: The Progress Parties in Denmark and Norway. Acta Sociologica, 33(1), 195–217. Betz, H.-G. (1993). Radical Right-wing Populism in Western Europe. London: Macmillan. Blank, F. (2011). Soziale Rechte 1998–2005. Die Wohlfahrtsstaatsreformen der rot-grünen Bundesregierung. Wiesbaden: Springer VS. Blum, S., and Kuhlmann, J. (2019). Stories of How to Give or Take – Towards a Typology of Social Policy Reform Narratives. Policy and Society, 38(3), 339–55. Blum, S., and Kuhlmann, J. (2020). Landscapes in Motion: Welfare System Reform in 28 European Countries, 1998–2018. In S. Blum, J. Kuhlmann, and K. Schubert (eds), Routledge Handbook of European Welfare States (2nd edn, pp. 577–91). London: Routledge. Blum, S., Kuhlmann, J., and Schubert, K. (eds) (2020). Routledge Handbook of European Welfare States (2nd edn). London: Routledge. Blyth, M. (2013). Austerity. The History of a Dangerous Idea. Oxford: Oxford University Press. Bonoli, G. (2005). The Politics of the New Social Policies: Providing Coverage against New Social Risks in Mature Welfare States. Policy & Politics, 33(3), 431–49. Borosch, N., Kuhlmann, J., and Blum, S. (2016). Opening up Opportunities and Risks? Retrenchment, Activation and Targeting as Main Trends of Recent Welfare State Reforms across Europe. In K. Schubert, P. de Villota, and J. Kuhlmann (eds), Challenges to European Welfare Systems (pp. 769–91). Cham: Springer International Publishing. Briggs, A. (1961). The Welfare State in Historical Perspective. European Journal of Sociology, 2(1), 221–58. Caiani, M., and Graziano, P. (2019). Understanding Varieties of Populism in Times of Crises. West European Politics, 42(6), 1141–58. Castles, F.G., Leibfried, S., Lewis, J., Obinger, H., and Pierson, C. (2010). Introduction. In F.G. Castles, S. Leibfried, J. Lewis, H. Obinger, and C. Pierson (eds), The Oxford Handbook of the Welfare State (pp. 1–15). Oxford: Oxford University Press. Emmenegger, P., Kvist, J., Marx, P., and Petersen, K. (2015). Three Worlds of Welfare Capitalism: The Making of a Classic. Journal of European Social Policy, 25(1), 3–13. Ennser-Jedenastik, L. (2016). A Welfare State for Whom? A Group-based Account of the Austrian Freedom Party’s Social Policy Profile. Swiss Political Science Review, 22(2), 409–27. Esping-Andersen, G. (1990). The Three Worlds of Welfare Capitalism. Cambridge: Polity Press. Esping-Andersen, G. (ed.) (1999). Social Foundations of Postindustrial Economies. Oxford: Oxford University Press. Farnsworth, K., and Irving, Z. (eds) (2015). Social Policy in Times of Austerity. Global Economic Crisis and the New Politics of Welfare. Bristol: Policy Press. Fenger, M. (2018). The Social Policy Agendas of Populist Radical Right Parties in Comparative Perspective. Journal of International and Comparative Social Policy, 34(3), 188–209. Ferrera, M. (1996). The ‘Southern Model’ of Welfare in Social Europe. Journal of European Social Policy, 6(1), 17–37. Garland, D. (2016). The Welfare State: A Very Short Introduction. Oxford: Oxford University Press. Greve, B. (ed.) (2019a). The Routledge Handbook of the Welfare State (2nd edn). London, New York: Routledge. Greve, B. (2019b). Welfare, Populism and Welfare Chauvinism. Bristol: Policy Press. Häusermann, S. (2006). Changing Coalitions in Social Policy Reforms: The Politics of New Social Needs and Demands. Journal of European Social Policy, 16(5), 5–21. Hemerijck, A. (2017). Social Investment and Its Critics. In A. Hemerijck (ed.), The Uses of Social Investment (pp. 3–39). Oxford: Oxford University Press.

92  Handbook on austerity, populism and the welfare state Herke, B. (2020). Investigating the Welfare Deservingness of Single Mothers: Public Image and Deservingness Perceptions in Hungary. East European Politics and Societies and Cultures. https://​doi​ .org/​10​.1177/​0888325420937773. Holliday, I. (2000). Productivist Welfare Capitalism: Social Policy in East Asia. Political Studies, 48(4), 706–23. Ivaldi, G. (2013). A New Radical Right Economic Agenda?: The Transformation of the Front National in France. Workshop on ‘Radical Right-wing Populists and the Economy’. Retrieved from https://​halshs​ .archives​-ouvertes​.fr/​halshs​-00879951 (accessed 19 January 2021). Jensen, C., Wenzelburger, G., and Zohlnhöfer, R. (2019). Dismantling the Welfare State? After Twenty-five Years: What Have We Learned and What Should We Learn? Journal of European Social Policy, 29(5), 681–91. Ketola, M., and Nordensvard, J. (2018). Reviewing the Relationship between Social Policy and the Contemporary Populist Radical Right: Welfare Chauvinism, Welfare Nation State and Social Citizenship. Journal of International and Comparative Social Policy, 34(3), 172–87. Kitschelt, H., and McGann, A. (1995). The Radical Right in Western Europe. A Comparative Analysis. Ann Arbor, MI: University of Michigan Press. Kriesi, H., and Pappas, T.S. (2015). Populism in Europe during Crisis: An Introduction. In H. Kriesi and T.S. Pappas (eds), European Populism in the Shadow of the Great Recession (pp. 1–19). Colchester: ECPR Press. Kuhlmann, J. (2019). What is a Welfare State? In B. Greve (ed.), The Routledge Handbook of the Welfare State (2nd edn, pp. 13–22). London, New York: Routledge. Lewis, J. (1992). Gender and the Development of Welfare Regimes. Journal of European Social Policy, 2(3), 159–73. Manow, P. (2018). Die politische Ökonomie des Populismus. Frankfurt a.M.: Suhrkamp. Moffitt, B. (2015). How to Perform Crisis: A Model for Understanding the Key Role of Crisis in Contemporary Populism. Government and Opposition, 50(2), 189–217. Morel, N., Palier, B., and Palme, J. (2012). Beyond the Welfare State as We Knew it? In N. Morel, B. Palier, and J. Palme (eds), Towards a Social Investment Welfare State? Ideas, Policies and Challenges (pp. 1–30). Bristol: Policy Press. Mudde, C. (2004). The Populist Zeitgeist. Government and Opposition, 39(4), 542–63. Mudde, C., and Rovira Kaltwasser, C. (2013). Exclusionary vs. Inclusionary Populism: Comparing Contemporary Europe and Latin America. Government and Opposition, 48(2), 147–74. Nyby, J., Nygård, M., Autto, J., Kuisma, M., and Blum, S. (2017). The Role of Discourse in Family Policy Reform: The Case of Finland. Critical Social Policy, 38(3), 567–88. Obinger, H., and Petersen, K. (2019). Die historische Entwicklung des Wohlfahrtsstaates: Von den Anfängen bis zum Ende des Goldenen Zeitalters. In H. Obinger and M.G. Schmidt (eds), Handbuch Sozialpolitik (pp. 9–31). Springer VS: Wiesbaden. Pierson, P. (1994). Dismantling the Welfare State? Reagan, Thatcher, and the Politics of Retrenchment. Cambridge: Cambridge University Press. Pierson, P. (2001). Coping with Permanent Austerity: Welfare State Restructuring in Affluent Democracies. In P. Pierson (ed.), The New Politics of the Welfare State (pp. 410–56). Oxford: Oxford University Press. Rathgeb, P. (2020). Makers against Takers: The Socio-economic Ideology and Policy of the Austrian Freedom Party. West European Politics. https://​doi​.org/​10​.1080/​01402382​.2020​.1720400. Röth, L., Afonso, A., and Spies, D.C. (2018). The Impact of Populist Radical Right Parties on Socio-economic Policies. European Political Science Review, 10(3), 325–50. Rovny, J. (2013). Where Do Radical Right Parties Stand? Position Blurring in Multidimensional Competition. European Political Science Review, 5(1), 1–26. Schumacher, G., and Van Kersbergen, K. (2016). Do Mainstream Parties Adapt to the Welfare Chauvinism of Populist Parties? Party Politics, 22(3), 300–12. Tálos, E. (2013). Bestimmungsfaktoren der Sozialpolitik. Am Beispiel des österreichischen Sozialstaates. In K. Armingeon (ed.), Staatstätigkeiten, Parteien und Demokratie (pp. 331–52). Wiesbaden: Springer.

The ‘welfare state’ in the context of austerity and populism  93 Tausz, K. (2020). Pathway to a Punitive Workfare System: Hungary. In S. Blum, J. Kuhlmann, and K. Schubert (eds), Routledge Handbook of European Welfare States (2nd edn, pp. 256–74). London: Routledge. Taylor-Gooby, P. (2004). New Risks, New Welfare: The Transformation of the European Welfare State. Oxford: Oxford University Press. Taylor-Gooby, P., Leruth, B., and Chung, H. (eds) (2017). After Austerity: Welfare State Transformations in Europe after the Great Recession. Oxford: Oxford University Press. Van der Waal, J., De Koster, W., and Van Oorschot, W. (2013). Three Worlds of Welfare Chauvinism? How Welfare Regimes Affect Support for Distributing Welfare to Immigrants in Europe. Journal of Comparative Policy Analysis: Research and Practice, 15(2), 164–81. Van Kersbergen, K., Vis, B., and Hemerijck, A. (2014). The Great Recession and Welfare State Reform: Is Retrenchment Really the Only Game Left in Town? Social Policy & Administration, 48(7), 883–904.

8. Austerity, populism, and the politics of blame: an ideational perspective Daniel Béland and Alex Waddan

INTRODUCTION In rich democracies, especially in Europe, austerity and populism are two major political phenomena that have drawn much scholarly attention since the 2008 financial crisis. First, in European countries as different as Greece and the UK, during the last decade, austerity became a central and contested policy orientation. Second, populism, especially on the right of the political spectrum, has become increasingly influential in European countries such as Hungary, Italy, Poland, and the UK, but also in the US, in the aftermath of the 2016 election of Donald Trump to the presidency. Yet, in rich democracies, both austerity and populism predate the 2008 financial crisis (Mudde and Rovira Kaltwasser, 2017). Because of the contemporary prominence of these two concepts and their importance for ongoing social policy debates, working on the relationship between austerity and populism is a useful endeavor that could improve our understanding of these debates. To shed light on these concepts, it is necessary to define them before exploring their potential relationship. In this chapter, we do this using an ideational perspective that stresses the role of framing processes in the politics of social policy, with a particular focus on the politics of blame avoidance and blame generation (Weaver, 2018). Then, we use the example of the UK before and during the 2016 Brexit referendum to illustrate our claims.

