The Oxford Handbook of Job Quality 9780198749790, 0198749791

Citation preview

T H E OX F O R D H A N D B O O K O F

J OB QUA L I T Y

THE OXFORD HANDBOOK OF

JOB QUALITY Edited by

CHRIS WARHURST, CHRIS MATHIEU AND RACHEL E. DWYER

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Oxford University Press 2022 The moral rights of the authors have been asserted First Edition published in 2022 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2021952675 ISBN 978–​0–​19–​874979–​0 DOI: 10.1093/oxfordhb/9780198749790.001.0001 Printed and bound in the UK by TJ Books Limited Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Acknowledgements

Job quality is as important as ever, if not more so given the double assault on jobs from the Global Financial Crisis and the global Covid pandemic. But if this Handbook is timely, it has been a long time in the making. Pulling together and keeping together so many of the world’s leading authorities on job quality has been challenging at times. It has been a marathon, not a sprint. We would like to thank our commissioning editors at Oxford University Press, first David Musson and more recently Jenny King, for their support. Likewise our contributors, most of whom have been with us since the beginning—​we want to thank them for both their patience and efforts. Amanda Kerry, formerly of the Warwick Institute for Employment Research, supported our editing work and we are grateful to her for that work. Of course, as we laboured over the book, other responsibilities suffered. In this respect, we thank our families and friends for their understanding—​in particular, Doris and Hannah, and Katja, Ossian, Malaïka and Emilia, and David, Henry and Anna. Finally, the study of job quality is expanding, and we would like to thank all longstanding and new researchers to the field for their commitment to both wanting to understand and, often, wanting to improve job quality, the latter sometimes working with likeminded policymakers and practitioners. We hope that this Handbook is a useful tool for them and look forward to learning from their work in turn.

Contents

List of Figures  List of Tables  List of Contributors 

xi xv xvii

Job Quality Matters  Chris Warhurst, Chris Mathieu and Rachel E. Dwyer

1

PA RT I   T H E F O U N DAT ION S OF J OB Q UA L I T Y 1. The Quality of Working Life  David Guest

23

2. The Swedish Contribution to Job Quality  Ian Hampson and Åke Sandberg

41

3. Job Quality: A Family Affair?  Chris Warhurst, Sally Wright and Chris Mathieu

63

PA RT I I   U N DE R STA N DI N G J OB Q UA L I T Y 4. Understanding Differences and Trends in Job Quality: Perspectives from Cross-​National Research  Sven Hauff and Stefan Kirchner

87

5. Understanding Job Quality Using Qualitative Research  Angela Knox and Sally Wright

107

6. Quantitative Approaches to Assess Job Quality  Rafael Muñoz de Bustillo, Enrique Fernández-​Macías and José-​Ignacio Antón

126

7. Job Quality as the Realization of Democratic Ideals  John Godard

148

viii   CONTENTS

PA RT I I I   K E Y I S SU E S I N J OB Q UA L I T Y 8. Job Polarization: Its History, an Intuitive Framework and Some Empirical Evidence  Maarten Goos, Emilie Rademakers, Anna Salomons and Marieke Vandeweyer

169

9. Geographies of Job Quality  Sally Weller, Tom Barnes and Nicholas Kimberley

203

10. The Cornerstone of Job Quality: Occupational Safety and Health  Maria Albin, Chris Mathieu, Esa-​Pekka Takala and Töres Theorell

220

11. Innovation and Job Quality  Rafael Muñoz de Bustillo, Rafael Grande and Enrique Fernández-​Macías

244

12. Immigration and Job Quality  Amada Armenta and Shannon Gleeson

274

13. Job Quality for Service and Care Occupations: A Feminist Perspective  Orly Benjamin 14. Inequality in Job Quality: Class, Gender and Contract Type  Duncan Gallie

299 318

PA RT I V   R E G IONA L DE V E L OP M E N T S I N J OB QUA L I T Y 15. Job Quality in the United States and Canada  Arne L. Kalleberg, Sylvia Fuller and Ashley Pullman

339

16. The Great Recession and Job Quality Trends in Europe  Christine Erhel, Mathilde Guergoat-​Larivière, Janine Leschke and Andrew Watt

363

17. Job Quality in Emerging Economies through the Lens of the OECD Job Quality Framework  Sandrine Cazes, Paolo Falco and Bálint Menyhért

389

CONTENTS   ix

PA RT V   SE C TOR A L DE V E L OP M E N T S I N J OB QUA L I T Y 18. Job Quality in High-​Touch Services  Mary Gatta

415

19. The Changing Quality of Office Work  Chris Baldry

433

20. The Steady but Uneven Decline in Manufacturing Job Quality  Jeffrey S. Rothstein

452

21. Neoliberalism’s Impact on Public-​Sector Job Quality: The US and Germany Compared  Carsten Sauer, Peter Valet, Vincent J. Roscigno and George Wilson 22. Job Quality and the Small Firm  Paul Edwards and Monder Ram

469

485

PA RT V I   I M P ROV I N G J OB Q UA L I T Y 23. Human Resource Management and Job Quality  Peter Boxall and John Purcell 24. Using Efficiency, Equity and Voice for Defining Job Quality, and Legal Regulation for Achieving It  Stephen F. Befort, Silvia Borelli and John W. Budd

505

522

25. Trade Unions and Job Quality  Melanie Simms

542

Index 

559

List of Figures

3.1 Job quality categorization using subjective and objective approaches

66

4.1 Job quality as a relational concept

101

5.1 Job quality categorization

119

6.1 Average index of job satisfaction in 33 countries and territories, 2005

128

6.2 Average and differences in importance given to job characteristics

131

6.3 JQI and EJQI in the EU (15), 2010

143

8.1 Smoothed average 5-​year employment share changes by ICT-​capital intensity (in percentage points, 1980–​2005)

183

8.2 Smoother average 5-​year output price changes by ICT-​capital intensity (in percentage points, 1980–​2005)

186

8.3 Smoothed average 5-​year output volume changes by ICT-​capital intensity for manufacturing and all sectors (in percentage points 1980–​2005)

187

8.4 Smoothed average 5-​year employment share changes by ICT-​capital intensity for manufacturing and all sectors (in percentage points, 1980–​2055)

191

8.5 Share of employment in routine occupations over 1-​digit sectors, ranked by their ICT-​capital intensity

193

10.1 Interaction between the human and work

229

10.2 The work system Modified from Smith and Sainfort 1989

232

11.1 Relations between technological innovation and job quality

245

11.2 GDP per capita in Europe (EU-​12) and USA, 1870–​2010

246

11.3 Wages and productivity growth. OECD countries, 1970–​2006

248

11.4 Telephone and telegraph operators in England and Wales

250

11.5 Job quality by major sectors of economic activity; JQI 2005 and 2010, EU-​15

250

11.6 Probability of computerization and JQI in 39 sectors of economic activity

252

11.7 Five-​year averages of annual number of deaths related to coal mine explosions in the United States, 1901–​1995

254

xii   List of Figures 11.8 Summary Innovation Index and Technical Innovation Index from the EWCS in the EU-​27, 2010

261

11.9 Innovation: new processes or technologies introduced in current workplace during the last 3 years by Job Quality Index. EU15 countries, EWCS 2010

264

15.1 Trade union density in the United States and Canada, 1960–​2020

344

15.2 Employment rate in the US and Canada, total working-​age population 25–​54

346

15.3a Share of involuntary part time among part-​time workers in the US and Canada, men 1998–​2020

349

15.3b Share of involuntary part time among part-​time workers in the US and Canada, women 1998–​2020

349

15.4 Real hourly minimum wage, 1965–​2020

351

15.5a US occupational percentage shares, indexed to 2003

353

15.5b Canadian occupational percentage shares, indexed to 2003

353

16.1 Changes in sub-​indices of Job Quality Index between 2005 and 2010, EU27

368

16.2 Developments in overall Job Quality Index by country, 2005–​2010

370

16.3 Change in the unemployment rate between 2005 and 2010 and change in the overall Job Quality Index, 2005–​2010

371

16.4 Changes in unemployment rate and changes in share of people ‘not worried about losing job’, 2010–​2005

373

16.5a Box plot of level-​2 residuals (i.e. country-​specific effects): not employed vs job quality staying the same or improving

379

16.5b Box plot of level-​2 residuals (i.e. country-​specific effects): job quality deteriorating vs job quality staying the same or improving

379

17.1 Job quantity, job quality and well-​being

391

17.2 Earnings quality

398

17.3 Labour market insecurity due to unemployment

399

17.4 Labour market insecurity due to extreme low pay

401

17.5 Overall labour market insecurity

402

17.6 Incidence of very long hours

403

17.7 Job quality and quantity outcomes by sociodemographic groups

404

17.8 Job quality among formal and informal workers

406

21.1 The share of part-​time employees in the public sector for the US and Germany from 1985 to 2015

477

List of Figures    xiii 21.2 The share of exempt public-​sector workers for the US and the share of fixed-​term contracted workers in Germany from 1985 to 2015

478

21.3 The share of public-sector workers who work overtime for the US and Germany from 1985 to 2015

479

24.1 An aggregate scorecard for US job quality

532

24.2 An aggregate scorecard for job quality in the European Union

534

List of Tables

4.1 Dimensions of job quality captured in EWCS 2005 and ISSP data and their uses in different studies

90

4.2 Approaches to analysing cross-​national differences and trends in job quality 94 4.3 Index approaches based on the EWCS 2005 data

98

6.1 Dimensions of job quality suggested by the different traditions

133

6.2 List of acronyms, complete names, sources and databases of the indices of job quality reviewed

136

6.3 Summary of the main indicators of job quality

138

6.4 European Job Quality Index and Job Quality Indicator

142

7.1 Frege-​Godard measures of civic principles

154

8.1 ICT-​capital intensity, employment shares, output volume and output price by sector averaged across countries, 1980–​2005

180

8.2 Changes in employment shares by ICT-​capital intensity, 1980–​2005

182

8.3 Changes in output price by ICT-​capital intensity, 1980–​2005

186

8.4 Changes in output volume by ICT-​capital intensity, 1980–​2005

187

8.5 Changes in employment shares, output price and volume by ICT-​capital intensity, 1980–​2005

189

8.6 Changes in employment shares, output price and volume by ICT-​capital intensity, 1980–​2005

190

10.1 Ergonomics actions as a process

233

11.1 Index of job quality

258

11.2 Index of innovation. EU-​27, 2010

260

11.3 Pearson correlation coefficient among Job Quality Index and index of innovation by sector, EU(15), 2010

263

11.4 Determinants of job quality (linear regression), EU-​15, 2010

266

12.1 Job quality and immigration (select dynamics facing immigrant workers)

279

12.2 South Philly Survey descriptive statistics by gender

282

15.1 Labour market policies and public expenditures

343

15.2 Employment characteristics

347

xvi   List of Tables 15.3 Earnings quality and labour market insecurity in Canada and the United States