IDEAS Both austerity and populism explicitly relate to the role of ideas in politics (Blyth, 2013; Mudde and Rovira Kaltwasser, 2017). Before we turn to these two phenomena and stress their ideational components, a few general remarks about the ideas literature in political science and social policy research are necessary. First, ideas refer to the changing “causal beliefs” of individuals and collective actors (Goldstein and Keohane, 1993), which include their values and perceptions (Béland and Cox, 2011). This actor-centric understanding of ideas (Genieys and Smyrl, 2008) points to how assumptions about the social and political environment evolve over time, and how these ideational constructions affect both policy decisions and political struggles over them (Béland, 2016). Second, ideas are not simple “epiphenomena” but a key aspect of social and political life that can shape policy decisions on its own (Campbell, 2004). Third, different types of ideas are involved in the policy process and breaking down these ideas is essential to explore their possible policy and political impact (Béland and Waddan, 2015). Types of ideas especially central to the policy process include frames, programs, policy paradigms, and public sentiments (Campbell, 2004). Although it is clear that “ideas matter” in social policy, understanding how they matter requires rigorous conceptual and methodological 94

Austerity, populism, and the politics of blame  95 analysis (Jacobs, 2009; Mehta, 2011). In this chapter, we do not seek to demonstrate the causal impact of ideas on the politics of austerity and populism. Instead, we draw on a broad ideational perspective centered on the analysis of framing processes to map these two phenomena and explore the ways in which they might interact.

AUSTERITY In social policy research, Paul Pierson (1994, 1998) helped popularize the concept of austerity in social policy research. For Pierson (1994: 164), the Thatcher and Reagan years marked the advent of an “era of austerity” in which “the politics of retrenchment” replaces the long push for welfare state expansion that took place during the three decades that followed the Second World War. This led him to theorize what he called “permanent austerity,” which he described as a new state of affairs in social policy stemming from endogenous socio-economic changes such as population aging and post-industrial labor markets increasingly centered on the service sector (Pierson, 1998). Facing new economic and fiscal pressures, governments used various strategies to constrain costs and restructure the welfare state. In this context, for Pierson (1998), austerity is something that takes place over long periods while paving the way to retrenchment and restructuring, which both seek to control social spending. In contrast to Pierson (1994, 1998), Mark Blyth (2013) shows that austerity is not only a fiscal and institutional context but a powerful economic idea that has a long intellectual history that can be traced back to the work of classical liberal authors such as John Locke, David Hume, and Adam Smith. The central assumption of the modern idea of austerity is that public debt is a scourge that creates unsustainable fiscal burdens over time, which is why a war on debt is an absolute necessity (Blyth, 2013: 108). As Blyth (2013: 108) suggests, the criticisms of Italian and Greek fiscal policy and debt level in the aftermath of the 2008 financial crisis are strikingly similar to the anti-debt ideology Hume sketched more than 300 years earlier. Simultaneously, while the US rejected austerity under the leadership of left-of-center, Democratic President Barack Obama, it became dominant in the UK under the right-of-center, Conservative–Liberal Democrat coalition government of David Cameron, which took power in 2010 (Blyth, 2013: 59). Because it spreads the belief that public debt is the enemy, this idea of austerity can legitimize welfare state retrenchment. Yet, political actors pursuing retrenchment strategies in the name of an anti-debt ideology do not always accept the pejorative label of “austerity” to define what they are doing. Frequently, the claim is that the cutbacks they enact are purely pragmatic in nature as they are the product of an economic and fiscal “necessity,” a claim that mirrors Margaret Thatcher’s “there is no alternative” (TINA) motto, which has become prevalent in many other jurisdictions than the UK since the 1980s (Blyth, 2013: 171). Consequently, political actors seeking retrenchment and restructuring in the name of cost control may reject “austerity” as an ideological construct, using their insistence that “there is not alternative to cuts” (Blyth, 2013: 171) as a way to blame outside force for their decisions, which they frame as inevitable. Paradoxically, this type of technocratic yet neoliberal discourse becomes a key aspect of what Robert Henry Cox (2001) calls “the social construction of the need to reform.” In other words, austerity is a powerful economic idea that can influence decision-makers to pursue retrenchment strategies, even if they deny such strategies, which they depict as inevitable (TINA), and have no ideological underpinnings.

96  Handbook on austerity, populism and the welfare state

POPULISM Populism is a concept even slipperier and more contested than austerity. According to Jan-Werner Müller (2016: 2), when we talk about populism, “it is far from obvious that we know what we are talking about. We simply do not have something like a theory of populism, and we seem to lack coherent criteria for deciding when political actors turn populist in some meaningful sense.” Importantly, criticizing the elites is “a necessary but not a sufficient condition … to qualify as a populist” (20). This is why Müller’s definition of populism has not one but two main components: the critique of the elites and, additionally, the claim to speak on behalf of a unified, coherent people: “In addition to being antielitist, populists are always antipluralist: populists claim that they, and only they, represent the people” (20). By framing any opposition to it as an enemy of the people understood as a monolithic entity, populism features an anti-pluralist and moralistic frame rooted in the claim that only “a part of the people is the people – and that only the populist authentically identities and represents this real or true people” (22–3). Finally, Müller (2016) and other well-known specialists of populism such as Cas Mudde and Cristóbal Rovira Kaltwasser (2017) remind us that populism is not always located on the right of the political spectrum. In fact, although left-wing populism has proved more popular in Latin America, it does exist in Europe as well, for instance Greece’s Syriza and Spain’s Podemos (Mudde and Rovira Kaltwasser, 2017: 37). Moreover, political theorists such as Ernesto Laclau (2005) have long promoted left-wing populism as a positive force capable of bringing positive social and political change. Like the one of austerity, the concept of populism lends itself to ideational analysis, something that Mudde and Rovira Kaltwasser (2017: 6) make clear, arguing that populism is a “thin-centred ideology” that opposes the “pure people” and the “corrupt elite.” Yet, as suggested above, because populism exists on both the left and the right of the political spectrum, we prefer to understand populism as a type of discourse (i.e. a way to frame the issues) that proponents from different ideological traditions can draw upon. This is what Mudde and Rovira Kaltwasser (2017: 6) suggest when they claim that “populism almost always appears attached to other ideological elements, which are crucial for the promotion of political projects that are appealing to a broader public.” This ideological malleability of populism also helps us understand why this type of discourse can be located on either the right or the left of the political spectrum. Finally, like austerity, populism can take a pejorative meaning, which is why political actors who embrace a populist discourse may reject the label altogether. In the case of both austerity and populism, scholars should scrutinize existing labels and use rigorous definitions to characterize the actors and discourses of the political actors they study.

THE POLITICS OF BLAME: HOW AUSTERITY AND POPULISM CAN INTERACT One way to grasp the potential relationship between austerity and populism is to turn to the work of Kent Weaver (1986, 2018) on blame avoidance and blame generation. On the one hand, blame avoidance is about minimizing their responsibility for unpopular actions taken and denying or obscuring their positional distance from any group or from the electorate as a whole. They may use a variety of strategies to

Austerity, populism, and the politics of blame  97 do this, such as obscuring their own positions, passing the buck for decision-making to other policymakers …, or “circling the wagons” to reach a consensus among a supermajority of politicians on a position so that blame is broadly diffused. (Weaver, 2018: 260)

On the other hand, politicians “engage in blame generating or negative messaging about the past, present, or projected future actions, as well as the positions, character, and other attributes, of political opponents” (Weaver, 2018: 260). Following Weaver (2018), we can define blame avoidance as a set of strategies through which politicians attempt to shield themselves against criticisms for unpopular measures, and blame generation as strategies through which politicians attempt to put a negative spin on the actions of their opponents to undermine their political support. At the most general level, it is possible to argue that blame avoidance is prevalent in the context of the politics of welfare state retrenchment stemming from austerity (Pierson, 1994). Conversely, while populists can use blame avoidance strategies to shield themselves against criticism for unpopular decisions, populism is also likely to feature blame generating strategies according to which opponents are depicted as members of a corrupt elite acting against the “people” they claim to defend. In the context of the politics of retrenchment, it is possible for populists on both the left and the right of the political spectrum to use blame generating strategies to embarrass the ruling “elite” through negative framing of austerity policies. Beyond austerity, right-wing populists are likely to draw on blame generating strategies to attack the “elite” regarding immigration policies and the social and economic problems they allegedly foster. Sometimes, immigrants can also become culprits for austerity, when they are depicted as a fiscal burden and a source of welfare dependency the ruling elite supports instead of siding with the people. This trend leads to welfare chauvinism, a form of anti-immigrant sentiment depicting migrants as a menace to existing social programs (on the relationship between populism and welfare chauvinism see Greve, 2019). In the following sections, we illustrate some of the key issues raised above through a discussion of the politics of austerity and populism in the UK over the last decade. This case is particularly interesting because, during this relatively short period, the UK witnessed both austerity policies and the rise of populist politics associated with Brexit. Our analysis starts with the onset of the financial crisis and how the expanded government deficits and overall debt that resulted from that crisis shifted both the political and policy landscape. We then focus on the policies pursued by the Conservative-led Coalition government from 2010 to 2015 and the subsequent Conservative governments. Importantly these governments focused on austerity rather than revenue raising as a means of reducing debt. As we explain, the retrenchment policies required to implement austerity were justified in related but distinct ways. First, the overall strategy was presented as an unfortunate necessity. There was no alternative to cutting spending and so the incumbent government should not be blamed for the pain incurred. Second, some of the specific cuts to welfare programs were pushed in ideological terms, bringing into play populist style themes that distinguished between the “deserving” and “underserving.” Here the aim was not simply blame avoidance but blame generation.

98  Handbook on austerity, populism and the welfare state

AUSTERITY IN THE UK PRIOR TO THE 2008 FINANCIAL CRISIS In the fall of 2007, the Conservative Party Shadow Chancellor, George Osborne, sought to alleviate fears that a future Conservative government would cut government spending. With stories circulating that the newly installed Labour Prime Minister, Gordon Brown, was contemplating calling an early general election, Osborne committed the Conservatives to maintaining the Labour government’s spending proposals at least in the short term. He claimed: “The result of adopting these spending totals is that under a Conservative government there will be real increases in spending on public services, year after year. The charge from our opponents that we will cut services becomes transparently false” (Barker, 2007). The Conservative home website (2007) explicitly laid out the political calculations behind this statement, “Mr Osborne says that the Tories will deliver year-on-year real increases in public spending and will match Labour’s spending plans for the next two years. The announcements are an attempt to pre-empt any Labour charge that the Tories will cut public services.” Hence, at that point in the political and economic cycle, the Conservative Party leadership clearly felt that it would be an electoral liability to be perceived as being wedded to spending reduction plans. In this context it is important to reflect how, between 1997 and 2010, somewhat belying the narrative of permanent austerity, public spending in the UK had increased “by an average of 4.4% a year in real terms under Labour, significantly faster than the 0.7% a year average seen under the Conservatives from 1979 to 1997” (Chote et al., 2010a: 1). While spending on public services is not the same as spending on social policy, the increases included an average annual growth rate of 5.7 percent on the National Health Service (NHS), 3.9 percent on education, and 2.2 percent on Social Security (Chote et al., 2010a: 9).1 Prime Minister Brown did not call an early election in fact and the Parliament lasted its full term. Hence, the actual election did not take place until spring 2010, by which time the political and economic environment had dramatically shifted from where it had been three years earlier. After over five years of constant economic growth as measured by the quarterly gross domestic product (GDP) figures, the UK saw five consecutive quarters of decline. Overall, the country’s economy contracted “by more than 6% between the first quarter of 2008 and the second quarter of 2009” (Office for National Statistics, 2018). As spelled out by the Office for Budget Responsibility, the impact of the worst recession since the end of the Second World War and the emergency measures taken to support the financial sector were to play havoc with the government’s finances: The budget deficit quadrupled between 2007–08 and 2009–10, reaching £157 billion or 11 per cent of GDP – and it was still £115 billion or 7 per cent of GDP three years later. Public sector net debt increased by £647 billion, doubling to 74 per cent of GDP, over the same five years. (Riley and Chote, 2014: 1)

As political actors scrambled to react to the evidence that the financial crisis was causing major disruption to economic and fiscal forecasts, the Conservatives quickly ended discussion about maintaining spending increases and the discourse switched to the necessity of austerity. This turn was required, it was alleged, because of the profligacy of the Labour government. In a speech in spring 2009, David Cameron emphasized the need for a “responsible politics.” He announced, the “money has run out” and that hence the “age of irresponsibility is giving way to the age of austerity” (Cameron, 2009). At the same conference, Osborne’s promise to match Labour’s spending commitments was washed from the record: “Gordon Brown’s only

Austerity, populism, and the politics of blame  99 argument was Labour spending versus Tory cuts. His only policy was to throw money at every problem. Now that there is no money left, he has no policy left. And no argument either. No policy for how to govern in an age of austerity” (Osborne, 2009). When, a year later, Cameron and Osborne took occupancy of 10 and 11 Downing Street their message about the need for austerity was given extra credibility by the outgoing Chief Secretary to the Treasury, Liam Byrne. In a note for his successor, which Byrne later described as a misplaced joke, he wrote, “Dear chief secretary, I’m afraid to tell you there’s no money left” (Owen, 2010). The reporting of that episode gave at least short-term apparent legitimacy to the effort at pre-emptive blame avoidance for the oncoming austerity measures.

FRAMING AUSTERITY TO JUSTIFY IT Prior to the election there was consensus across the mainstream political parties that fortifying the government’s finances would be a priority for the next administration, but there was no uniformity about how much austerity this need entailed. For example, the party manifestos contained important differences in the balance of spending cuts to tax rises in the effort to control the deficit. All did advance plans that emphasized cuts over tax hikes, but Labour suggested a ratio of 2:1 and the Liberal Democrats went for 2.5:1. For the Conservatives, the appropriate mix was 4:1 (Chote et al., 2010b: 2). Hence, the incoming Conservative-led Coalition government entered office with “austerity” as its watchword, even if Conservative plans were somewhat tempered by the need to accommodate some Liberal-Democrat wishes. Within weeks of taking over as Prime Minister, Cameron pronounced that the government’s overall debt and the projected deficits were “even worse than we thought.” While acknowledging that “How we deal with these things will affect our economy, our society – indeed our whole way of life,” he also claimed that “I want this Government to carry out Britain’s unavoidable deficit reduction plan in a way that strengthens and unites the country.” Hence, he insisted, “we are all in this together” (Tapsfield and Branagh, 2010). This was in line with the pre-election message that the Conservative Party’s realism about the inevitability of painful spending cuts was compatible with the notion that the Tories were nevertheless a “party of the NHS … and of social justice too” (Cameron, 2009). This framing, almost attempting to draw on the “spirit of the Blitz” and the enforced austerity of the Second World War and the immediate aftermath, was clearly an attempt to frame government policy in terms of there being “no alternative” and as a device to “gain political support” for the forthcoming spending cuts (MacLeavy, 2011: 357). Yet, in contrast to the spirit of the wartime Beveridge Report and the emphasis on the need for greater government intervention to combat the structural causes of the “five giants” (want, ignorance, disease, squalor, and idleness) that afflicted so many (Timmins, 1995), the Cameron government was more inclined to stress the role of individual agency when it came to the particular giants of “Want” and “Idleness.” Thus, the insistence that the population was “all in it together” was supplemented by a quite distinct frame of “scroungers versus strivers.” This was exemplified in the speech given by Chancellor George Osborne (2012) to the Conservative Party conference in October 2012, in which he outlined his plans for an extra £10 billion in spending cuts. At the start of his speech, he declared the continuing importance of a shared burden: “We’re not going to get through this as a country if we set one group against another, if we divide, denounce and demonise …

100  Handbook on austerity, populism and the welfare state We are still all in this together.” But that notion of “together” was qualified by the requirement that all individuals abide by what the “British people” understood “by fair.” In this context, Osborne enquired of the delegates: “Where is the fairness, we ask, for the shift-worker, leaving home in the dark hours of the early morning, who looks up at the closed blinds of their next-door neighbour sleeping off a life on benefits?” This stigmatization of some welfare beneficiaries can be seen as an attempt to play into traditional conservative tropes distinguishing between the “deserving” and “undeserving” recipients of state assistance.

AUSTERITY IN ACTION As suggested by these mixed messages about how the burden of austerity was an unfortunate and unforeseen one, to be shared by the wider public, yet that there were also particular groups that deserved the coming disruption to their lifestyles, the austerity axe did not swing down evenly across all areas of social policy. Some big ticket items were protected or at least spared the sharpest reductions, while elsewhere the cuts were deep. One area where the Cameron government diverged from the Thatcherite model was in its attitude towards the state pension. In 1980, the Thatcher government changed the indexing of the basic state pension so that it rose only in line with inflation, rather than with the higher of inflation or wage growth. This had the result of diminishing the relative value of the basic pension to 14 percent of mean male earnings in 1997, compared to 20 percent in 1981 (Evandrou and Falkingham, 2009: 162). The Blair governments had reversed this trajectory and through the 2010 campaign the Tories promised to maintain the framework put in place by Labour. The Coalition government then went even further by saying that the state pension would increase by the higher of inflation or wage growth annually, or by a minimum of 2.5 percent should both those other metrics fall short of that figure. Hence, existing pensions were protected from the austerity bite by the so-called “triple lock.” In 2015, the Office of Budget Responsibility (2015: 45) forecast that state spending on pensions would in fact decline as a percentage of GDP over the next five years as a consequence of policy retrenchment that took the form of the planned rise in the pensionable age. That change did not go unnoticed but it did not have an impact on current seniors already receiving their benefits, which diminished its immediate political impact (Barr and Diamond, 2010: 140–1). Prime Minister Cameron was also much more forceful in his defense of the NHS than Thatcher had been. The NHS, which is often referred to as a “national treasure” or the “crown jewel” of the nation’s welfare state, had prospered during Tony Blair and Gordon Brown’s time in Downing Street as spending doubled in real terms (Vizard et al., 2016: 155). Despite these numbers and the fact that the NHS was an area where Labour would traditionally expect to have “issue ownership,” during the 2010 campaign David Cameron promised that the Tories were now the “party of the NHS” maintaining, “We are the only party committed to protecting NHS spending. I’ll cut the deficit, not the NHS” (Winnett, 2010). As things turned out, the Coalition government kept its promise in terms of sparing the NHS from cuts to its current spending levels, but there were effective reductions to the service’s capacity over time as annual expenditure growth may have been “positive,” but it was in fact “exceptionally low,” falling well below historical trends (Vizard et al., 2016: 157). Therefore, it was simultaneously true that in the years from 2010 to 2018, “Relative to other areas of public spending, health spending has actually been more favoured … than it was in the previous decade”

Austerity, populism, and the politics of blame  101 as it accounted for an increasing share of public service spending, yet with annual average growth rates of 1.1 percent during the Coalition years and then 2.3 percent though the next Conservative government public spending on health was considerably short of the average annual growth rate of 3.7 percent since the foundation of the NHS (Institute for Fiscal Studies and the Health Foundation, 2018: ii). Under such financial pressure it was not surprising that public discourse focused on the strains the NHS was facing, especially as the service regularly missed its targets for treating patients in a timely fashion. By 2018, the health-focused think tank, the King’s Fund, reported: The NHS has not met the standard that 95 per cent of patients should be seen within four hours in A&E [Accident and Emergency departments] since July 2015. The number of people on waiting lists for elective treatment is now more than 4 million, and the standard that 92 per cent of patients should start treatment within 18 weeks of referral has not been achieved since February 2016. The standard that 85 per cent of patients should begin definitive treatment for cancer within 62 days of referral was last achieved in quarter three 2013/14. (Ham and Murray, 2018)

As Martin Powell (2019: 9) points out, official publications by the NHS’s supervisory bodies in the year of the service’s 70th anniversary may not have used the term “crisis” but did include references to the situation being “particularly challenging” (Monitor, 2018: 26). The NHS England (2018: 14) performance report for 2017/18 noted how the NHS in England had “met or exceeded all financial goals set by the Government for the fifth year in a row,” but that the system had suffered from “intensifying and inescapable pressure” as “GP workloads are up, but GP numbers down. There were longer waits last year for A&E and routine surgery, and a particularly difficult winter.” Hence, despite the promise to spare the NHS the brunt of austerity this most popular organization found itself struggling to operate within its fiscal limits. As we shall see below, the NHS became a key battleground in the Brexit referendum with the Leave side explicitly talking about how money “saved” from payments to the European Union (EU) could be spent on the NHS and implicitly suggesting that much of the strain on the service’s resources was a consequence of the open door immigration status for any EU citizen. Yet, even if the attempt at “winning” the political battle over the NHS was unsuccessful, it was clear that the government understood the need to minimize the blame they might take for underfunding the service. On the other hand, the absolute protection afforded the state pension to existing beneficiaries and the relative protection offered to the NHS budget was not a simple hat-tip to shielding all universal programs.2 The Child Benefit program that paid out to all parents with children up to 18 years old, and which had long been subject to skepticism in Conservative ranks with its level being periodically frozen when Tories were in government, finally saw its universal status ended in 2013 (Béland and Waddan, 2014). Importantly, the decision to impose an income test on Child Benefit allowed the government to claim that it was targeting higher income households for benefit reductions, while leaving the program intact for most families with children. Osborne, for example, justified the move in the following by saying, “It is very difficult to justify taxing people on low incomes to pay for the child benefit of those earning so much more than them” (BBC News online, 2010).