351

16.1 Most pronounced improvements and deteriorations in Job Quality Index (total and sub-​indices)—​2005–​2010

369

16.2 Results of the empty model

377

16.3 Role of individual characteristics on job quality change between 2007 and 2009

377

16.4 Results of the ‘best’ model

378

23.1 Job facet priorities of British workers

511

List of Contributors

Maria Albin MD PhD is Professor in Occupational and Environmental Medicine, Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden. She is former head of the Unit of Occupational Medicine and principal investigator for the research programme, ‘A sustainable new working life—​trends, health effects and governance’. Her research has focused mainly on chemical and physical occupational exposures and health but more recently more generally on sustainable working life, including the ageing workforce. She has served as an expert to, amongst others, the Swedish Criteria Group for Occupational Standards, the Swedish Work Environment Agency and the Swedish National Commission on Social Inequalities in Health. José-​ Ignacio Antón PhD is Associate Professor at the Department of Applied Economics of the University of Salamanca. His research interests include labour, health and public economics. He has been a consultant for the European Commission, the European Parliament, the International Labour Organization and several Spanish public administrations and agencies. Amada Armenta PhD is Associate Professor of Urban Planning at the UCLA Luskin School of Public Affairs. She is the author of Protect, Serve and Deport: The Rise of Policing as Immigration Enforcement (University of California, 2017). Her articles have appeared in Social Problems, Annual Review of Sociology, Annual Review of Law and Social Science and Sociology of Race and Ethnicity. Chris Baldry is Emeritus Professor attached to the University of Stirling Management School, Scotland. He has published in the areas of information technology and employment relations, occupational health and safety, and space and the built working environment. Past books include, with colleagues, The Meaning of Work in the New Economy (Palgrave, 2007). He is a former editor of the journal New Technology, Work and Employment. Tom Barnes PhD is an economic sociologist at the Institute for Religion, Politics and Society at Australian Catholic University in Melbourne. His current research project focuses on the demise of Australian automotive manufacturing. He has written several articles and chapters on insecure, precarious and informal work in Asia (especially India) and Australia, including his new book on Indian auto workers, Making Cars in the New India (Cambridge University Press, 2018).

xviii   list of Contributors Stephen F. Befort is the Gray, Plant, Mooty, Mooty, and Bennett Professor of Law at the University of Minnesota Law School. He teaches courses in Labour Law, Employment Law, Employment Discrimination, Disability in the Workplace and Alternative Dispute Resolution, and seminars in Comparative Labor and Employment Law and Advanced Topics in Labor and Employment Law. He has authored eight books and more than 50 articles on labour and employment subjects. He is the former Chair of the Labor Law Group, a member of the National Academy of Arbitrators, and a fellow of the American College of Labor and Employment Lawyers. Orly Benjamin PhD is Associate Professor at the Sociology Department at Bar-​Ilan University, Israel. She is the author of Gendering Israel’s Outsourcing: The Erasure of Employees’ Caring Skills (Palgrave, 2016) and co-​author of Feminism, Family and Identity in Israel: Women’s Marital Names (Palgrave, 2011). Her work on the topic of women and precarious employment and on mothers’ poverty employment has been published in Human Relations, Journal of Family Issues, Social Policy and Administration, The Sociological Review and Sociological Perspectives. She chairs the committee on the sociology curriculum for Israeli high-​schools and is an activist with the coalition for direct and fair employment. Silvia Borelli is Associate Professor in Labour Law at the University of Ferrara. She is a member of the editorial committee of Lavoro e Diritto and Rivista giuridica del lavoro (journals members of the International Association of Labour Law Journals). She participates in the Réseau académique sur la charte sociale européenne et les droits sociaux and collaborates with the European secretariat of the Italian General Confederation of Labour. During her career, she has taken part in several national and European research projects on gender equality, anti-​discrimination law, the quality of employment, networks of enterprises and transnational labour law. Peter Boxall PhD is a Professor in Human Resource Management and leader of the disciplinary area in HRM and Organisational Behaviour at the University of Auckland Business School. He is interested in strategic HRM and employee well-being. With John Purcell and Patrick Wright, he co-edited the Oxford Handbook of Human Resource Management (OUP, 2007) and with Richard Freeman and Peter Haynes, he co-edited What Workers Say: Employee Voice in the Anglo-American Workplace (Cornell University Press, 2007). He is the co-author with John Purcell of Strategy and Human Resource Management (Bloomsbury Academic), now in its fifth edition. John W. Budd holds the Industrial Relations Land Grant Chair at the University of Minnesota’s Carlson School of Management. His current research interests include employee voice, frames of reference, conflict resolution and other industrial relations topics. His books include Employment with a Human Face: Balancing Efficiency, Equity, and Voice; Invisible Hands, Invisible Objectives: Bringing Workplace Law and Public Policy Into Focus; The Thought of Work and Labor Relations: Striking a Balance. Professor Budd has served on the executive board of the Labor and Employment Relations Association,

list of Contributors    xix and has been Director of the Center for Human Resources and Labor Studies at the University of Minnesota. Sandrine Cazes is a Senior Economist at the OECD in the Directorate for Employment, Labour and Social Affairs. Prior to joining the OECD, she worked at the OFCE-​Sciences Po French observatory and at the International Labour Organization, both in the Budapest sub-​regional office and in the Geneva headquarters where she was the Head of the Employment Analysis and Research Unit. Her main topics of interest include comparative analysis of labour market policies and institutions, job quality, employment protection and labour relations. She has authored several publications in the fields of labour economics, flexicurity, labour market segmentation and collective bargaining, and is a regular contributor to the OECD Employment Outlook. Since September 2017, she is also co-​Chair of the Evaluation Committee of the ‘Ordonnances’ French Reforms. Rachel E. Dwyer PhD is Professor of Sociology and Faculty Affiliate of the Institute for Population Research at The Ohio State University. She studies the causes and consequences of rising economic inequality and insecurity in the US across several social arenas. She has contributed a series of institutional analyses of job polarization and economic restructuring in US labour markets, including as intersecting with racial and gender disparities in job quality. In related work, she analyses credit, debt and inequality, and the relationship between financialization and rising economic insecurity. Her published work on these issues has appeared in, for example, American Sociological Review, Social Forces, Annual Review of Sociology, Social Science Research, Gender & Society and Social Problems. Paul Edwards is Emeritus Professor of Employment Relations at Birmingham Business School in the UK, where he was Head of the Department of Management. He was previously at Warwick Business School, where he was Director of the Industrial Relations Research Unit as well as holding several Associate Dean positions. He is a Fellow of the British Academy and a former editor-​in-​chief of Human Relations. He served on two sub-​panels for the UK Research Excellence Framework. His research interests include workplace employment relations and critical realism, on which he continues to publish. Christine Erhel is a Professor in Economics at the Conservatoire National des Arts et Métiers (CNAM, Paris) and Director of the Employment and Labour Research Centre (Centre d’Etudes de l’Emploi et du Travail). She is a labour economist with a particular focus on researching comparative European labour markets, labour market policy reforms and job quality. In her research she has regularly participated in comparative projects, including a book published in 2016 on productivity trends in France, Germany, Spain and the UK. She has been regularly involved in EU-​funded research projects, such as the QuInnE project (http://​qui​nne.eu/​), analysing the relationships between innovation, job quality and employment, in which she coordinated the quantitative work package.

xx   list of Contributors Paolo Falco is a Labour Market Economist at the OECD in the Directorate for Employment, Labour and Social Affairs, where he contributed to the design of the OECD Job Quality Framework and is currently investigating how different megatrends will shape the future of work. He previously worked at the University of Oxford and collaborated with various international organizations, including the World Bank and the International Monetary Fund. His academic work uses experimental methods to understand how labour markets in developing countries can be made more efficient and inclusive. His research has been published in a number of academic journals and he is a regular contributor to the OECD Employment Outlook. He holds a PhD from the University of Oxford. Enrique Fernández-​Macías PhD is a researcher working in the Joint Research Centre of the European Commission. He previously worked in the Employment Unit of Eurofound and as a lecturer and researcher at the University of Salamanca. He has a PhD in Economic Sociology from the University of Salamanca, and his main research interests are in the areas of job quality, occupational change, technology and employment, and inequality. Sylvia Fuller is Associate Professor of Sociology at the University of British Columbia, and the Academic Director of the British Columbia Inter-​University Research Data Centre. Her research centres on understanding how labour market inequalities develop and erode and the implications of changing employment relations and social policy for people’s economic security and mobility. Recent publications explore how organizational context mediates parenting pay gaps, temporary workers’ employment trajectories, divergence in the career pathways of new immigrants and the medicalization of welfare among lone mothers. Duncan Gallie CBE FBA is an Emeritus Fellow of Nuffield College and Professor of Sociology in the University of Oxford. His research has involved comparative European studies of the quality of employment and of unemployment. Most recently, he has published on issues of inequality in work conditions, job insecurity and participation at work. He has advised the French government as a member of an expert group on psychosocial risks at work. He was a member of the advisory committee of a recent OECD initiative to provide guidelines to national governments for monitoring the quality of work. Mary Gatta PhD is the Director of Research and Policy at National Association of Colleges and Employers (NACE). She has published numerous books and articles including Waiting on Retirement: Aging and Economic Insecurity in Low Wage Work (2018) and All I Want Is a Job! Unemployed Women Navigating the Public Workforce System (2014). Shannon Gleeson is Professor of Labor Relations, Law, and History at the ILR School of Cornell University. Her books include Conflicting Commitments: The Politics of Enforcing Immigrant Worker Rights in San Jose and Houston (Cornell University Press, 2012) and Precarious Claims: The Promise and Failure of Workplace Protections in the United States (University of California Press, 2016).

list of Contributors    xxi John Godard is at the Asper School of Business, the University of Manitoba. His main interest is in the relationship between institutions and the attainment of democratic values at work. Although a sociologist by inclination, his work has appeared primarily in industrial relations journals, and from 2012 to 2017 he served as chief editor of the British Journal of Industrial Relations. Maarten Goos is Professor of Economics at Utrecht University. His research focuses on labour markets, including technological progress, labour market intermediation, inequality and institutions. He received his PhD from the London School of Economics (LSE) and held positions at Erasmus University Rotterdam, University College London and KU Leuven before joining Utrecht University. He also held visiting positions at Princeton University, the Centre for Economic Performance at LSE, the Massachusetts Institute of Technology and Boston University. He is currently coordinating the Future of Work initiative at Utrecht University and is involved in various initiatives by the European Commission, national policy institutes and multinational corporations. Rafael Grande PhD is a lecturer and researcher at the University of Málaga in Spain. He has a PhD in Social Science from the University of Salamanca, a Master’s degree in Latin-​American Economics Studies and a BA in Sociology. His fields of interest are sociology of labour, demography and international migration. He is currently working on the effects of digitalization and innovation on working conditions, the relationship between fertility and employment patterns, and the issues of an ageing population. Mathilde Guergoat-​Larivière PhD works as a Full Professor of Economics at University of Lille (Clersé) and is a research associate in Cnam-CEET in Paris. She has worked on the issue of job quality from a European comparative perspective for fifteen years and published a number of articles on that topic. Her research topics also include gender equality, labour market transitions and the impact of innovation on employment, job quality and inequalities. She has participated in various international projects over the last years for the European Commission, OECD and World Bank. David Guest is Emeritus Professor of Organizational Psychology and Human Resource Management at King’s Business School, King’s College, London. He has previously worked at Birkbeck, University of London, and at the London School of Economics. He is a previous editor of the Journal of Occupational and Organizational Psychology and the British Journal of Industrial Relations. His research addresses the link between HRM, well-​being and performance, the concept of the new career, the role of the psychological contract and developments in quality of working life. He has published numerous articles, and his recent books, with colleagues, are HRM and Performance: Achievements and Challenges (Wiley, 2013) and the Oxford Dictionary of Human Resource Management (Wiley, 2014). Ian Hampson is Honorary Professor at the Centre for Workplace Futures at Macquarie University, and an associate of the Industrial Relations Research Centre at the University of New South Wales, Australia. Ian gained a PhD in Politics, and Science and Technology Studies, from the University of Wollongong in 1994. He has published and taught in the