102  Handbook on austerity, populism and the welfare state

CUTS TO WELFARE If, however, the retrenchment of the Child Benefit program did target higher income earners in other areas the spending cuts did fall upon more vulnerable households (Edmiston, 2017). When the Cameron-led Coalition government took office and made clear that it would be cutting much of the social welfare budget even as pensions and the NHS were to be spared, there was some debate amongst social policy analysts as to whether this represented a significant change in policy, or whether it was more of a continuation of the previous Labour government’s record of imposing increased conditionality on means-tested welfare programs (Driver, 2011; Lister and Bennett, 2010; Taylor-Gooby, 2012). The benefit of hindsight, however, suggests that this debate was somewhat misplaced, as the financial consequences for beneficiaries of the actions taken to place extra conditionality on receipt of cash welfare after 2010 were clearly more significant than any actions taken under the Blair and Brown governments (Edmiston, 2017; McEnhill and Taylor-Gooby, 2018). From 1997, the Labour governments had increased transfer benefits to pensioners and families with children, while containing transfers for the working aged. As the Coalition government’s policies began to take effect, transfers to pensioners were maintained, but support to families with children was reduced. The government “made a series of specific cuts with large effects on particular groups, including to disability benefits, Council Tax support and Housing Benefit, as well as much greater use of sanctions cutting off benefits from increasing numbers of people” (Hills et al., 2016: 30). One feature of the austerity program was the manner in which local authorities were tasked with implementing many of the cuts. Despite devolution of some powers to the constituent nations of the UK, the country remains a highly centralized state, especially in England, as local governments are often highly dependent on central government funding. Some areas of social policy that are within the remit of local authority are deemed essential and hence it is mandatory for local government to maintain provision. Yet that did not mean that the budgets for those services were protected, and as Mia Gray and Anna Barford (2018: 556–7) observe, “mandatory services, such as adult social care and children’s services, are being squeezed most acutely” meaning that “many local authorities were only able to provide the most basic functions and dropped most preventative measures, early outreach and supplementary services.” Overall, although most local governments prioritized protecting the social care budget, which is by far the biggest single aspect of local government spending, during the years of the Coalition government there was an average cut of 16.7 percent spending per person on social care (Innes and Tetlow, 2015: 318). In the UK, politics is not really local so it is unclear that these cuts were an effective means of “blame avoidance” for the central government, but it was a means of handing down the responsibility for decision-making about exactly how to implement austerity. For its part, the government insisted that it was passing decision-making capacity to local governments, which in turn could act more effectively. For example, speaking in September 2015 as head of a now Conservative government, Prime Minister Cameron (2015) explained how he thought local governments could act more flexibly by taking on a more business-oriented model: The best businesses would never shy away from allowing their customers to shape the way they improve their services. If we are bold enough, government can go one better by actually putting many of those services in the hands of local people. It is also a proven reality that money spent closer to people is often money spent wiser – so we can really deliver more for less.

Austerity, populism, and the politics of blame  103 Simultaneously, however, he also underlined how little autonomy some local authorities actually had by adding that if they failed in their duties they would face sanction, “transform the way you provide services, or those services will be taken over by non-profit trusts or other partnerships.” As it was, local councils responded differently to the cuts, adapting in their own fashion but often in a way that redefined the relationship between local government and vulnerable individuals in their community as the former could no longer simply directly support to the latter with the same level of resource (Hastings et al., 2015).

AUSTERITY AS CHOICE? As described above, Prime Minister Cameron and fellow members of his governments were keen to emphasize that austerity was a necessity rather than a choice, while simultaneously suggesting that those most negatively impacted by government expenditure cuts were often responsible for their own misfortune and needed to change their own behavior rather than continue to look to government for assistance. For some observers, however, the latter part of this framing narrative undermined the claim made in the opening part. In an extraordinary speech in November 2018 that previewed his final report issued six months later, Philip Alston, the United Nations Special Rapporteur on extreme poverty and human rights, denounced the evidence of hardship he had encountered in a visit to the UK and explicitly rejected the notion that the reductions to the social welfare budget were an unfortunate, but necessary, consequence of the austerity drive: Although the provision of social security to those in need is a public service and a vital anchor to prevent people being pulled into poverty, the policies put in place since 2010 are usually discussed under the rubric of austerity. But this framing leads the inquiry in the wrong direction. In the area of poverty-related policy, the evidence points to the conclusion that the driving force has not been economic but rather a commitment to achieving radical social re-engineering. Successive governments have brought revolutionary change in both the system for delivering minimum levels of fairness and social justice to the British people, and especially in the values underpinning it. Key elements of the post-war Beveridge social contract are being overturned. (Alston, 2018)

Unsurprisingly, the Conservative government rejected this analysis. Responding to Alston’s findings, the then newly appointed Secretary for Work and Pensions, Amber Rudd, forcefully rebutted the conclusions and she emphasized, “I must say I was disappointed to say the least by the extraordinary political nature of his language” (Walker, 2018). Yet, even if the case that the UK really did need to engage in extensive debt and deficit reduction in 2010 is accepted, there were alternative means of implementing that agenda. The government could have chosen to raise taxes. This might have been done through increasing so-called “sin taxes” such as those on petrol or alcohol in a way that might have constituted a different plan of blame avoidance. Hence, the depth of the austerity drive and the manner in which welfare state restructuring took place, targeting the most vulnerable was not inevitable.

104  Handbook on austerity, populism and the welfare state

BREXIT, POPULISM, AND THE SOCIAL POLITICS OF BLAME The Brexit referendum was not centrally a question about the direction of social policy or the future of the welfare state. While EU regulations laid down some minimal requirements about working conditions and its presiding institutions might issue calls for social justice (see, for example, European Commission, 2017), the jurisdiction for the development of key areas of social policy such as pensions, healthcare, and social assistance remained with domestic governments. Those countries in the Eurozone had ceded control over key aspects of monetary policy to the European Central Bank (ECB) and this could diminish the sovereignty of national governments to manage their own budgetary affairs in order to maintain the support of the ECB. Most significantly, in the case of Greece, the ECB along with other international financial institutions such as the International Monetary Fund (IMF) could effectively impose a drastic level of austerity on ailing Eurozone member states.3 The UK, however, remained outside the Eurozone, and did not suffer from such external pressure to impose austerity. The Cameron government did still invoke the example of Greece’s plight as a justification of its self-imposed austerity, and some commentators saw this EU-wide obsession with the need for austerity as a response to the Great Recession as a key factor in fracturing politics across the continent. The economics correspondent Larry Elliott (2018), for example, bemoaned how the policies pursued from 2010 onwards “stretched almost to breaking point” the “old social contract between leaders and people” based on the principle that if people “worked hard, they would earn a decent wage and the state would look after them if times were hard.” While it is important not to draw overly simplistic causal lines, the economic impact of the financial crisis and the sustained squeeze on living standards that resulted, combined with the seeming insistence of governments that more pain was necessary, created a febrile political environment. In such circumstances, the emergence of populism across Europe should have been unsurprising. That populism, if sometimes ideologically heterodox, was mostly evident on the political right and most explicitly manifested in the rise of political movements that were hostile to immigration. In this context, the UK did not experience a major and resilient challenge from the far right, but the increasing popularity of the United Kingdom Independence Party (UKIP) did have important consequences. Critically, Prime Minister Cameron offered the prospect of a referendum on EU membership, partially at least to check the rise of UKIP and the defection of Conservative voters to that party. As Tim Bale (2018) observes, the populist Eurosceptic messages that UKIP brought to the fore in the years after the Great Recession had in fact previously been fomented over time by factions within the Conservative Party. In a speech at his first Conservative conference as party leader, David Cameron (2006) had urged the party to move on from its arguments over the EU. Speaking of why the party had lost successive elections he proclaimed, “While parents worried about childcare, getting the kids to school, balancing work and family life – we were banging on about Europe.” That warning did not quieten the Eurosceptic voices within the Conservative Party, and as the referendum approached that voice expressed an increasingly populist rhetoric against “Eurocrats Living in the Brussels Bubble” oppressing the British people (Charteris-Black, 2019). In the aftermath of the Brexit referendum, there was animated discussion about the relationship between austerity and the vote to leave the EU. One particularly influential paper by Thiemo Fetzer (2019) argued that economic frustration caused by austerity mattered at the margins, and that in a close vote the margins matter. Peter Taylor-Gooby (2017: 828) observes that those groups most negatively impacted by “long-term structural factors” such

Austerity, populism, and the politics of blame  105 as “globalisation and labour market change and deindustrialisation” had found their problems “exacerbated” by social policy choices that favored “the better off and pensioners” and were more likely to have voted to leave the EU. Other voices were more skeptical that there was an identifiable causal relationship between austerity and voting leave (McKay and Bailey, 2018), while others prioritized cultural explanations for the referendum result (Norris and Inglehart, 2019). The purpose here is not to re-litigate or adjudicate that argument. More simply, it is to note one important way in which advocates of Brexit insisted that leaving the EU would release funds that could be spent on improving the welfare of the country, most especially through an extra influx of cash for the NHS. This was not explicitly framed in terms of a release from “austerity,” which was unsurprising given that many of the leaders of the “Leave” campaign had been part of the government implementing austerity. From this angle, it could be understood as a form of blame generation for past domestic austerity decisions. Simultaneously, there was a distinctly populist edge to the (in)famous slogan on the Brexit bus, which was used as a backdrop for many rallies, reading “We send the EU £350 million a week. Let’s fund our NHS instead.” This prompted an extraordinary intervention from Remain-supporting former Conservative Prime Minister, John Major, who ridiculed the notion that leaders of the Leave campaign cause such as Boris Johnson and Michael Gove could be seen as champions of the NHS. Speaking on the BBC, Major mocked, “The NHS is about as safe with them as a pet hamster would be with a hungry python” (Mason, 2016). Nevertheless, one of the Labour figures most prominently associated with the Leave campaign, MP Gisela Stuart, commented, “Every week we send £350m to Brussels. I’d rather that we control how to spend that money, and if I had that control I would spend it on the NHS” (Reuben, 2016). This is a clear example of blame generating strategy depicting the EU as a burden responsible for some of the problems facing the NHS. The blame generating message about how the money sent to the EU could be spent on the NHS was also sometimes attached to the more explicitly populist theme of immigration control. In particular, the Leave campaign referred to how there “were 475,000 live births to mothers from other EU countries between 2005 and 2014 … The cost of providing NHS services to those families could be over £1.33bn. If we remain in the EU, the NHS will be put under more and more pressure and the A&E crisis will get even worse” (Vote Leave, 2016). This was a blame generating message that emphasized the increased strain on the NHS over the preceding years, which framed that strain as a result of immigration rather than austerity. The fact was that the £350 million figure was an exaggeration and many NHS workers were foreign EU nationals. In 2018, this included 10 percent of doctors and 7 percent of nurses working for the NHS (Bloor, 2018). These data points confirm the assertion, in the weeks before the referendum, of Remain-backing and then Tory MP, Sarah Woollaston: “if you meet a migrant in the NHS, they are more likely to be treating you than ahead of you in the queue” (Bloor and Street, 2016). Such sentiments, however, were swept aside by the populist tide. This analysis shows that there was a clear if not always logically consistent interaction between the policies of austerity and the use of populist frames from 2010 onwards. The case of the NHS provides an interesting case study of the dynamics at play. Prime Minister Cameron insisted that his governments were protecting spending on this “jewel in the crown” of the welfare state even though real spending increased at the slowest pace in the organization’s history. Here Cameron was trying to avoid blame for the increasing evidence of the NHS missing its treatment targets. Then during the referendum campaign as Cameron led the

106  Handbook on austerity, populism and the welfare state Remain camp, the Prime Minister found himself outflanked by Leave campaigners, mostly to his ideological right, who insisted that the NHS, a bastion of state collectivism, was being starved of cash because of monies being sent to Brussels. That argument, dismissed by most experts as nonsensical and often made by those most skeptical of government-run social programs, was perceived as critical to the success of the Leave campaign. According to Dominic Cummings, the campaign director for Vote Leave, the “£350m/NHS was … clearly the most effective argument not only with the crucial swing fifth but with almost every demographic” (Brown, 2017).

CONCLUSION This chapter stresses the ideational components and the potential interaction of austerity and populism, two key phenomena closely associated with the contemporary politics of social policy. As the analysis of the UK case suggests, austerity and populism are two distinct phenomena that can feed one another under specific historical and political circumstances. Simultaneously, the concepts of blame avoidance and blame generation (Weaver, 2018) are especially helpful to capture the interaction between austerity and populism. This is the case in part because populist rhetoric can serve as blame avoidance strategy for austerity measures by finding scapegoats and/or by diverting attention away from them. More generally, our analysis suggests that paying close attention to the politics of ideas and, especially, framing processes, is most appropriate to grasp the nature of austerity and populism as well as the ways in which they can interact in the field of social policy and beyond. Future research on austerity and populism could use our ideational and blame-centric lens to explore other empirical cases capable of providing more insight on how these two phenomena relate to one another. Here comparative and historically minded case studies could prove especially helpful to complement the above analysis of the UK case.

NOTES 1. The increase in Social Security spending over New Labour’s time in power was in fact lower than under the Thatcher and Major governments from 1979 to 1997. This is partially explained by lower unemployment rates (Chote et al., 2010a: 8). 2. The UK basic state pension should not really be classed as a universal program in the proper sense of the word as it remained a social insurance program, with eligibility for the full pension determined by contributions. 3. On the Greek crisis in the aftermath of the financial crisis see Bond (2018) for a condemnation of the EU’s “deliberate, slow destruction of a small and vulnerable society of fellow Europeans, all in the name of financial prudence and necessity” from a writer sympathetic to the EU’s broader project.

REFERENCES Alston, Philip (2018) Statement on Visit to the United Kingdom, by Professor Philip Alston, United Nations Special Rapporteur on Extreme Poverty and Human Rights, https://​www​.ohchr​.org/​ documents/​issues/​poverty/​eom​_gb​_16nov2018​.pdf (accessed January 22, 2021).

Austerity, populism, and the politics of blame  107 Bale, Tim (2018) Who leads and who follows? The symbiotic relationship between UKIP and the Conservatives – and populism and Euroscepticism. Politics, 38 (3): 263–77. Barker, Alex (2007) Osborne makes key spending pledge. Financial Times, September 3, https://​www​.ft​ .com/​content/​c61c3b9c​-598b​-11dc​-aef5​-0000779fd2ac (accessed January 22, 2021). Barr, N. and P. Diamond (2010) Pension Reform. New York: Oxford University Press. BBC News online (2010) Child benefit to be scrapped for higher taxpayers, October 4, http://​www​.bbc​ .co​.uk/​news/​uk​-politics​-11464300 (accessed January 22, 2021). Béland, D. (2016), Ideas and Institutions in Social Policy Research, Social Policy & Administration, 50 (6): 734–50. Béland, Daniel and Robert Henry Cox (2011) Introduction: ideas and politics. In D. Béland and R.H. Cox (eds), Ideas and Politics in Social Science Research. New York: Oxford University Press, pp. 3–20. Béland, Daniel and Alex Waddan (2014) Conservatives, partisan dynamics, and the politics of universality: reforming universal social programmes in the United Kingdom and Canada. Journal of Poverty and Social Justice, 22 (2): 83–97. Béland, Daniel and Alex Waddan (2015) Breaking down ideas and institutions: the politics of tax policy in the U.S. and the U.K. Policy Studies, 36 (2): 176–95. Bloor, Karen (2018) NHS reliance on EU staff in numbers – full scale of Brexit risk revealed. The Conversation, October 23, https://​theconversation​.com/​nhs​-reliance​-on​-eu​-staff​-in​-numbers​-full​ -scale​-of​-brexit​-risk​-revealed​-105326 (accessed January 22, 2021). Bloor, Karen and Andrew Street (2016) The truth about migrants and the NHS. The Conversation, June 13. Blyth, Mark (2013) Austerity: The History of a Dangerous Idea. Oxford: Oxford University Press. Bond, Jonathon (2018) A Greek tragedy: how the EU is destroying a country. The Spectator, October 6, https://​www​.spectator​.co​.uk/​2018/​10/​a​-greek​-tragedy​-how​-the​-eu​-is​-destroying​-a​-country/​ (accessed January 22, 2021). Brown, Jessica (2017) The infamous 350 million for the NHS won the Brexit referendum, according to the boss of the Brexit campaign. The Independent, February 8, https://​www​.indy100​.com/​article/​voted​ -brexit​-nhs​-promise​-campaign​-director​-350million​-battle​-bus​-voting​-7568381 (accessed January 22, 2021). Cameron, David (2006) Leader’s Speech at the Bournemouth Conservative Party Conference, http://​ www​.br​itishpolit​icalspeech​.org/​speech​-archive​.htm​?speech​=​314 (accessed January 22, 2021). Cameron, David (2009) The Age of Austerity, speech by David Cameron, Cheltenham, UK, April 26, https://​conservative​-speeches​.sayit​.mysociety​.org/​speech/​601367 (accessed January 22, 2021). Cameron, David (2015) Prime Minister: My Vision for a Smarter State, https://​www​.gov​.uk/​government/​ speeches/​prime​-minister​-my​-vision​-for​-a​-smarter​-state (accessed January 22, 2021). Campbell, John L. (2004) Institutional Change and Globalization. Princeton, NJ: Princeton University Press. Charteris-Black, J. (2019) Conceptual metaphors: “Eurocrats living in the Brussels bubble”. In Metaphors of Brexit. Basingstoke: Palgrave Macmillan, pp. 137–60. Chote, Robert, R. Crawford, C. Emmerson, and G. Tetlow (2010a) Public Spending under Labour, Institute for Fiscal Studies, 2010 Election Briefing Note No. 5 (IFS BN92), https://​www​.ifs​.org​.uk/​ bns/​bn92​.pdf (accessed January 22, 2021). Chote, R., R. Crawford, C. Emmerson, and G. Tetlow (2010b) Filling the Hole: How Do the Three Main UK Parties Plan to Repair the Public Finances? London: Institute for Fiscal Studies. Retrieved from http://​www​.ifs​.org​.uk/​bns/​bn99​.pdf (accessed January 22, 2021). Conservative home (2007) Tories will match Labour’s public spending for next three years, http://​ conservativehome​.com/​. Cox, Robert Henry (2001) The social construction of an imperative: why welfare reform happened in Denmark and the Netherlands but not in Germany. World Politics, 53: 463–98. Driver, S. (2011) Welfare reform and coalition politics in the age of austerity. In S. Lee and M. Beech (eds), The Cameron-Clegg Government: Coalition Politics in the Age of Austerity. Basingstoke: Palgrave Macmillan, pp. 105–17. Edmiston, Daniel (2017) Review article: welfare, austerity and social citizenship in the UK. Social Policy and Society, 16 (2): 261–70.

108  Handbook on austerity, populism and the welfare state Elliott, Larry (2018) Greece’s bailout is finally at an end – but has been a failure. Guardian, August 19, https://​www​.theguardian​.com/​world/​2018/​aug/​19/​greeces​-bailout​-is​-finally​-at​-an​-end​-but​-has​-been​ -a​-failure (accessed January 22, 2021). European Commission (2017) European Pillar of Social Rights, https://​ec​.europa​.eu/​commission/​ publications/​european​-pillar​-social​-rights​-booklet​_en (accessed January 22, 2021). Evandrou, M. and J. Falkingham (2009) Pensions and income security in later life. In John Hills, Tom Sefton, and Kitty Stewart (eds), Towards a More Equal Society? Poverty, Inequality and Policy since 1997. Bristol: Policy Press, pp. 157–76. Fetzer, Thiemo (2019) Did austerity cause Brexit? American Economic Review, 109 (11): 3849–86. Genieys, William and Marc Smyrl (eds) (2008) Elites, Ideas, and the Evolution of Public Policy. London: Palgrave. Goldstein, Judith and Robert O. Keohane (eds) (1993) Ideas and Foreign Policy: Beliefs, Institutions, and Political Change. Ithaca, NY: Cornell University Press. Gray, Mia and Anna Barford (2018) The depths of the cuts: the uneven geography of local government austerity. Cambridge Journal of Regions, Economy and Society, 11: 541–63. Greve, Bent (2019) Welfare, Populism and Welfare Chauvinism. Bristol: Policy Press. Ham, Chris and Richard Murray (2018) The NHS 10-year Plan: How Should the Extra Funding be Spent? The King’s Fund, https://​www​.kingsfund​.org​.uk/​publications/​nhs​-10​-year​-plan (accessed January 22, 2021). Hastings, Annette, Nick Bailey, Maria Gannon, Kirsten Besemer, and Glen Bramley (2015) Coping with the cuts? The management of the worst financial settlement in living memory. Local Government Studies, 41 (4): 601–21. Hills, John, Paola DeAgostini, and Holly Sutherland (2016) Benefits, pensions, tax credits and direct taxes. In Ruth Lupton, Tania Burchardt, John Hills, Kitty Stewart, and Polly Vizard (eds), Social Policy in a Cold Climate: Policies and Their Consequences since the Crisis. Bristol: Policy Press, pp. 11–34. Innes, David and Gemma Tetlow (2015) Delivering fiscal squeeze by cutting local government spending. Fiscal Studies, 36 (3): 303–25. Institute for Fiscal Studies and the Health Foundation (2018) Securing the Future: Funding Health and Social Care to the 2030s: Executive Summary. London: NHS Confederation. Jacobs, Alan M. (2009) How do ideas matter? Mental models and attention in German pension politics. Comparative Political Studies, 42 (2): 252–79. Laclau, Ernesto (2005) On Populist Reason. London: Verso. Lister, Ruth and Fran Bennett (2010) The new “champion of progressive ideals”? Cameron’s Conservative Party: poverty, family policy and welfare reform. Renewal, 18 (1): 84–109. MacLeavy, Julie (2011) A “new politics” of austerity, workfare and gender? The UK coalition government’s welfare reform proposals. Cambridge Journal of Regions, Economy and Society, 4: 355–67. Mason, Rowena (2016) John Major: NHS at risk from Brexit “pythons” Johnson and Gove. Guardian, June 5, https://​www​.theguardian​.com/​politics/​2016/​jun/​05/​john​-major​-nhs​-risk​-brexit​-pythons​ -joh​nson​-and​-gove (accessed January 22, 2021). McEnhill, Libby and Peter Taylor-Gooby (2018) Beyond continuity? Understanding change in the UK welfare state since 2010. Social Policy & Administration, 52 (1): 252–70. McKay, L. and J. Bailey (2018) Was Brexit really caused by austerity? Here’s why we’re not convinced. The Conversation, https://​theconversation​.com/​was​-brexit​-really​-caused​-by​-austerity​-heres​-why​ -we​re​-not​-convinced​-102124 (accessed January 22, 2021). Mehta, Jal (2011) The varied roles of ideas in politics: from “whether” to “how”. In D. Béland and R.H. Cox (eds), Ideas and Politics in Social Science Research. New York: Oxford University Press, pp. 23–46. Monitor (2018) Annual Report and Accounts 1 April 2017 to 31 March 2018. London: NHS Improvement, https://​assets​.publishing​.service​.gov​.uk/​government/​uploads/​system/​uploads/​attachment​_data/​file/​ 725242/​Monitor​_annual​_report​_and​_accounts​_201718​_web​.pdf (accessed January 22, 2021). Mudde, Cas and Cristóbal Rovira Kaltwasser (2017) Populism: A Very Short Introduction. Oxford: Oxford University Press. Müller, Jan-Werner (2016) What is Populism? Philadelphia, PA: University of Pennsylvania Press.