xxii   list of Contributors fields of politics, industrial relations, sociology of work, training policy and skill recognition. He was lead author of the Australian Research Council funded report The Future of Aircraft Maintenance in Australia, and is a member of the editorial committee of the Economic and Labor Relations Review. Sven Hauff PhD is Professor of Human Resource Management and holds the Chair of Labor, Human Resources and Organization at the Helmut Schmidt University in Hamburg. His main research interests are in the interrelationships between the design of human resource management and its effects on employees and organizational performance. Most recently, he has worked on HRM systems, job quality and job satisfaction, as well as the influence of national institutions and culture. He has published in journals such as Human Resource Management, Human Resource Management Journal, International Journal of Human Resource Management and International Business Review. Arne L. Kalleberg is a Kenan Distinguished Professor of Sociology at the University of North Carolina at Chapel Hill. He has published extensively on topics related to the sociology of work, organizations, occupations and industries, labour markets, and social stratification. His recent books are Good Jobs, Bad Jobs (2013) and Precarious Lives (2018). He is currently editing (with David Howell) an issue of the RSF of the Social Science on ‘Changing Job Quality’. He served as the President of the American Sociological Association in 2007–​2008 and is currently the editor of Social Forces, an international journal of social research. Nicholas Kimberley PhD is a candidate in economic geography at the University of South Australia in Adelaide, Australia. His current research explores the labour market changes following major job losses in regional economies. Prior to studying at ACU, he completed a Bachelor of Arts (Honours) in the School of Geography at the University of Melbourne. Stefan Kirchner PhD is Professor in the Sociology of Working Worlds’ Digitalization at the Technische Universität Berlin. His research revolves around changes in work, organization and the economy. Currently, he focuse on how digital technology affects job quality and enables work on digital marketplaces (e.g. crowdsourcing and gig economy). He has published his work in journals such as the Socio-​Economic Review and Zeitschrift für Soziologie. Angela Knox PhD is an Associate Professor of Work and Organisational Studies at the University of Sydney Business School. Her research focuses on job quality, insecure and precarious work and employment regulation. She has co-​edited (with Chris Warhurst) a book on job quality in Australia and published in various leading international journals, including Gender, Work and Organisation, the Human Resource Management Journal, the International Journal of Human Resource Management and Work, Employment and Society. Janine Leschke PhD is Professor in Comparative Labour Market Analysis at the Department of Management, Society and Communication at Copenhagen Business School (CBS). Her main research area is comparative European labour market and

list of Contributors    xxiii welfare state analysis. She is particularly interested in the interface between labour market flexibility and security. She has been working extensively on non-​standard employment, EU cross-​border labour mobility and job quality as well as gender and labour markets. She is currently one of the Danish lead partners in the EU project Disruptive Technologies Supporting Labour Market Decision Making (HECAT). She is one of the editors of the Journal of European Social Policy. Chris Mathieu PhD is Reader in the Sociology of Work and Organisation at the Department of Sociology, Lund University. From 2015 to 2018 he coordinated the Horizon 2020 project QuInnE—​Quality of Jobs and Innovation Generated Employment Outcomes (quinne.eu), and currently works on the Horizon 2020 project Beyond 4.0. (beyond4-​0.eu). His primary research is on innovation, work, employment, organization, and careers in the arts/​creative industries and hospitals. He currently co-​edits the Sociology and Management of the Arts book series on Routledge. Bálint Menyhért is a Research Fellow at the Finance and Economy Unit of the European Commission’s Joint Research Centre in Ispra, Italy. His areas of expertise include economic history, labour markets and poverty measurement. He obtained his PhD in Economics from the Central European University in 2017. He has previous work experience from the Federal Reserve Banks of Boston and New York, the Hungarian Financial Supervisory Authority and the OECD. Rafael Muñoz de Bustillo PhD is Professor of Applied Economics at the University of Salamanca. He has published a large number of books and articles on labour issues and the economics of the welfare state, among them Measuring More than Money: The Socio-​Economics of Job Quality (Edward Elgar, 2011) in collaboration with E. Fernandez-​ Macías and J. I. Antón. He has been consultant for the European Commission, the European Parliament, the International Labour Organization and several Spanish public administrations and agencies. Ashley Pullman is a Postdoctoral Fellow at the Education Policy Research Initiative (EPRI) in the Graduate School of Public and International Affairs at the University of Ottawa, Canada. Ashley completed her PhD at the University of British Columbia in 2017 in education with a sub-​specialization in Measurement, Evaluation, and Research Methodology. Her current research focuses on the interplay among education, skill and work, with a particular focus on socio-demographic inequality. John Purcell DLitt held Chair appointments at Warwick and Bath universities with earlier appointments at Oxford and Manchester universities. His main book publica­ tions include Human Resource Management in the Multi-​Divisional Company (OUP, 1994), the Oxford Handbook of HRM (OUP, 2007) edited with Peter Boxall and Patrick Wright, and Strategy and Human Resource Management (written with Peter Boxall) (Bloomsbury Academic 2022). His career has been devoted to the study of the management of labour including the links with business performance, the role of front line managers, contingent workers and employee voice systems: see Consultation at Work: Regulation and Practice with Mark Hall (OUP 2012).

xxiv   list of Contributors Emilie Rademakers is an Assistant Professor at the Utrecht University School of Economics. She obtained her PhD in economics from the University of Leuven in 2019. Her research focuses on labour markets, including the consequences of technological change and gender inequalities in labour market outcomes. Her research has been published in international peer-​reviewed journals and she is a frequent speaker at international academic conferences and policy institutes. Monder Ram is Professor and Director of the Centre for Research in Ethnic Minority Entrepreneurship at Aston Business School in the UK. He publishes widely on small business and ethnic minority entrepreneurship, employment relations and small firm policy. Much of his work involves collaborations with academics, practitioners and communities from diverse backgrounds. Vincent J. Roscigno PhD is Distinguished Professor of Arts & Sciences in Sociology at Ohio State University. His research foci and interests include social inequality and mobility, relational dynamics in the workplace, sociology of education, and politics and social movements. Some of his work has been published in American Sociological Review, Human Relations, Academy of Management Review, American Journal of Sociology, Work & Occupations and Work, Employment & Society. Jeffrey S. Rothstein PhD is Professor of Sociology at Grand Valley State University in Allendale, Michigan. He is the author of When Good Jobs Go Bad: Globalization, De-​Unionization, and Declining Job Quality in the North American Auto Industry (Rutgers University Press, 2016). His research on the impact of globalization on labour relations, job quality and prospects for economic development can also be found in Social Forces, Critical Sociology, Competition & Change, Research in the Sociology of Work and New Labor Forum. Anna Salomons is Professor of Economics at Utrecht University. She obtained her PhD in economics from the University of Leuven in 2012 and held visiting positions at the Centre for Economic Performance at LSE, the Massachusetts Institute of Technology and Boston University. Her research focuses on labour markets, including the consequences of technological change, and on applied micro-​econometrics. Her research has been published in international peer-​reviewed journals and she is a frequent speaker at international academic conferences and policy institutes. Åke Sandberg holds an MBA from the Gothenburg School of Economics and a PhD in sociology from Uppsala. He is Emeritus Professor in sociology at Stockholm University. Earlier he worked at KTH, The Royal Institute of Technology and at Arbetslivsinstitutet (the National Institute for Working Life). He has published Nordic Lights: Work, Management and Welfare in Scandinavia (SNS, 2013) and På jakt efter framtidens arbete (Tiden, 2016), on work–life research policy (in Swedish with English abstracts). An earlier book was Enriching Production: Perspectives on Volvo’s Uddevalla Plant as an Alternative to Lean Production. The two latter books may be

list of Contributors    xxv downloaded at the homepage, akesandberg.se. A recent volume is Arbete & välfärd (Studentlitteratur 2019) in which 27 researchers contribute chapters on Swedish working life and welfare. Carsten Sauer PhD is Professor of Sociology at Bielefeld University, Germany. He is interested in labor markets and organizations, social inequalities, justice perceptions, and survey experiments. Recent publications have appeared in the American Sociological Review, the American Journal of Sociology, and Social Forces. Melanie Simms PhD is Professor of Work and Employment at the University of Glasgow in Scotland. Her research focuses on trade unions, the future of work and young people’s transitions into employment. Her latest book is What Do We Know and What Should We Do About the Future of Work? (Sage, 2019). Esa-​Pekka Takala MD PhD is Docent in Ergonomics at the University of Eastern Finland and Physiatricts at the University of Helsinki. He has studied relationships between work and musculoskeletal disorders since the 1980s at the Finnish Institute of Occupational Health (FIOH). The approach of his research has varied from work physiology and biomechanics to epidemiology, clinical trials and rehabilitation. The number of his scientific publications covers over 100 original peer-​reviewed reports and 25 reviews or chapters in textbooks. Töres Theorell became a licensed physician in 1967 and completed his doctoral dissertation at the Karolinska Institute in 1971. He worked clinically in internal medicine, cardiology, primary care and occupational medicine until 1986, but became more involved in research work. He became Professor at the National Institute for Psychosocial Factors and Health in 1981 and Professor at the Karolinska Institute in 1995 and at the same time Director of the National Institute for Psychosocial Factors and Health. After retirement in 2006 he has remained active as a researcher. His focus is physiological stress mechanisms, occupational health epidemiology and interventions. Peter Valet PhD is an Assistant Professor and Chair of Sociology at the University of Bamberg in Germany. His main research interests are in the areas of social inequality, distributive justice, and empirical methods. His recent work has been published in the American Journal of Sociology, Work & Occupations, and Social Psychology Quarterly. Marieke Vandeweyer is a labour market economist in the Employment, Labour and Social Affairs Directorate of the OECD. She has authored several OECD reports in the areas of skills, activation policies and labour market reforms. Marieke holds a PhD in Economics from the University of Leuven in Belgium and has published a range of articles related to employment, skills and wage inequality. Chris Warhurst PhD FRSA is Professor and Director of the Warwick Institute for Employment Research at the University of Warwick in the UK and a Research Associate of SKOPE at Oxford University. With colleagues, he has published a number of books and articles on job quality, including, Are Bad Jobs Inevitable? (Palgrave, 2012) and Job