Austerity, populism, and the politics of blame  109 NHS England (2018) Performance Report – NHS England Annual report 2017/8, https://​www​.england​ .nhs​.uk/​wp​-content/​uploads/​2018/​07/​performance​-report​-201718​.pdf (accessed January 22, 2021). Norris, Pippa and Ronald Inglehart (2019) Cultural Backlash: Trump, Brexit and Authoritarian Populism. New York: Cambridge University Press. Office for National Statistics (2018) The 2008 recession 10 years on, https://​www​.ons​.gov​.uk/​economy/​ gros​sdomesticp​roductgdp/​articles/​the200​8recession​10yearson/​2018​-04​-30 (accessed January 22, 2021). Office of Budget Responsibility (2015) Welfare Trends Report, 2015, https://​obr​.uk/​docs/​dlm​_uploads/​ 49754​-OBR​-Welfare​-Accessible​-v0​.2​.pdf (accessed January 22, 2021). Osborne, George (2009) Speech to the Conservative spring conference, April 27, https://​www​.politics​ .co​.uk/​comment​-analysis/​2009/​04/​27/​tory​-spring​-conference​-speeches​-in​-full (accessed January 22, 2021). Osborne, George (2012) George Osborne’s speech to the Conservative conference: full text. New Statesman, October 8, https://​www​.newstatesman​.com/​blogs/​politics/​2012/​10/​george​-osbornes​ -speech​-conservative​-conference​-full​-text (accessed January 22, 2021). Owen, Paul (2010) Ex-Treasury secretary Liam Byrne’s note to his successor: there’s no money left. Guardian, May 17, https://​www​.theguardian​.com/​politics/​2010/​may/​17/​liam​-byrne​-note​-successor (accessed January 22, 2021). Pierson, Paul (1994) Dismantling the Welfare State? Reagan, Thatcher, and the Politics of Retrenchment. New York: Cambridge University Press. Pierson, Paul (1998) Irresistible forces, immovable objects: post-industrial welfare states confront permanent austerity. Journal of European Public Policy, 5 (4): 539–60. Powell, Martin (2019) The English National Health Service in a cold climate. In Elke Heins, James Rees, and Catherine Needham (eds), Social Policy Review 31, Bristol: Policy Press, pp. 7–28. Reuben, Anthony (2016) Reality check: would Brexit mean extra £350m a week for NHS? BBC News, April 15, https://​www​.bbc​.co​.uk/​news/​uk​-politics​-eu​-referendum​-36040060 (accessed January 22, 2021). Riley, Jon and Robert Chote (2014) Crisis and consolidation in the public finances. Office for Budget Responsibility. Tapsfield, James and Ellen Branagh (2010) Tackling debt will change everyone’s lives, says David Cameron. Independent, June 7, https://​www​.independent​.co​.uk/​news/​uk/​politics/​tackling​-debt​-will​ -change​-everyones​-lives​-says​-david​-cameron​-1993587​.html (accessed January 22, 2021). Taylor-Gooby, Peter (2012) Root and branch restructuring to achieve major cuts: the social ambitions of the Coalition. Social Policy and Administration, 46 (1): 61–82. Taylor-Gooby, Peter (2017) Re-doubling the crises of the welfare state: the impact of Brexit on UK welfare politics. Journal of Social Policy, 46 (4): 815–35. Timmins, Nicholas (1995) The Five Giants: A Biography of the Welfare State. London: Fontana Press. Vizard, Polly, Polina Obolenskaya, and Emily Jones (2016) Health. In Ruth Lupton, Tania Burchardt, John Hills, Kitty Stewart, and Polly Vizard (eds), Social Policy in a Cold Climate: Policies and Their Consequences since the Crisis. Bristol: Policy Press, pp. 147–86. Vote Leave (2016) The EU immigration system is immoral and unfair, http://​www​.​voteleavet​akecontrol​ .org/​briefing​_immigration​.html (accessed January 22, 2021). Walker, Peter (2018) Amber Rudd condemns UN poverty report in combative return to frontline politics. Guardian, November 19, https://​www​.theguardian​.com/​politics/​2018/​nov/​19/​amber​-rudd​-un​-poverty​ -report​-return​-frontline​-politics (accessed January 22, 2021). Weaver, R. Kent (1986) The politics of blame avoidance. Journal of Public Policy, 6 (4): 371–98. Weaver, R. Kent (2018) The nays have it: how rampant blame generating distorts American policy and politics. Political Science Quarterly, 133 (2): 259–89. Winnett, Robert (2010) David Cameron: Tories are the party of the NHS. Daily Telegraph, January, https://​www​.telegraph​.co​.uk/​news/​election​-2010/​6932043/​David​-Cameron​-Tories​-are​-the​-party​-of​ -the​-NHS​.html (accessed January 22, 2021).

9. The social legitimacy of European welfare states after “the age of austerity” Femke Roosma

INTRODUCTION After the Great Recession that caused varying levels of economic contraction in European countries, welfare states responded differently with either harsh, mild or balanced austerity measures, depending on the severity and specificities of the crisis in their country and the current social policy schemes in place (Farnsworth and Irving 2012; Ólafsson et al. 2019; Taylor-Gooby et al. 2017; van Kersbergen et al. 2014). Together with the influx of refugees in 2015, the rise of populist parties and welfare chauvinism (Greve 2019), these developments changed (the images of) European welfare states as we knew them. The main idea was that the welfare state had become unaffordable, summarized in the words of the President of the European Commission, Barroso: “There is no such thing as a ‘free deficit’” (Farnsworth and Irving 2012, p. 135). It is easy to imagine that there will be an impact of these contextual changes on the social legitimacy of the welfare state and solidarity with welfare beneficiaries (Kumlin and Stadelmann-Steffen 2014; Laenen 2019; Naumann et al. 2016). Traditional welfare state questions like “who should get what and why” are debated in relation to pension and health care reform, increasing conditionality of (unemployment) benefits, limiting welfare benefits for migrants and new proposals for minimum-income schemes to protect the poor, and support levels for these proposals show substantial cross-national variation (Meuleman et al. 2018). Although previous studies of public support for the welfare state have shown that welfare attitudes are remarkably stable over time and remain quite unaffected by changes in contextual circumstances (Brooks and Manza 2006; Laenen and Meuleman 2017; Meuleman 2019; Svallfors 2011), the shock of the Great Recession and the policy responses afterwards might have had a more substantial impact on how people support welfare policies and evaluate their social system. For a systematic analysis of how welfare state legitimacy is affected in the decade of austerity after the crisis, a multidimensional perspective of welfare state attitudes is warranted (Roosma et al. 2013; Svallfors 1991; van Oorschot and Meuleman 2012b). Taking a multidimensional perspective allows us to think about potential differences in impact on different welfare attitudes. On the one hand, we might expect, for instance, that support for the government in providing a decent standard of living for our elderly and the sick might have become stronger after an economic crisis as people seek more income security after insecure times. On the other hand, it can be anticipated as well that people might have become more critical about the actual ability of the welfare state to provide this decent standard of living, especially in cases where austerity measures were implemented. Expectations about change in perceptions of abuse of benefits can be rather ambiguous: either abuse perceptions could have become stronger as people fear that others take advantage of scarce provisions, while a crisis-induced idea that people “like us” are in need of support can lower 110

The social legitimacy of European welfare states after “the age of austerity”  111 abuse perceptions as well (Roosma et al. 2016a). A multidimensional perspective allows the possibility that the welfare state can become more socially legitimate on certain dimensions, while for other dimensions support can decline or critical evaluations can take over. And it is likely that this will be different for European countries and populations that are experiencing varying levels of austerity and material deprivation. In this chapter I analyse the possible impact of severe material deprivation and differences in responses to the crisis on the (change in) attitudes towards different dimensions of the welfare state. I take a multidimensional perspective and analyse how public support across a wide range of welfare state dimensions has changed – if at all – between 2008/9 and 2016/17, in 17 European countries controlling for composition effects. I will use data from two waves of the European Social Survey (ESS) from 2008/9 and 2016/17, which include many different welfare attitudes and capture the important dimensions of welfare state legitimacy. Results will show a systematic analysis of change and stability in welfare state legitimacy in Europe after a decade of austerity and financial hardship. In this chapter, firstly a multidimensional perspective on welfare state legitimacy is outlined and the support for different dimensions in 2008/9 – at the beginning of the crisis – are discussed. Secondly, differences in experiences with the different crises, policy responses and welfare outcomes are discussed, as well as expectations about the impact on different dimensions of welfare state legitimacy. After discussing the data and methodological approach to study change in attitudes, the results section presents a systematic overview of change in welfare attitudes between 2008/9 and 2016/17, across dimensions and countries. Finally, results are discussed together with conclusions in the final section.

WELFARE STATE LEGITIMACY: AN ANALYTICAL PERSPECTIVE A Multidimensional Framework of Welfare State Legitimacy In the literature on welfare state attitudes, for decades scholars have suggested studying opinions towards the welfare state in a multidimensional perspective, instead of assuming that people are either pro- or anti-welfare state (Cnaan 1989; Sihvo and Uusitalo 1995a; Svallfors 1991; van Oorschot and Meuleman 2012b). However, the unavailability of comparative cross-national data hampered systematic analysis until 2008 when the ESS ran a module with a broad range of questions about the support for and the functioning of the welfare state. A theoretical framework of welfare dimensions distinguished four conditions for the welfare state to be legitimate – substantial, redistributional and procedural justice and just outcomes (Roosma 2016; Rothstein 1998) – by which dimensions of welfare support should be analysed. Substantive justice can be obtained if people consider the goals and range and degree of government activities as just. Questions relate to whether the welfare state should redistribute, tackle poverty and inequality and promote social inclusion, as well as in which domains the government should take a role and increase or decrease social spending (Roller 1995). Redistributional justice refers to dimensions that assess, on the one hand, who is deserving of benefits and under what conditions and, on the other hand, who should contribute (more or less) to the welfare state in terms of taxes and contributions (Roosma et al. 2016b;

112  Handbook on austerity, populism and the welfare state van Oorschot 2006). Both substantive and procedural justice refer to what, according to the people, the welfare state should do. Procedural justice entails that the implementation of welfare benefits and services is efficient and effective. People’s evaluations of the welfare bureaucracy, levels of free riding and welfare abuse, and of welfare non-take-up are central here. The question is to what extent the welfare state is able to redistribute in a fair, simple and cheap manner (Roosma et al. 2014b; Rothstein 1998)? Finally, the condition of just outcomes refers to the evaluation of welfare benefits and services and their unintended consequences. Does the welfare state succeed in tackling poverty and inequality and providing sufficient benefits and services for those who need it and are deserving of it? And to what extent do people perceive unintended outcomes, for instance, for the economy or moral or social consequences of (a strong role of) the welfare state (Fenger et al. 2011; van Oorschot 2010)? To draw conclusions about whether or not the welfare state is still legitimate or whether support for the welfare state has changed after the crisis in Europe, we should take into account multiple conditions for welfare state legitimacy and analyse support towards multiple dimensions. This analysis starts with an overview of what we know about support for the welfare state in 2008/9. Social Legitimacy of European Welfare States: High and Stable Support, but Also Criticism In 2008/9, at the beginning of the financial crisis, support for the substantive dimension of European welfare states and their efforts to provide benefits and services for those in need of support was very high across Europe. Among the general public there seems to be no doubt that the welfare state should have a substantial role in redistributing between the rich and the poor and tackle severe poverty (Roosma et al. 2013; Svallfors 2012). Support was especially high in Eastern and Southern European countries. For instance, in Latvia, Bulgaria, Spain, Hungary and Portugal the demand for a strong role of government was high in 2008/9 (Roosma et al. 2014a). But also in the Western and Nordic countries the majority of the population expressed enthusiasm for a substantial role of the welfare state in providing a reasonable standard of living for the elderly and the unemployed as well as health care and child care services and paid care leave. In countries like Switzerland, the Netherlands, France, Belgium and Great Britain, support for a stronger role of the government was relatively lower in 2008/9 compared to the average (Roosma et al. 2014a). This high support is in sharp contrast with evaluations of the outcomes of the welfare state, especially the concrete policy outcomes. Where people support a strong role of the welfare state, they express strong negative opinions about the level of benefits and quality of services and about the unintended outcomes of the welfare state. This is especially the case in the East and South of Europe, while in Nordic countries and Switzerland and the Netherlands, people are relatively more positive about welfare outcomes (Roosma et al. 2014a; van Oorschot and Meuleman 2012a; van Oorschot et al. 2012). And the procedural dimension is considered to be “the Achilles’ heel” of welfare state support; with high levels of perceived benefit abuse as well as benefit non-take-up (Ervasti 2012; Goul Andersen 1999; Roosma et al. 2016a). As a result of the availability of new cross-sectional data, scholars find that the Europeans do not have a comprehensive attitude towards the welfare state, but they combine both positive attitudes and critical evaluations

The social legitimacy of European welfare states after “the age of austerity”  113 towards different dimensions of the welfare state (Ervasti et al. 2012; Roosma 2016; Svallfors 2012; van Oorschot et al. 2012). Is there reason to believe that these attitudes would have substantially changed after a decade of crises and political interventions? Maybe not. We know from previous studies that welfare attitude change is modest and welfare support is remarkably stable over time (Brooks and Manza 2007; Laenen and Meuleman 2017; Pekari et al. 2018; Sihvo and Uusitalo 1995b; Svallfors 2011). And a study of welfare attitudes in the United States shows that the economic crisis hardly affected support (Brooks 2012). However, there studies also present differences for different dimensions in different countries. Especially in Great Britain, there seems to be a growing concentration of criticism, anti-welfare attitudes and negative opinions towards welfare beneficiaries in the previous decades (Deeming and Johnston 2018). A cross-national study of welfare attitudes (Edlund 2009) shows that support for health care and pensions is stable over time across almost all societies, while change in attitudes towards unemployment benefits is more dependent on the context (see also Naumann et al. 2016). And in liberal, market-oriented societies attitude change is more likely to occur (Edlund 2009). Welfare attitudes can thus be expected to be relatively stable, although change can occur for specific dimensions and depending on the contextual circumstances. The Great Recession and the Welfare State If there is a period in which contextual circumstances can be expected to impact welfare state legitimacy, it is in the wake of the financial crisis. Although it is often suggested that the universal policy response to the crisis was welfare retrenchment and austerity, several studies show a far more nuanced picture (Farnsworth and Irving 2012; Ólafsson et al. 2019; Shahidi 2015). Contextual circumstances were very different for countries and this led to different possibilities and barriers for policymakers to act. Ólafsson et al. (2019) distinguish the key factors that were central in explaining differences in welfare outcomes – in terms of financial hardship, unemployment, inequality, and so on – for European welfare states. Firstly, differences in the depth and persistence of the crisis was probably the most important factor in explaining outcomes for welfare states. Where some countries were severely exposed to the crisis of the financial sector (Great Britain and Iceland), other countries were more strongly hit by the Eurocrisis that followed after (Greece, Ireland, Spain and Portugal), or by the worldwide economic downfall (Hungary and Estonia) (Farnsworth and Irving 2012). Several countries suffered from high levels of contraction of gross domestic product (GDP) and high unemployment rates (both were strongly related). Of course Greece is among these countries, but also the Baltic States, other Southern European welfare states and Ireland. In the years after the crisis, some of these countries bounced back and strongly recovered from it, such as the Baltic States and Ireland (Ólafsson et al. 2019). Secondly, different welfare regimes, with differences in strength, resources and effectiveness, were capable of softening the impact of the crisis to varying degrees (Ólafsson et al. 2019). The Nordic and Western European countries with generous social protection were able to dampen the effects of the crisis. While in Eastern and Southern European countries these possibilities were much more limited. An important aspect of this initial capacity position was also the governmental debt level. Depending on the debt level countries had the possibility to opt for stimulus packages or retrenchment and austerity measures. Highly indebted governments had little room to maneuver (Farnsworth and Irving 2012; Ólafsson et al. 2019).

114  Handbook on austerity, populism and the welfare state Thirdly, there were of course differences in policy responses. Some countries had the possibility to choose to either expand, maintain or retrench social policy programs, to increase certain taxes or not, and to either stimulate the economy to maintain demand or to pursue austerity (Ólafsson et al. 2019). The choice in the policies that were implemented also depended on the political actors in charge (Blyth 2013); in many European countries, after the crisis right-wing parties were elected (Farnsworth and Irving 2012). Other countries that were strongly affected by the Eurozone crisis, however, had little option but to implement budget cuts and opt for austerity, under the influence of the European Union (Blyth 2013). Ólafsson et al. (2019) distinguish three groups of countries, based on changes in underlying government revenues and government spending, with different levels of emphasis on austerity: harsh, mild and balanced austerity countries. They also show a relationship with changes in financial hardship in countries; harsh austerity regimes showed the strongest increase in financial hardship. The levels of financial hardship were substantially different in European welfare states. In the Nordic countries, the initial low levels of financial hardship were largely stable during the crisis, and also the Continental countries counteracted the impact of the crisis and financial hardship levels did not change much. In Ireland and Great Britain, however, financial hardship rose significantly during the crisis, and recovery occurred only at modest speed. The Mediterranean countries saw their initial differences in living standards with the Continental countries grow during the crisis years and financial hardship levels were higher after the crisis years than before. The Baltic countries, as well as Bulgaria and Hungary, who already had high levels of financial hardship before the crisis, saw some increase, but recovered relatively well from it. Other European countries showed a varying pattern (Ólafsson et al. 2019). It is not easy to formulate expectations on the impact of the Great Recession on welfare state legitimacy based on the variation in factors involved in determining welfare outcomes after the crisis. I therefore suggest analysing the impact of two factors on the different welfare state dimensions without formulating clear expectations. Firstly, the role of policy responses that were partly bound to the initial capacity of welfare states to act. Harsh, mild and balanced responses can be distinguished and countries are categorized accordingly. Secondly, the impact of severe material deprivation as the most important welfare outcome of the crisis. What is the different impact of poverty in 2008/9 compared to 2016/17?

DATA AND METHODS Data To systematically analyse change in attitudes since 2008, I use data from the ESS round 4 and round 8, which include a repeated cross-section module of welfare attitudes. Data are collected in 2008 and 2009 (round 4) in the early stages of the financial crisis, and in 2016 and 2017 (round 8) in the years after recovery from the crises. The surveys have 19 overlapping countries of which I selected the 17 European countries (excluding Russia and Israel): Belgium (BE), Czech Republic (CZ), Estonia (EE), Finland (FI), France (FR), Germany (DE), Great Britain (GB), Hungary (HU), Ireland (IE), the Netherlands (NL), Norway (NO), Poland (PL), Portugal (PT), Slovenia (SI), Spain (ES), Sweden (SE) and Switzerland (CH). From both surveys, 16 overlapping variables are selected based on their match with the multidimensional framework discussed above. Unfortunately, not all dimensions could be

The social legitimacy of European welfare states after “the age of austerity”  115 operationalized due to limited data availability (the redistribution dimension in particular is lacking), but the substantial dimensions of goal and range, the procedural dimension of abuse and underuse of benefits and several welfare outcome dimensions are covered. The selected survey questions are presented in Table 9.1. Because there is ample variation within the operationalization of dimensions, the individual measures are not added to sum scales as done in previous research (Roosma et al. 2014a; van Oorschot 2010). Measures have either a 5-point scale (offering the response categories (strongly) agree, neither agree nor disagree, (strongly disagree), recoded such that a higher score indicates agreement with the presented statement) or a 11-point scale (offering a response scale running from not at all the government’s responsibility (0) to entirely the government’s responsibility (10), or from extremely bad (0) to extremely good (10)). For the descriptive statistics, design and population weights provided by the ESS are used. In the second stage of the analysis, the selected welfare attitude measures are standardized to facilitate better interpretability in the multivariate analyses. In these multivariate analyses several control variables are added to control for population composition: subjective income, employment status, educational level, political affiliation, age and gender. In the final stage of the analysis a measure of “austerity regime” is included, using dummies of countries implementing harsh austerity (EE, ES, GB, HU, IE, PL, PT and SI), mild austerity (BE, CZ, FI, FR and NL) and balanced policy responses (DE, NO, SE, CH). The categorization is based on the assessment and calculations of Ólafsson et al. (2019) of countries’ emphasis on austerity determined on the basis of changes to underlying government revenues and spending between 2009 and 2013. Unfortunately, this assessment is based on data from 2009 to 2013 only. However, it is assumed that the level of emphasis on austerity has impacted the development of the welfare state in the following years. In addition, a variable measuring severe material deprivation is used, which is the percentage of the population that cannot afford at least four of the following nine items: to pay their rent, mortgage or utility bills; to keep their home adequately warm; to face unexpected expenses; to eat meat or protein regularly; to go on holiday; a television set; a washing machine; a car; a telephone. This indicator is taken from Eurostat for the years before the survey (2007 and 2015 respectively). Methods The analyses proceed in the following steps. Firstly, the descriptive statistics for the 16 different welfare dimensions are presented in Table 9.1. Secondly, the estimates of the survey year (for 2016/17) from the random intercept multilevel regression model, controlled for the individual level control variables, are presented in Table 9.2. These estimates indicate the direction and significance of the change in attitudes for the full sample of countries. To make effect sizes comparable, the dependent variables are standardized. In addition, the same models were estimated with a random slope for the year dummy. A random slope allows the effect of the year 2016/17 to vary for different countries. In other words, the effect of the year 2016/17 on the welfare attitudes is different in different countries. A likelihood ratio test was performed for the two models to test whether including a random slope results in a substantially different and better model. The likelihood ratio is presented in Table 9.2 as well. Based on both indicators the welfare dimensions with the most substantial change in attitudes over the two waves and strongest differences across countries (i.e. the models that improved the most in terms of model fit after including a random slope), were selected for further analyses. These include

116  Handbook on austerity, populism and the welfare state two attitudes towards the role of government (in providing a decent standard of living for the elderly and child care), two evaluations of the intended outcomes of the welfare state (state of health care and standard of living of the unemployed) and one attitude towards an unintended outcome (the costs of social benefits and services for businesses). In a third step, the five selected dimensions are analysed in more detail and related to approaches to the crisis. For the selected dimensions, random slopes for the year of the survey (2016/17) are presented per country, indicating change in welfare attitudes per country controlled for the individual level control variables mentioned above. Results are presented in Figures 9.1 to 9.5. In relating these dimensions to approaches to the crisis, “austerity regime” dummies and the measure of severe material deprivation are included separately in the models (taking account of the fact that the number of countries is limited) to explain differences across countries. For both these austerity dummies and severe material deprivation we test models using interaction effects with the survey year, to see if the effect gets stronger or weaker in 2016/17.