xxvi   list of Contributors Quality in Australia (Federation Press, 2015). He has been expert advisor on job quality to Oxfam Scotland and the Carnegie Trust, and on skills policy to the UK, Scottish and Australian Governments and an International Expert Adviser to the OECD’s LEED programme. He was also a member of the Measuring Job Quality Working Group established to develop recommendations on Good Work from the UK Government’s Taylor Review of Modern Working Practices. Andrew Watt is Head of the European Economic Policy Unit and Deputy Head at the Macroeconomic Policy Institute (IMK), part of the Hans-​Böckler Foundation. His main fields are European economic and employment policy and comparative political economy, with a particular interest in the interaction between wage-​setting and macroeconomic policy. Recent work has focused on reform of the economic governance of the euro area, emphasizing the need to coordinate monetary, fiscal and wage policy in order to achieve balanced growth and favourable employment outcomes. He has served as adviser to numerous European and national institutions, including the European Commission and Eurofound. Regular commentary can be found on https://​twit​ter. com/​Andre​wWat​tEU and http://​andrewwatt.eu/​ Sally Weller PhD is an economic geographer and Associate Professor at the University of South Australia in Adelaide, Australia. Her work on labour market change, regional development and industrial transformation in Australia. She is a frequent contributor to public policy and has conducted consultancy projects for the Australian Government, Victorian State government and non-​profit organizations. George Wilson PhD is Professor of Sociology and department Chair at the University of Miami. His research interests include social stratification and mobility, public-​sector work, minority vulnerability, and social class, racial/​ethnic and gender inequalities in mobility pathways. His work has appeared in journals such as Social Forces, Social Problems, Research in the Sociology of Work, Work & Occupations, and the DuBois Review. Sally Wright PhD is a Senior Research Fellow at the Warwick Institute for Employment Research (IER) at the University of Warwick in the UK. Since joining IER, she has undertaken research on job quality for sponsors including the European Commission, Eurofound, Oxfam UK and UK CIPD. She is currently involved in several European-​ funded research projects on the impact of technology on work, convergence in job quality in the EU, and measuring outcomes from active labour market policies for the long-​term unemployed. Her doctoral thesis investigated job quality in Australia.

J OB QUALIT Y MAT T E RS chris warhurst, chris mathieu and rachel e. dwyer

The aim of this Handbook of Job Quality is to provide an interdisciplinary and international benchmark text for anyone wanting to understand job quality. Job quality matters and has long and continually done so, even if the terminology used to describe job quality has, and continues, to vary. For example, struggling in many cases to physically subsist, British workers in the first industrial revolution of the eighteenth century agitated against the poor working conditions of the new factory system (Hall 2009). In the mid-​twentieth century, expressed as the Quality of Working Life, job quality was held up as the route to improved organizational productivity and efficiency and what would now be termed ‘employee well-​being’ (e.g. Trist and Bamforth 1951; also the chapter in this Handbook titled ‘The Quality of Working Life’ by Guest). By the late twentieth century trends in job quality became a concern. Often contrasting even competing, grand claims of historical inevitability of distinct trajectories for job quality were made. Some scholars predicted that job quality was inevitably improving as the nature of production and what was being produced changed (e.g. Bell 1973). Others argued that job quality would degrade due to the competitive nature of capitalism (e.g. Braverman 1974). Job quality still matters. Understanding trends continues to be important, as the debate about a polarization in job quality exemplifies (see the chapter in this Handbook titled ‘Job Polarization: Its History, an Intuitive Framework and Some Empirical Evidence’ by Goos and colleagues and the chapter titled ‘Job Quality in the United States and Canada’ by Kalleberg and colleagues). Debate about the future of work and job quality in the twenty-​first century centres on the impact of the new digital technologies of the putative fourth industrial revolution (Schwab 2016). Whilst most of the debate about the impact of these technologies has foregrounded job loss, another strand has attempted to focus on the quality of the remaining jobs, with, on the one hand, fears that it might be little more than machine minding or, alternatively, hopes that it will be more complex and socially interactive (Dwyer and Wright 2019; Warhurst and Hunt 2019). This debate about the coming of the clever robots compounds existing concerns about the restructuring of employment (e.g. Fernández-​Macías et al. 2010), the transformation of work (Doogan, 2009) and, importantly, a worrying proliferation of poor-​ quality jobs (e.g. Gautié and Schmitt 2010) often within the context of neo-​liberal political-​economic hegemony since the early 1980s or the economic crisis that followed

2    chris warhurst, chris mathieu and rachel e. dwyer the Global Financial Crisis (GFC) of the late 2000s (see the chapter titled ‘The Great Recession and Job Quality Trends in Europe’ by Erhel and colleagues). Job quality is thus the focus of significant international and multi-​disciplinary academic interest and debate, much of which has been shaped by the contributors to this volume. It is also increasingly a concern of policymakers. Governments and politicians around the world are now not just taking an interest in job quality but actively advocating its improvement, as the 2015 Ankara Declaration signed by the G20 countries illustrates.1 Underneath this growing consensus about its importance lurks a number of significant, and in some cases unresolved, issues about job quality. If job quality matters, there are matters to do with job quality that need to be explored and understood. This introductory chapter sets out these matters and the ways that they are explored in the Handbook.

Why Job Quality An obvious starting point is the question, why job quality? This simple question can be interpreted in at least two ways. The first is why job quality is important. The answers to this first interpretation of the question comprise the fundamental rationale and justification for this volume. The central importance of job quality to a range of social, economic and political challenges is what warrants devoting a Handbook to the topic. Job quality is offered as a solution to challenges such as health, welfare, productivity, innovation, economic competitiveness, democracy and democratic participation, Bildung/​cultivation, societal equality, individual and collective quality of life and environmental sustainability. Along with poverty eradication and environmental sustainability, the International Labour Organization (ILO) (2015) even claims that decent work, as an expression of job quality, is one of ‘the three defining challenges of the twenty-​first century’ (p.4). As job quality is a key factor in addressing these and the other challenges, it needs to be understood in all its complexity in terms of what it affects as well as what affects it. This Handbook draws together into a single volume: first, an explicit focus on job quality both as a significant factor in and of itself and as producing instrumental effects on a range of other processes and outcomes; second, a catalogue of the diverse range of multiple contributions and applications related to job quality; and third, the complexity and multiple interpretations of the concept of job quality. Each chapter provides distinct responses to the question of why job quality matters, coupled to a contention about for whom or for what job quality matters most. As the chapters with their respective answers and arguments attest, there are a range of ways in which job quality is relevant to an equally broad range of social, economic and political concerns.

1 

www.dol.gov/​ilab/​media/​pdf/​2015-​G20-​Ministerial-​Declaration.pdf/​.

JOB QUALITY MATTERS    3 The second interpretation centres on why the term or concept of job quality is now being used or preferred over other possible terms or concepts in framing and discussing solutions to social, economic and political challenges. The obvious alternative is skills. Over the last quarter century, education and training policy has shifted from being a secondary concern for government to a primary concern, and regarded as the solution to a range of social and economic problems (Keep and Mayhew 2010). This shift, Keep and Mayhew note, was epitomized in the UK Government’s Leitch Review of Skills which stated that ‘where skills were once a key driver of prosperity and fairness, they are now the key driver’ (Leitch 2006: 9).2 Unfortunately boosting skills supply, typically by increasing rates of higher education participation amongst young people, has failed to significantly shift the dial on these problems. Part of the reason is that skill deployment is as important as skill development (Livingstone 2017). Wider workplace issues within which skills are embedded have also been marginalized in debate. Keep and Mayhew argue that the shift to skills ignores key issues that need to be addressed: work organization and job design; employee ‘voice’, wage setting and income distribution and progression opportunities—​all of which feature as part of the dimensions of job quality in contributions to this volume. Thus, skills remain important. However, skills alone are insufficient. Instead, what is needed, Keep and Mayhew insist, are ‘improvements to the quality of working life’ (p.574). If job quality really is the answer, the obvious question is to ask why it took so long to be recognized as such by policymakers. One explanation forwarded by Guest in this volume is that the wider political and economic tilt to neo-​liberalism from the 1970s pushed job quality off the policy agenda. Another explanation is that job quality has been overshadowed by industrial relations. Whilst interest in some aspects of job quality existed throughout the mid-​to-​late twentieth century, as Bell (1973) and Braverman (1974) illustrate, during this period those concerns were ‘masked and subsumed’ by the policy focus on industrial relations or, more specifically the ‘labour problem’ that was said to hamper productivity and stifle organizational efficiency (Knox et al. 2011: 8). The primary concern of industrial relations was trade union and management negotiations over the terms and conditions of employment, employee voice and pay, and with variation in these negotiations depending on whether they occur at workplace, enterprise, sector and national levels (Bamber et al. 2016). Industrial relations became synonymous with collective action but also associated with confrontation, strikes and disruption (Bray et al. 2009; Kochan 1998). Hence, within the shift to neo-​liberalism came the undermining of trade unions and collectivism and the assertion of management’s right to manage. By the 2010s, with state-​led dismantling of industrial relations, the labour problem had been fixed many governments believed, at least in the liberal market economies. In these economies, human resource management (HRM) has emerged, displacing industrial relations. Within the enterprise-​ level focus on HRM, collectivism is

2 

Emphasis in the original.

4    chris warhurst, chris mathieu and rachel e. dwyer replaced with individualism, and the concern is to align organizations’ needs with the needs of individual employees or, in its harder US version, the ‘auditing’ of human resources to better match organizational supply and demand (Legge 1995). The dismantling of industrial relations might have enabled broader concerns with job quality to re-​emerge. However, it is telling that in the displacement of industrial relations by HRM there has again been little interest in job quality (Boxall and Purcell in the chapter titled ‘Human Resource Management and Job Quality’ in this volume). Nonetheless, Boxall and Purcell argue that more recent HRM research is increasingly focused on what they call the ‘double agenda’ of organizational performance and employee well-​being. Significantly, employee well-​being is often argued to be the function of job quality (e.g. Muñoz de Bustillo and colleagues in the chapter on ‘Quantitative Approaches to Assessing Jobs’ in this volume). The outcome of this recent HRM research is a focus on helping policymakers and managers understand the factors that contribute to more productive and more fulfilling employment relationships. It is this outcome that makes Boxall and Purcell hopeful that job quality will come to sit at the heart of HRM research. What these debates highlight is that both alternatives to job quality—​skills and industrial relations/​HRM—​maintain their importance as fields of research and practice but debate in both is now being peeled back to reveal the underlying importance of job quality. There is a logic to this development given that skills and employee voice are often argued to be aspects of job quality as various contributions to this volume highlight. But it is also an outcome of the recognition that single-​factor solutions to social, economic and political challenges have limited efficacy and that ‘bundles’ of workplace practices are needed to meet those challenges, as research on High Performance Work Systems illustrates (Combs et al. 2006). In this context, job quality,-especially when understood as a multi-​dimensional concept, is an obvious and useful focal point for both research and policy.

What is Job Quality? Conceptualization of job quality extends into another prominent issue in the volume—​ questions about the measurement of job quality. Research and policy formulation of job quality rely upon defining it and specifying its components in order to measure it. Despite emerging consensus that job quality is best understood as a multi-​dimensional concept, what is striking about past and current foci on job quality is a lack of common agreement about those dimensions and therefore lack of agreement on what constitutes job quality. As a term, job quality has high recognition but, as this Handbook illustrates, little consensus about what it comprises. Muñoz de Bustillo et al. (2011: 4) refer to job quality as ‘elusive’ because ‘it is one of those concepts . . . which everyone understands yet it is difficult to define precisely’. Tellingly, in 2015 the Group of 20 (G20) declared a

JOB QUALITY MATTERS    5 commitment to improving job quality but without defining it or advancing a means of measuring improvements to it.3 Reflecting the conceptual heterogeneity, numerous lists exist of what is said to comprise job quality (Wright 2015). Some of these lists’ contents overlap, some are distinct, some extend beyond the job to include labour market conditions. Whilst these conditions are important and deserve attention, for example the use of child labour and the persistence of gender inequality (see ILO 1999; Ghai 2003), they do not directly constitute the ‘properties’ of the job. This variety of indicators and the lack of conceptual consensus is vividly reflected in this Handbook, and intentionally so if the field is to be represented. Some chapters highlight one particular dimension of job quality, other chapters offer novel combinations of multiple dimensions, some chapters draw on broad and others a narrow combination, and some include subjective, others objective measures. Despite Albin and colleagues in the chapter titled ‘The Cornerstone of Job Quality: Occupational Safety and Health’ in this volume arguing that physical and mental health are foundational to other dimensions of job quality, contributions to the Handbook typically focus on indicators related to work and employment, and sometimes both. The distinction between work and employment is important in discussing job quality, despite the use of the term ‘work’ as a shorthand for ‘jobs’ in much US research (Warhurst and Knox 2022). In terms of ‘work’, Sauer and colleagues in the chapter titled ‘Neoliberalism’s Impact on Public Sector Job Quality: The US and Germany Compared’, for example, analyze work intensification and enhanced managerial control in the analysis of changing job quality amongst white-​collar office workers. Job control is also used as a marker in Gallie’s discussion in the chapter on ‘Inequality in Job Quality: Class, Gender, and Contract Type, along with opportunities for skill development and earning levels. The high rate of employee disposability—​in other words, the existence of high levels of job insecurity—​features strongly in Gatta’s discussion in this volume of ‘Job Quality in High Touch Services’ and reflects the concern with ‘employment’ as a key marker of job quality, in this case the use and impact of certain labour contracts. Building on pluralist industrial relations theory with its emphasis on the terms and conditions of employment again, Befort and colleagues in the chapter titled ‘Using Efficiency, Equity and Voice for Defining Job Quality, and Legal Regulation for Achieving It’ elevate three dimensions as central to job quality—​efficiency, equity and voice—​that they posit are mutually interdependent and need to be seen and pursued as a whole. This argument reinforces the point made earlier that job quality is likely best conceived and researched as an interactive bundle of practices. Weighing in on the conceptual heterogeneity, in 2015 the United Nations’ Economic Commission for Europe issued its Handbook on Measuring Quality of Employment: A Statistical Framework (UNECE 2015), making refinements in 2019. It offers seven dimensions of job quality, one of which, reflecting the priorities of its ILO

3 

www.dol.gov/​ilab/​media/​pdf/​2015-​G20-​Ministerial-​Declaration.pdf/​.

6    chris warhurst, chris mathieu and rachel e. dwyer agency, is portrayed as an ethical labour market issue—​the use of child and forced labour. These seven dimensions are: Safety and ethics of employment; Income and benefits from employment; Working time and work–​life balance; Security of employment and social protection; Social Dialogue; Skills development and training; and Employment-​ related relationships and work motivation. About the same time, the OECD developed its Job Quality Framework (Cazes et al. 2015) highlighting three dimensions: earnings quality, labour market insecurity, and quality of the working environment. Cazes and colleagues’ contribution in the chapter titled ‘Job Quality in Emerging Economies through the Lens of the OECD Job Quality Framework’ offers an application of this framework to job quality in emerging economies and explains the framework in greater detail. Trying to cut through the conceptual heterogeneity, some researchers argue that job quality can be best measured through just one dimension—​pay. Indeed, pay is an oft-​ used marker of job quality. The reason is pragmatic according to Osterman and Shulman (2011): typically, good data exists, and that data is often longitudinal and is often comparable across occupations, industries and countries. Even research that extends beyond pay, to skill for example, does so suggesting that skill can be a proxy for pay (e.g. Goos and Manning 2007). In this volume Goos and colleagues use the same combination to assess the nature and extent of job polarization across the US and European economies. This continuing diversity of concepts of job quality matters. It undermines common understanding by researchers and constrains action by government. There can be a number of reasons why this heterogeneity exists. First, for Hurley et al. (2012), it reflects the different disciplinary traditions that underpin research into job quality. Second, it can reflect different methodological approaches. The chapter by Muñoz de Bustillo and colleagues, ‘Quantitative Approaches to Assessing Jobs’ provides a review of the different quantitative approaches to measuring job quality and, with it, more than twenty proposals of indicators of job quality. Knox and Wright in the chapter titled ‘Understanding Job Quality Using Qualitative Research’ focus on measures within qualitative research and they point out that quantitative and qualitative research methods tend to be utilized for different, though often complementary, purposes. Third, what is offered as a marker of job quality can simply be a sign of the times. In this respect, the conception of job quality during the period in Sweden examined by Hampson and Sandberg in the chapter on ‘The Swedish Contribution to Job Quality’ should be seen against the backdrop of struggles over control of the firm predicated on ideas about democratization of working life current at the time. By the standards of the early twenty-​first century, this co-​determination of production would appear ambitious, almost off the scale in terms of what today’s advocates of improved job quality are seeking. Similarly, following the Covid pandemic and extensive working from home, opportunity for hybrid working might become an indicator of job quality in the future. Fourth, and an extension perhaps of the last point, it might be that context determines particular conceptualizations and measurement of job quality. In the chapter by Hauff and Kirchner, ‘Understanding Differences and Trends in Job Quality: Perspectives from Cross-​National Research’, they argue that job quality

JOB QUALITY MATTERS    7 arises from and should be understood in context. Their review of cross-​national research of job quality leads them to identify a conceptual incommensurability at the heart of this research that makes it difficult to develop a common understanding of job quality. As a consequence, job quality should be recognized and accepted as a relational concept they conclude. Relatedly, Godard in his chapter on Job Quality as the Realization of Democratic Ideals’ argues that job quality should be regarded as an expression of how societies value and treat labour. As a consequence, he argues that it is more fruitful to analyse job quality at the societal rather than individual or organizational level regardless of methodological approach. A key challenge here will be to include job quality for migrants as well as the role of job quality in both push and pull factors driving emigration and immigration, issues considered by Armenta and Gleeson in their chapter, ‘Immigration and Job Quality.’ Fifth, rather than paradigms, politics or place, it can be pragmatism that drives the use of particular markers of job quality. Here data availability is the key, so what gets collected gets counted. For example, amongst European researchers of job quality a key dataset is the European Working Conditions Survey administered by the European Commission. There are a number of weaknesses with this dataset (Warhurst et al. 2018) but the information collected shapes much analysis of what is construed as job quality (e.g. Holman 2013). What information this survey does not collect is not factored into these constructs of job quality. In other words, data availability can drive the conceptualization and subsequent measurement of job quality. As we noted earlier, the key reason for Osterman and Shulman’s (2011) advocacy of pay as the measure of job quality is that good pay data is readily available. Sometimes the weakness of this approach is recognized and legitimized by suggestions that the set of measures being offered are a ‘short form’ of job quality (e.g. Felstead et al. 2019). Highlighting this heterogeneity in understanding job quality is important but needs to be resolved, Muñoz de Bustillo and colleagues contend in their chapter, ‘Quantitative Approaches to Assessing Jobs’, and efforts made to develop a unifying, single set of measures. This task is discussed in this volume by Warhurst and colleagues in the chapter titled ‘Job Quality: A Family Affair’. The authors’ premise is twofold: first, agreeing with Muñoz de Bustillo et al. (2009: 25), that there should be a ‘purging’ of dimensions and indicators not directly related to the job and, second, that a double consensus is required: first amongst the research community, second with non-​academic stakeholders. Existing research indicates that there are several ways in which indicators and measures of job quality are developed. First, as we have already noted, researchers can determine what is to be included based on data availability. Second, workers can be asked to self-​identify aspects of jobs that cause dis/​satisfaction, with researchers then extrapolating the main items. Third, workers can be asked what researcher-​identified features make a ‘good job’ with items then ranked by researchers. Fourth, stakeholders, such as trade unions, can be asked what they believe is important in jobs and these views accepted. Fifth, periodic literature reviews can be undertaken of existing international, multidisciplinary research to identify common or recurring foci along with new emerging emphases.

8    chris warhurst, chris mathieu and rachel e. dwyer Building on the earlier work of other European researchers (Muñoz de Bustillo et al. 2011; Davoine et al. 2008), Warhurst and colleagues adopt this last approach in order to generate consensus amongst the academic community. The six dimensions that emerge from the review are: Pay and other rewards; Intrinsic characteristics of work; Terms of employment; Health and safety; Work–​life balance; and Representation and voice. These dimensions were later road-​tested with key stakeholders in the UK drawn from employer, trade union and civil society organizations, plus government statisticians and officials. One dimension, Intrinsic characteristics of work, was subsequently split to pull out Social support and cohesion and then renamed ‘Job design and nature of work’, resulting in seven dimensions (Measuring Job Quality Working Group 2018). These dimensions can provide a common basis for measuring conceptual variations of job quality (see Zemanik 2020) and also provide the basis for developing a set of minimum standards of job quality that might improve work and employment in the future (Warhurst and Knox 2022).

The Future of Job Quality As with past economic crises, the Covid pandemic badly impacted national economies and led in some to rising unemployment. Falling back on the creation of any jobs rather than good jobs is always a temptation for governments in such situations. However, since the GFC of 2007–​08, evidence-​based understanding of the relationship between job quality and job quantity has improved, and policy thinking shifted accordingly. Many OECD governments now regard job quality as offering a route to national economic growth and competitiveness, as the G20 Ankara Agreement shows. It is also increasingly advocated for the emerging economies, as Cazes and colleagues demonstrate in their chapter in this volume. Indeed, expressions of job quality such as ‘good work’, ‘decent work’ and ‘fair work’ (for a discussion, see the chapter by Warhurst and his colleagues) are now mainstreamed in government policy. In the past it was often argued that a trade-​off exists between job quantity and job quality. Countries could not have both, particularly in times of economic crisis. This argument resulted in a ‘jobs first’ policy over the 1980s and 1990s emphasizing job creation at the expense of job quality. However, research has exposed this trade-​off to be a myth (see Davoine et al. 2008; Kenworthy 2008; Osterman 2012). Updating their previous analysis of Europe undertaken with their colleague Davoine, Erhel and Guergoat-​ Larivière (2016) again demonstrate that countries can have high job quality whilst also enjoying high levels of employment. Policy can thus focus on both needs simultaneously and deliver on both simultaneously. Illustrating this shift in policy thinking, even a right-​of-​centre government such as that of the UK can, for the first time, declare that ‘fair and decent work should be available to all’ and that it now ‘plac[es] equal importance on quality work as well as quantity’ (HM Government 2018: 6).