RESULTS Table 9.1 summarizes the descriptive statistics. At first glance, the results show very minimal changes in welfare attitudes between 2008/9 and 2016/17. The major financial and economic crises and the policy responses of austerity, overall seem to have hardly affected Europeans’ welfare attitudes. In general, people support redistribution of welfare and a strong role of the government in providing benefits and services. At the same time, people worry about both welfare abuse and underuse, and they show (strong) criticism towards the standard of living of the unemployed and the elderly, as well as the state of health care. People have differentiated opinions about the welfare state’s impact on poverty, equality, the economy and business, and its consequences for moral hazard or social outcomes. The results of the 2016/17 welfare attitude survey more or less present the same picture as in 2008/9 (Ervasti et al. 2012; Roosma 2016; Svallfors 2012). In Table 9.2 the effect sizes of the year dummy are presented, for which the dependent variables are standardized and individual level control variables are taken into account. The same pattern of relatively limited welfare attitude change can be observed. For some welfare dimensions no significant change can even be observed. Although it could have been expected that austerity has led to stronger support for the government to protect incomes from market volatility and to more negative evaluations of the outcomes of social policy efforts, results instead show tendencies towards a better evaluation of the intended outcomes and less support for government involvement in the domain of income protection for the elderly and the unemployed. There is, however, somewhat more support for a stronger role for government in providing child care services. People also see somewhat less abuse of welfare benefits. In addition, people seem to be more satisfied with the state of health care and the standard of living of the unemployed. The strongest attitude change can be found for the item measuring peoples’ perceived impact of the welfare state on businesses. Substantially less people agree with the statement that social benefits and services cost businesses too much in terms of taxes and charges. This might be an effect of the impact of the crisis: people are more likely to reject the neoliberal discourse that social protection harms businesses.

7.49 (2.09)

0–10

Government responsibility to provide child care services for working parents

1–5 1–5

Abuse: Most unemployed people do not really try to find a job

Underuse: Many with very low incomes get less benefit than legally entitled to 1–5 1–5 0–10 0–10 0–10 1–5 1–5 1–5 1–5

Intended: Social benefits/services prevent widespread poverty

Intended: Social benefits/services lead to a more equal society

Intended: Perceived state of health care

Intended: Perceived standard of living of pensioners

Intended: Perceived standard of living of the unemployed

Unintended: Social benefits/services place too great strain on economy

Unintended: Social benefits/services cost businesses too much in taxes/charges

Unintended: Social benefits/services make people lazy

Unintended: Social benefits/services make people less willing to care for one another

Outcomes

1–5

Abuse: Many manage to obtain benefits/services to which they are not entitled

3.12 (1.04)

3.21 (1.11)

3.29 (1.03)

3.13 (1.04)

3.86 (1.92)

4.72 (2.14)

5.42 (2.38)

3.25 (0.99)

3.46 (0.96)

3.34 (0.96)

3.10 (1.09)

3.67 (0.94)

6.53 (2.21)

Implementation

8.17 (1.77)

0– 0 0–10

3.77 (1.04)

Government responsibility to provide decent standard of living for the unemployed

1–5

Mean (STD) 2008/9

Government responsibility to provide decent standard of living for the old

Range

Should the government reduce differences in incomes?

Goals

Scale

Operationalizations of welfare dimensions and their descriptive statistics (2008/9 and 2016/17)

Population and design weights applied according to ESS standards Welfare dimensions

Table 9.1

3.12 (1.07)

3.23 (1.12)

2.93 (1.07)

3.09 (1.06)

3.68 (1.97)

4.77 (2.21)

5.79 (2.38)

3.27 (0.99)

3.51 (0.95)

3.33 (0.95)

3.02 (1.10)

3.57 (0.98)

7.71 (2.08)

6.37 (2.20)

7.89 (1.83)

3.83 (1.02)

N = 64 704; N countries = 17 Mean (STD) 2016/17

The social legitimacy of European welfare states after “the age of austerity”  117

118  Handbook on austerity, populism and the welfare state

Figure 9.1

Government responsibility for the elderly

In Figures 9.1 to 9.5 the random slopes of the survey wave dummy per country are presented to show the effect of the year 2016/17, that is, change in welfare attitudes, for different welfare states who experience varieties of austerity. Models are controlled for individual level variables. The graphs show substantial differences per country. Figure 9.1 shows that in 15 out of 17 countries, support for a strong role of the government to provide a decent standard of living for the elderly has decreased in the past decade. There is only a small positive effect in Germany and no effect in the Czech Republic. Support has especially decreased in Hungary, Poland, Great Britain and Ireland. But also in the Nordic countries, Norway and Sweden. In some of these countries right-wing governments took office in the years after the crisis and advocated a more flexible government and budget cuts for expensive pensions and health care services (Farnsworth and Irving 2012), which could hint at an explanation for this trend in some of the countries. In Figure 9.2 we see more or less an opposite pattern in support for child care services. In 12 countries support for a role for government in providing child care services has increased, while in Hungary and Spain it has decreased and in Finland, the Netherlands and Sweden changes are rather small. Especially in Poland, Portugal, the Czech Republic and Germany support for the role of government in child care services has increased. Germany and the Czech Republic show in both substantive dimensions a direction towards more or stable government influence, while in Hungary in both cases support decreases. Trends in the other

The social legitimacy of European welfare states after “the age of austerity”  119

Figure 9.2

Government responsibility for child care

countries might be interpreted as higher support for social investment policies over benefits (van Kersbergen et al. 2014). Figure 9.3 shows strong differences across countries in evaluation of health care. Although in some countries there was not much change, we see strong change in attitudes in Spain, the Czech Republic, Germany, Norway and Portugal; people in Spain have become much more negative about the state of health care in their country, in the other aforementioned countries people have become much more positive. The result in Spain, as well as the negative effect in Great Britain and the Ireland, may be explained by harsh austerity measures that affected the (perceived) state of health care in these countries. The positive effect in Portugal, as a country experiencing a relatively deep crisis and also harsh austerity, is therefore more remarkable. In Figure 9.4, however, there is a negative effect in Portugal for the perceived standard of living of the unemployed. Also here Spain deviated from the general pattern and in Great Britain there is no change, which is in contrast with most other European countries where people have become more positive about the standard of living of the unemployed. In Figure 9.5 we see, except for Slovenia and Estonia, a clear decreasing trend in the perception that welfare benefits and services cost business too much in taxes and charges. Again Hungary stands out as a country with the strongest opinion change. Especially in countries affected by austerity – Spain, Great Britain and Ireland – the idea that welfare benefits cost businesses too much in taxes and charges has become smaller, together with Germany, Belgium and Norway.

120  Handbook on austerity, populism and the welfare state

Figure 9.3

Perceived state of health care

All in all, there are some indications that countries where there were harsh austerity responses (Spain, Portugal, Hungary, Great Britain and Ireland), either with or without (populist) right-wing parties coming into power (Hungary and Great Britain), stand out as having the sharpest opinion change. However, there are also unexplained deviations from this pattern. In order to analyse whether these trends can be explained by indicators of responses to or results from the crisis, in Tables 9.3 and 9.4 several multilevel models are presented. Firstly, in Table 9.3 the effects of the austerity regime dummies are estimated in the models for all five dimensions (Models 1–5A) and interaction effects with the survey year are included in Models 1–5B (a random slope for the survey year is included in the models with the interaction effects). All models are controlled for individual level variables (not presented) and the balanced countries are the reference category. Model 1A shows that countries where harsh austerity measures were implemented are more positive towards a stronger role of the government in providing a decent living standard for the elderly. The insignificant interaction effect in Model 1B shows that this does not change in 2016/17; these countries remain most positive towards a strong role of government and differences between the groups of countries do not change significantly. For the government responsibility in child care the differences between the groups of countries are not significant and different responses to austerity do not seem to have impacted attitudes in 2016/17. When the effects on the outcome dimensions are assessed, we see a different pattern. Countries experiencing harsh austerity are more critical about the state of health care and the standard of living of the unemployed than the two other groups of countries. And regarding the state of health care, people in harsh austerity regimes

The social legitimacy of European welfare states after “the age of austerity”  121

Figure 9.4

Perceived standard of living of unemployed

have become more negative in 2016/17. We note that Hungary, Poland and Estonia (within the group of countries experiencing harsh austerity) have become more positive about the standard of living of the unemployed, while within other countries in this group – Spain, Portugal and Slovenia – people have become more negative. These differences within the group seem to cancel each other out. Finally, in Model 5, we see that people in both mild and harsh austerity regime countries are more in agreement with the statement that the welfare state costs businesses too much in taxes and charges than people in balanced regimes. This might not be a surprise, as especially in the balanced regime countries – Germany, Sweden, Norway and Switzerland – the welfare state is strong and people are likely to believe that businesses are affected by this. Model 5B shows that differences between these groups of countries do not become smaller or bigger in 2016/17, as there is no significant interaction effect with the year dummy. The negative trend is equally strong across the different country groups. Finally, we look at how severe material deprivation has impacted welfare attitudes before and after the crisis. Table 9.4 shows the contextual level effect of severe material deprivation controlled for individual level variables in Models 6–10A. In Models 6–10B an interaction effect with the survey year was included as well as a random slope for the survey wave. Results show that severe material deprivation of a country has no effect on the support for a role of government for the elderly, while there is an interesting negative effect on support for providing child care. In countries with higher levels of material deprivation people might prefer income protection rather than (expensive) services. There is an expected negative effect of severe material deprivation on both the state of health care and the perceived standard of living of the unemployed. If there is a higher level of poverty, people are less satisfied with the

122  Handbook on austerity, populism and the welfare state

Figure 9.5

Welfare benefits cost businesses too much in taxes/charges

social protection and services system. In all these cases the interaction effect with the survey year (Model B) was not significant. The effect of severe material deprivation did not become stronger in 2016/17 and the crisis does not seem to have impacted the relation between poverty and welfare support here. Only for the dimension of the unintended outcome, the effects of benefits and services for businesses, is there a significant interaction effect. In countries with higher levels of severe material deprivation there is less agreement with the idea that welfare benefits and services will cost businesses too much. The negative interaction effect suggests that in 2016/17 there is even less support for this statement in more materially deprived countries. The experience of the crisis made people less convinced of the idea that businesses are harmed by welfare, especially in countries where deprivation is the highest.

DISCUSSION AND CONCLUSION In the history of the welfare state, the Great Recession has been one of the most impactful and elusive episodes. After the so-called golden and silver years of the welfare state in the previous century (Taylor-Gooby 2002), at the beginning of this century the welfare state was confronted with some major challenges following the financial, economic and Eurozone crisis and the policy responses captured under the phrase “austerity” (Taylor-Gooby et al. 2017). As countries were experiencing different levels of crises, had different welfare systems in place to mitigate the effects and adopted different policy strategies from harsh austerity to balanced

.009

Notes: *p