JOB QUALITY MATTERS    9 Since the GFC, the European Union (EC 2012) and OECD (2011) have both championed policies to encourage more and better jobs amongst their member states, and the ILO (2020) has re-​affirmed its commitment to promoting its Decent Work Agenda globally. Moreover, the UN’s 2030 Agenda for Sustainable Development explicitly includes job quality among its Sustainable Development Goals (SDG).4 It is most explicit in SDG 8—​Decent Work and Economic Growth—​that ties improved job quality to productivity. Job quality also features in other SDGs. SDG 6—​Gender Equality includes female employment and opportunities to move into managerial positions, a challenge explored by Benjamin in this volume in the chapter titled ‘Job Quality for Service and Care Occupations: A Feminist Perspective’, using a feminist lens to explore service and care occupations. SDG 16—​Peace, Justice and Strong Institutions directly mentions the protection of trade unionists; although as Simms shows in the chapter titled ‘Trade Unions and Job Quality’, the actions of trade unions can have both positive and negative impacts on workers’ job quality. The European Union has also been active in terms of research and legislation to improve job quality both within and beyond its borders. The Lisbon Strategy, Europe 2020 and, most recently, the European Pillar of Social Rights all underline the importance of, and elaborate aspirations for, improved job quality within Europe. Specialized agencies such as the European Centre for the Development of Vocational Training (CEDEFOP) and the European Agency for Safety and Health at Work (EU-​OSHA) develop standards and guidelines within their jurisdictions. The European Foundation for the Improvement of Living and Working Conditions (Eurofound) regularly researches and reports on job quality (e.g. Warhurst et al. 2020), both as a concern in itself and for its role in tackling wider social, economic and political challenges. As Eiffe writes illustratively, ‘The key message of the Eurofound Sustainable Work Framework is that job quality and working conditions are at the core of keeping people engaged and working longer at a better health’ (2021: 81). Reaching beyond Europe’s borders, European Union Commissioner for Justice Didier Reynders announced the introduction in 2021 of legislation mandating corporate due diligence obligations down their supply chains on human rights and environmental harms (RBC 2020). Such legislation intends to secure voice and health rights for both workers and communities affected by negative externalities of supply chain actors of firms doing business in the European Union. This extension of reach to the global level is largely in response to and in line with the higher ambition level in the report Study on due diligence requirements through the supply chain (EC 2020), underscoring the job quality research-​to-​policy link. National policy debates also engage job quality, even where there is less engagement with international bodies. As part of President Joe Biden’s post-​Covid rescue plan for the US economy, an American Jobs Plan was pursued that was intended to create millions of good jobs. These good jobs would pay ‘prevailing’ wages in safe workplaces,

4 Department

of Economic and Social Affairs, ‘Transforming our world: The 2030 Agenda for Sustainable Development’, (un.org)/​.

10    chris warhurst, chris mathieu and rachel e. dwyer with workers able to organize into unions and bargain with employers through those unions, all underpinned by strong labour standards.5 In tone and emphasis, this Plan marks an abrupt break with recent US federal policy but building on proposals long percolating in policy advocacy for improved job quality, including the advocacy of Black leaders and advocates who draw on the legacy of the 1963 March on Washington for Jobs and Freedom among efforts that link equality at work to civil rights (Aja et al. 2013). Future research on job quality should pay more attention to inequalities by race, ethnicity and indigeneity including the important role of vulnerable populations in creating policy and advocacy innovations. Efforts focus on pushing back against eroding worker bargaining power in the US by changing labour law (National Employment Law Project), while others also develop policy innovations that enhance job quality within existing legal frameworks. Policy efforts are aided by clarity on the very issues of conceptualization and operationalization of job quality that this volume addresses. The Canadian Index of Wellbeing6 includes an index of employment quality that parallels the OECD Job Quality Framework. The Economic Policy Institute’s data repository and publications on employment quality have been instrumental in US policy discussions contesting the retreat from worker rights (EPI 2021). Also influential has been the high-​ profile debate about the polarization of job quality in North America (see Kalleberg and colleagues in their chapter). Job quality policy must often engage the shifting terrain of developments and debates around economic growth and restructuring. New debates and concerns have emerged around digitalized automation, globalization, financialization and the possibilities and perils of maintaining job quality in mature and evolving market economies (Wright and Dwyer 2003; Dwyer 2013; Dwyer and Wright 2019). Job quality thus figures increasingly on the competitiveness agenda, not just on the sustainability, welfare and normative agenda. Job quality can be a key underpinning of innovation, for example, as the chapter in this volume by Muñoz de Bustillo and colleagues, ‘Innovation and Job Quality’ demonstrates and, of course, such innovation can lead to firm-​and national-​ level productivity gains, though research on the link between job quality and productivity is fragmented and patchy and in dire need of improvement (Bosworth and Warhurst 2020). At the firm level, the operative mechanisms in High Performance Work Systems, for example, are core job quality factors, even if rarely articulated as such—​see Boxall and Purcell in their chapter in this volume. There are several ‘win-​win’ assertions that link high job quality with business outcomes, be they output parameters such as productivity, performance or innovation, or cost-​reduction parameters associated with lower turnover and ensuing recruitment and on-​boarding costs, expanded recruitment pools, fewer work stoppages, avoided litigation and fines, fewer sick-​days due to better health and fewer injuries. Sceptics on improving job quality may point to technological change or globalization as unstoppable forces that erode quality. Yet the response and

5 

6 

The White House, ‘FACT SHEET: The American Jobs Plan’. University of Waterloo, ‘The Canadian Index of Wellbeing’, (uwaterloo.ca).

JOB QUALITY MATTERS    11 implementation of change within organizations and within policy regimes can lead to very different outcomes for job quality (Dwyer 2013). Indeed, as Eurofound (2015) has shown for the European Union (see also the chapter by Goos et al.), job quality trajectories vary—​there exist upgrading, downgrading and polarization by country, despite the general trends of globalization and technological change. Such variation affirms the importance of policy and the role of labour market and related institutions (see Godard’s chapter in this volume). Within the UK, Scotland is a good illustration of this point. Following the GFC, the Scottish Government adopted and actively advocated Fair Work as a means of generating benefits for individuals, organizations and society (Fair Work Convention 2016). “We are committed to Fair Work. More security, decent pay and a greater voice for workers in the companies whose wealth they help to create,’ declared the Scottish First Minister in 2018. ‘Fair Work is good for everyone. It drives innovation and productivity. And that makes for better businesses and higher profits.’7 Significantly, as the Covid pandemic unfolded, as with Biden in the US, both the Scottish Government and UK Government reaffirmed their aspirations for more Fair Work and Good Work respectively (Scottish Government 2021; HM Treasury 2021). Such pronouncements underline how embedded the desire to create a future of good jobs has become in current government policy thinking. Two other recent developments are likely to see job quality remain prominent in government policy thinking for the foreseeable future: the rise of populism and the need for a green transformation. That populism has gained traction on the back of workers’ disaffection with job quality has not gone unnoticed on both sides of the Atlantic. In the UK, there was a strong propensity for Brexit voters to largely have low-​skill, low-​wage jobs, lacking prospects (Goodwin and Heath 2016). Similarly, having a poor-​quality job has been linked to the rise of Trump and subsequent threats to democracy in the US (Blanchflower 2019). These concerns have led the European Commission to worry about the possible unravelling of the ‘European project’. In response it has directed policy to drive upward convergence in working conditions in Europe, recognizing that unity cannot come from a single currency alone but must also be reflected in better job quality for all (Eurofound 2018). Whether as part of economic recovery following the GFC or the Covid pandemic or simply because of environmental concerns, advocacy of a transformation to a green economy also references job quality (e.g. EC 2011; Martinez-​Fernandez et al. 2010; TUC 2020). For example, the ‘just transitions’ to environmentally sustainable economies and societies championed by the ILO (2015) rests on the creation of more decent jobs. At the national level too, promises by governments of new green jobs also claim that these jobs will be good jobs (e.g. Johnson 2020). However, there is an as yet unresolved issue about how to define and measure green jobs (Sofroniou and Anderson 2021), and there is recognition elsewhere that these jobs might not all be good; that they might ‘not [be] homogenous in terms of skill requirements, pay levels or working conditions’ in the

7 

https://​news.gov.scot/​news/​fair-​work-​first-​announ​ced-​by-​fm/​.

12    chris warhurst, chris mathieu and rachel e. dwyer measured words of the OECD/​Cedefop (2014: 20), and that policy interventions may be needed to make them better. A conflation of policy interests and concerns thus means that job quality looks set to continue to feature in policy thinking and, as a result, will ensure that research into job quality also continues strongly into the future. In the meantime, this volume represents the state of the art in that research, with contributions from many of the world’s leading authorities on job quality.

The Structure of the Handbook Reflecting these issues, debates and initiatives, the Handbook comprises six sections. Section I, ‘The Foundations of Job Quality’, could equally have been entitled ‘Job Quality and its Others’ as it contains three chapters with different takes on the more expansive field of the quality of jobs, work and working life. In the chapter ‘The Quality of Working Life’, Guest addresses the nature, evaluation, impact and, eventually, the fate of the international quality of working life movement that emerged out of the UK in the post-​war period. In ‘The Swedish Contribution to Job Quality’, Hampson and Sandberg chronicle the once leading-​edge Swedish approach to research, experimentation and improvement in job quality and wider working life as well as its international significance as the so-​called ‘Swedish model’ of collaborative industrial relations. In ‘Job Quality—​A Family Affair’, Warhurst, Wright and Mathieu analyse the array of leading concepts related to job quality and the key parameters by which these concepts can be understood and analysed. It concludes with the authors highlighting core dimensions of job quality as the basis for future research. Section II, ‘Understanding Job Quality’, focuses on various means of researching job quality. In ‘Understanding Differences and Trends in Job Quality’, Hauff and Kirchner outline the different approaches in cross-​national research on job quality. From this analysis they highlight the difficulties in developing a common understanding internationally. Two chapters follow that reflect methodological approaches within the economic and social sciences. In ‘Understanding Job Quality Using Qualitative Research’, Knox and Wright show how this methodology has progressed understanding of job quality and, they argue, it helps facilitate more precise conceptualization and more effective operationalization of job quality research. ‘Quantitative Approaches to Assessing Jobs’ scrutinizes various approaches, with Muñoz-​de-​Bustillo, Fernández-​Macías and Antón offering reflections on both the feasibility of measuring job quality and the adequacy of existing measures. In ‘Institutions, Societies and the Quality of Employment’, Godard argues that job quality should be assessed at the societal rather than individual level. Elevating the assessment of job quality to the societal level enquires directly into how it is formed by social institutions, which in turn, it is contended, are derived from or responsive to ‘civic principles’.

JOB QUALITY MATTERS    13 Section III, ‘Key Issues in Job Quality’, addresses central mechanisms, features and outcomes of job quality. ‘Job Polarization’ focuses on a key debate about job quality trends. Goos, Rademakers, Salomons and Vandeweyer provide a brief history of this debate and a framework for understanding it. Polarization is not automatic, they explain, supporting their point by reference to international empirical evidence. ‘Geographies of Job Quality’ redresses an analytical oversight in debate about job quality—​its spatiality. Weller, Barnes and Kimberley highlight five areas of geographical research within which job quality is central. Context, they state, in this case spatial, is critical to understanding job quality. In ‘The Cornerstone of Job Quality’, Albin, Mathieu, Takala and Theorell make a similar point about Occupational Safety and Health (OSH) by considering which contexts mitigate biological, ergonomic, and psycho-​social hazards and risks in work as well as promote beneficial health effects of work. In ‘Innovation and Job Quality’, Muñoz-​de-​Bustillo, Grande and Fernández-​Macías investigate the relationship between technological and organizational innovation and job quality. After outlining the ways that innovation and job quality can impact each other, empirical analysis finds strong association between the two. In ‘Immigration and Job Quality’, Armenta and Gleeson examine the job quality of immigrant labour in the US and the factors shaping that job quality. Immigrants face reduced power and lack workplace protections, and both highlight the intensified vulnerability of workers based on nativity as well as the broader precarity of low-​wage work in the US labour market. In ‘Job Quality for Service and Care Occupations’, Benjamin examines job quality through the lens of feminism. Using a gender-​sensitive model she exposes the undervaluation, de-​ skilling and deprofessionalization of women’s work in service and care occupations. As Gallie points out in the final chapter of this section, ‘Inequality in Job Quality’, inequality in job quality has long been a central issue in debate and theory. An initial concern with class-​based inequalities has now been complemented by recognition of the inequalities arising from gender and contract type, and the chapter outlines the trends in these inequalities over recent decades. Section IV, ‘Regional Developments in Job Quality’, highlights similarities and variations between countries in proximate geographic or developmental contexts. All regions face significant challenges in maintaining or improving job quality. ‘Job Quality in the United States and Canada’ demonstrates the link between flexible labour markets and neo-​liberal regulatory frameworks in growing rates of precarious work. Kalleberg, Fuller and Pullman highlight several differences between the US and Canada linked to stronger unions and greater employment protections in Canada, but they also catalogue shared challenges for job quality across these liberal market economies. In ‘The Great Recession and Job Quality Trends in Europe’, Erhel, Guergoat-​Larivière, Leschke and Watt examine how greater worker protections in Europe (relative to North America) limited job quality degradation following the GFC. Nonetheless, they raise concerns about particular dimensions of job quality including wages and for particular groups such as youth who bear the brunt of deteriorating quality. In this section, Cazes, Falco and Menyhért examine ‘Job Quality in Emerging Economies’ through the lens of the

14    chris warhurst, chris mathieu and rachel e. dwyer OECD Job Quality Framework. The chapter highlights the shared challenges between emerging economies and liberal market economies in North America and Europe. Workers face worse job quality in emerging economies but along the same dimensions and with similar needs, Cazes, Falco and Menyhért argue, for greater legal protections and social insurance. Section V, ‘Sectoral Developments in Job Quality’, dives deep in examining trajectories of job quality within particular sectors and broadly links industrial restructuring to threats to job quality. In ‘Job Quality in High Touch Services’, Gatta argues that as demand for services with interpersonal interaction has grown, so too have expectations that these services be available 24/​7. In the US context with low worker bargaining power, workers must adjust to often punishing and insecure schedules in order to stay employed. Baldry lays out similar challenges of work intensification in the context of office work in ‘The Changing Quality of Office Work’. Charting historical trends in office work, current tensions between a drive to create fun offices of the future and the routinization of work that reinforces proletarianism are highlighted. Taking an international view, Rothstein argues in ‘The Steady but Uneven Decline in Manufacturing Job Quality’, that manufacturing jobs became lower quality in places where worker power was diminished or always highly circumscribed. Only in places such as Germany where labour rights and unions remained stronger did manufacturing jobs continue to provide decent job quality. In ‘Neoliberalism’s Impact on Public Sector Job Quality’, Sauer, Valet, Roscigno and Wilson demonstrate declining job quality for public-​sector jobs after corporate-​styled efficiency reforms. Sauer and colleagues show that even in Germany where worker rights have remained more robust, neo-​liberalism has made inroads. These chapters highlight distinct challenges based on the particular work demands in different sectors but shared threats to job quality in the context of weak worker power. The final chapter in this section, ‘Job Quality and the Small Firm’, by Edwards and Ram addresses three issues: the paradox of job quality in small firms, variations in job quality across small firms and the relative neglect of small firms in policy debates about job quality. It ends with reflections from the authors about how to better promote job quality in small firms. The chapters in Section VI, ‘Improving Job Quality’, focus on actors, means and approaches or avenues to support good job quality. This issue—​ maintaining or improving job quality—​features strongly in both research and policy. Arguments for reducing it are seldom made, even if research shows that job quality can and does deteriorate. ‘Human Resource Management and Job Quality’ explores human resource management (HRM) within organizations as a lever of job quality. Boxall and Purcell note, however, that HRM research has tended to ignore job quality but that a strong case exists for it to now focus on it. In the chapter ‘Using Efficiency, Equity and Voice for Defining Job Quality and Legal Regulation for Achieving It’, Befort, Borelli and Budd argue that the basic elements to define and conceptualize job quality lie embedded in the nature of the employment relationship: efficiency, equity and voice. Contrasting the US with the EU, they assess to what extent existing regulatory regimes live up to the ideal of

JOB QUALITY MATTERS    15 providing this efficiency, equity and voice. Similarly, in the final chapter in this section, ‘Trade Unions and Job Quality’, Simms’ starting position is that job quality is an outcome of the employment relationship. Simms then reviews the links between trade unions and debates about job quality. The chapter notes that trade union strategies can both increase and diminish job quality, and considers the evidence of union opportunities to promote and improve it. This point of exit for the Handbook is deliberate. As we state, in addition to researchers firstly discerning and secondly explaining job quality, this third issue of maintaining or improving it now occupies the minds of policymakers. It will do so even more in the coming years. As national economies recover from the jobs crisis that followed the health crisis of Covid illustrates, all three issues will become important. Provided by some of the world’s leading experts, the contents of this Handbook provide an intellectual toolkit for researchers, students and policymakers who need or want to understand and address these issues.

References Aja, A., Bustillo, D., Darity, W. and Hamilton, D. (2013) ‘Jobs Instead of Austerity: A Bold Policy Proposal for Economic Justice.’ Social Research, 80/​3, 781–​794. Albin, M., Mathieu, C., Takala, E-​P, and Theorell, T. (2022), ‘The Cornerstone of Job Quality –​ Occupational Safety and Health’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, (eds.,) Oxford Handbook of Job Quality. (Oxford: Oxford University Press), 220–243. Armenta, A., and Gleeson, S. (2022), ‘Immigration and Job Quality’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 274–298. Baldry, C. (2022), ‘The Changing Quality of Office Work’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 433–451. Bamber, G. J., Lansbury, R. D., Wailes, N. and Wright, C. F. (2016), International and Comparative Employment Relations (Los Angeles: Sage). Befort, S. F., Borelli, S., and Budd, J.W. (2022), ‘Using Efficiency, Equity and Voice for Defining Job Quality, and Legal Regulation for Achieving It’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 522–541. Bell, D. (1973), The Coming of Post-​Industrial Society (New York: Basic Books). Benjamin, O. (2022), ‘Job Quality for Service and Care Occupations: A Feminist Perspective’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 299–317. Blanchflower D. (2019), Not Working: Where Have All the Good Jobs Gone? (Princeton: Princeton University Press). Bosworth, D., and Warhurst, C. (2020), ‘Does Good Work Have a Positive Effect on Productivity? Developing the Evidence Base’, in G. Irvine, ed., Can Good Work Solve the Productivity Puzzle? (Dunfermline: Carnegie UK Trust), 10–​19. Boxall, P., and Purcell, J. (2022), ‘Human Resource Management and Job Quality’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 505–521.

16    chris warhurst, chris mathieu and rachel e. dwyer Braverman, H. (1974), Labor and Monopoly Capital (New York: Monthly Review Press). Bray, M., Waring, P., and Cooper, R. (2009), Employment Relations: Theory and Practice (Sydney: McGraw-​Hill). Cazes, S., Falco, P., and Menyhért, B. (2021), ‘Job Quality in Emerging Economies through the Lens of the OECD Job Quality Framework’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 389–413. Cazes S., Hijzen, A., and Saint-​Martin, A. (2015), ‘Measuring and Assessing Job Quality: The OECD Job Quality Framework’, OECD Social, Employment and Migration Working Papers No. 174. Paris: OECD. Combs, J., Liui, Y, Hall, A., and Ketchen, D. (2006) ‘How Much Do High Performance Work Practices Matter? A Meta-​ Analysis of Their Effects on Organisational Performance’, Personnel Psychology, 59/​3, 501–​528. Davoine, L., Erhel, C., and Guergoat-​Larivière M. (2008) ‘Monitoring Employment Quality in Europe: European Employment Strategy Indicators and Beyond’, International Labour Review, 147/​2–​3, 163–​198. Doogan, K. (2009), New Capitalism? (Cambridge: Polity). Dwyer, R. E. (2013), ‘The Care Economy? Gender, Economic Restructuring, and Job Polarization in the U.S. Labor Market’, American Sociological Review, 78, 390–​416. Dwyer, R. E., and Wright E. O. (2019), ‘Low-​Wage Job Growth, Polarization, and the Limits of the Service Economy’, Russell Sage Foundation Journal of the Social Sciences, 5/​4, 56–​76. Economic Policy Institute (EPI) (2021), https://​www.epi.org/​. Edwards, P. and Ram, M. (2022), ‘Job Quality and the Small Firm’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 485–503. Eiffe, F. (2021) ‘Eurofound’s Reference Framework: Sustainable Work over the Life Course in the EU’, European Journal of Workplace Innovation, 6/​1–​2, 67–​83. Erhel, C., and Guergoat-​Larivière, M. (2016), ‘Innovation and Job Quality Regimes: A Joint Typology for the EU’, QuInnE Working Paper No. 3, QuInnE. Erhel, C., Guergoat-​Larivière, M., Leschke, J., Watt, A. (2022), ‘The Great Recession and Job Quality Trends in Europe’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 363–388. Eurofound (2015), ‘Upgrading or Polarisation? Long-​ term and Global Shifts in the Employment Structure’, European Jobs Monitor 2015. Dublin: Eurofound. Eurofound (2018), Upward Convergence in the EU: Concepts, Measurements and Indicators (Dublin: European Foundation for the Improvement of Living and Working Conditions). European Commission (EC) (2011), A Roadmap for Moving to a Competitive Low Carbon Economy in 2050, COM(2011) 112 Final (Brussels: European Commission). European Commission (EC) (2012), Europe 2020 http://​ec.eur​opa.eu/​eur​ope2​020/​. European Commission (EC) (2020), Study on Due Diligence Requirements through the Supply Chain (Directorate-​General for Justice and Consumers Directorate, Brussels: European Commission). Fair Work Convention (2016), Fair Work Framework 2016, www.fairworkconvention.scot/​. Felstead, A., Gallie, D., Green, F., and Henseke, G. (2019), ‘Conceiving, Designing and Trailing a Short-​Form Measure of Job Quality: A Proof-​Of-​Concept Study’, Industrial Relations Journal, 50/​1, 2–​19. Fernández-​Macías, E. (2010), ‘Job Polarisation in Europe? Changes in the Employment Structure and Job Quality, 1997-​2005’, Work and Occupations, 39/​2, 157–​182.

JOB QUALITY MATTERS    17 Gallie, D. (2022), ‘Inequality in Job Quality: Class, Gender and Contract Type’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 318–337. Gatta, M. (2022) ‘Job Quality in High Touch Services’, in C. Warhurst, C. Mathieu, and R. E. Dwyer (eds) Oxford Handbook of Job Quality (Oxford: Oxford University Press), 415–432. Gautié, J., and J. Schmitt (eds.) (2010), Low Wage Work in a Wealthy World (New York: Russell Sage Foundation). Ghai, D. (2003), ‘Decent Work: Concept and Indicators’, International Labour Review, 142/​2, 113–​145. Godard, J. (2022), ‘Job Quality as the Realization of Democratic Ideals’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 148–167. Goodwin, M., and Heath, O, (2016) Brexit vote explained: poverty, low skills and lack of opportunities. York: Joseph Rowntree Foundation. Goos, M., and Manning, A. (2007), ‘Lousy and Lovely Jobs: The Rising Polarization of Work in Britain’, Review of Economics and Statistics, 89/​1, 118–​33. Goos, M., Rademakers, E., Salomons A., and Vandeweyer, M. (2021), ‘Job Polarization: Its History, An Intuitive Framework and Some Empirical Evidence’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 169–202. Guest, D. (2022), ‘The Quality of Working Life’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 23–40. Hall, C. (2009), ‘The Reform Bill is Passed’, in The Great Turning Points in British History (London: Constable), 155–​163. Hampson, I., and Sandberg, Å. (2022), ‘The Swedish Contribution to Job Quality’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 41–62. Hauff, S. and Kirchner, S. (2022), ‘Understanding Differences and Trends in Job Quality: Perspectives from Cross-​National Research’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 87–106. HM Government (2018), Good Work Plan (London: HM Government). HM Treasury (2021), Build Back Better (London: HM Treasury). Holman, D. (2013), ‘Job Types and Job Quality in Europe’, Human Relations, 66/​4, 475–​502. Hurley, J., Fernández-​ Macías, E., and Muñoz de Bustillo, R. (2012), ‘Assessing Recent Employment Shifts in Europe Using a Multidimensional Job Quality Indicator’, in E. Fernández-​Macías, J. Hurley, and D. Storrie, eds., Transformation of the Employment Structure in the EU and USA, 1995–​2007 (London: Routledge), 147–​179. ILO (1999), Decent Work: Report of the Director General (Geneva: International Labour Organization). ILO (2015), Guidelines for a Just Transition towards Environmentally Sustainable Economies and Societies for All (Geneva: ILO). ILO (2020), Decent Work and the 2030 Agenda for Sustainable Development (Geneva: ILO). Johnson, B. (2020), ‘PM Outlines his Ten Point Plan for a Green Industrial Revolution for 250,000 Jobs’, https://​www.gov.uk/​gov​ernm​ent/​news/​pm-​outli​nes-​his-​ten-​point-​plan-​for-​ a-​green-​ind​ustr​ial-​rev​olut​ion-​for-​250​000-​jobs/​. Kalleberg, A. L., Fuller, S., and Pullman, A. (2021), ‘Job Quality in the United States and Canada’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 339–362.

18    chris warhurst, chris mathieu and rachel e. dwyer Keep, E., and Mayhew, K. (2010), ‘Moving Beyond Skills as a Social and Economic Panacea’, Work, Employment and Society, 24/​3, 565–​557. Kenworthy, L. (2008), Jobs with Equality (Oxford: Oxford University Press). Knox, A., Warhurst, C., and Pocock, B. (2011), ‘Job Quality Matters’, Journal of Industrial Relations, 53/​5, 5–​11. Knox, A., and Wright, S. (2022), ‘Understanding Job Quality Using Qualitative Research’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 107–125. Kochan, T. A. (1998), ‘What is Distinctive about Industrial Relations Research’ in K. Whitfield, and G. Strauss, eds., Researching the World of Work (Ithaca: Cornell University Press), 31–​46. Legge, K. (1995), Human Resource Management (Houndmills: Macmillan). Leitch Review of Skills (2006), Prosperity for All in the Global Economy: World Class Skills (London: HM Treasury). Livingstone, D. W. (2017), ‘Skill Under-​Utilisation’, in C. Warhurst, K. Mayhew, D. Finegold, and J. Buchanan, eds., Oxford Handbook of Skills and Training (Oxford: Oxford University Press), 281–​300. Martinez-​Fernandez, C., Hinojosa, C. and Miranda, G. (2010), Green Jobs and Skills: The Local Labour Market Implication of Addressing Climate Change (Paris: OECD). Measuring Job Quality Working Group (2018), Measuring Good Work (Dunfermline: Carnegie UK Trust). Muñoz-​de-​Bustillo, R., Fernández-​Macías, E., Antón, J., and Esteve, F. (2009), Indicators of Job Quality in the European Union (Brussels: Department of Employment and Social Affairs). Muñoz de Bustillo, R., Fernández-​Macías, E., Esteve, F., and Antón, J. (2011), Measuring More Than Money (Cheltenham: Edward Elgar). Muñoz-​ de-​ Bustillo, R. Fernández-​ Macías, E., and Antón, J-​ I. (2022a), ‘Quantitative Approaches to Assessing Job Quality’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 126–147. Muñoz-​de-​Bustillo, R., Grande, R., and Fernández-​Macías, E. (2022b), ‘Innovation and Job Quality’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 244–273. OECD (2011), Divided We Stand: Why Inequality Keeps Rising (Paris: OECD). OECD/​Cedefop (2014), Greener Skills and Jobs (Paris: OECD). Osterman, P. (2012), ‘Job Quality in the US: The Myths That Block Action’, in C. Warhurst, P. Findlay, C. Tilly, and F. Carré, eds., Are Bad Jobs Inevitable? (London: Palgrave), 45–​60. Osterman, P., and Shulman, B. (2011), Good Jobs in America (New York: Russell Sage Foundation). RBC (2020), European Commission Promises Mandatory Due Diligence Legislation in 2021 (Brussels: European Parliament Working Group on Responsible Business Conduct, Brussels). https://​op.eur​opa.eu/​en/​publ​icat​ion-​det​ail/​-​/​publ​icat​ion/​8ba0a​8fd-​4c83-​11ea-​ b8b7-​01aa7​5ed7​1a1/​. Rothstein, J. S. (2022), ‘The Steady but Uneven Decline in Manufacturing Job Quality’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 152–168. Sauer, C., Valet, P., Roscigno, V. J., and Wilson, G. (2022) ‘Neoliberalism’s Impact on Public Sector Job Quality: The US and Germany Compared’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 469–484. Schwab, K. (2016), The Fourth Industrial Revolution (London: Penguin).

JOB QUALITY MATTERS    19 Scottish Government (2021), Fair Work Action Plan: Annual Report (Edinburgh: Scottish Government). Simms, M. (2022), ‘Trade Unions and Job Quality’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 542–558. Sofroniou, N., and Anderson, P. (2021), ‘The Green Factor: Unpacking Green Job Growth’, International Labour Review, 160/​1, XX. Trist, E., and Bamforth, K. (1951), ‘Some Social and Psychological Consequences of the Longwall Method of Coal Getting’, Human Relations, 4, 21–​32. TUC (2020), Rebuilding after Recession: A Plan for Jobs (London: TUC). UNECE (2015), Handbook on Measuring Quality of Employment: A Statistical Framework (New York and Geneva: United Nations). Warhurst C., Erhel, C., Gallie, D., Guergoat-​Larivière, M., Muñoz de Bustillo, R., Obersneider, M., Postels, D., Sarkar, S., and Wright, S. (2018), ‘Data Evaluation Report: An Evaluation of the Main EU Datasets for Analysing Innovation, Job Quality and Employment Outcomes’, QuInnE Working Paper No. 14, QuInnE. Warhurst, C., and Hunt, W. (2019), ‘The Digitalisation of Future Work and Employment. Possible Impact and Policy Responses’, Joint Research Council Working Papers Series on Labour, Education and Technology 2019/​05. Seville: European Commission. Warhurst, C., and Knox, A. (2022), ‘Manifesto for a New Quality of Working Life’, Human Relations, 75(2): 304–​321. Warhurst, C., Muñoz de Bustillo Llorente, R., Antón Pérez, J-​I., Grande, R., Hernández, F. P., and Wright, S. (2020), Upward Convergence in Working Conditions Luxembourg: (Brussels: Publications Office of the European Union). Warhurst, C., Wright, S., and Mathieu, C. (2022), ‘Job Quality: A Family Affair?’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 63–85. Weller, S., Barnes, T., and Kimberley, N. (2022) ‘Geographies of Job Quality’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds, Oxford Handbook of Job Quality (Oxford: Oxford University Press), 203–219. Wright, S. (2015), ‘Challenges in Researching Job Quality’, in A. Knox and C. Warhurst, eds., Job Quality in Australia: Perspectives, Problems and Proposals (Annandale, NSW: Federation Press), 15–​36. Wright, E. O., and Dwyer, R. E. (2003), ‘The Patterns of Job Expansions in the USA: A Comparison of the 1960s and 1990s’, Socio-​Economic Review, 1, 289–​325. Zemanik, M. (2020), Working Lives Scotland (London: CIPD).

PA RT I

T H E F OU N DAT ION S OF J OB QUA L I T Y

Chapter 1

THE QUALI T Y OF WORKING L I FE david guest

Introduction The quality of working life emerged as an important topic for research and policy in the 1970s. Since then it has had a chequered history as research fragmented and policy priorities changed. However, interest in the general topic has remained and there have been recent calls to revitalize quality of working life research (Grote and Guest 2017). While job design has always been at the heart of quality of working life considerations, interest has ranged more widely, offering a context for job design improvements. The aim of this chapter is to trace the evolution of quality of working life research and associated policy interventions, to evaluate its impact and to assess its current relevance, including its relevance for work design.

From Work Redesign Towards Quality of Working Life: Initial Steps A common starting point for a consideration of research on the quality of working life (QWL) is the work of the Tavistock Institute of Human Relations. Set up in 1947, the Institute undertook pioneering research that led to the concept of socio-​technical systems theory. The key demonstration case was reported by Trist and Bamforth (1951) based on a study of UK coal mines. They explored the effect of transferring from the traditional method of extracting coal, based on socially integrated teams working fairly autonomously with their own internal supervision, to the technologically superior longwall method of coal-​getting. Despite its potential for efficiency gains, the new method

24   david guest broke up the team structure and led to suboptimal productivity as well as increases in absence and labour turnover. The important insight was that it was not sensible to focus on maximizing the technical system alone, as Taylorism seemed to advocate, or the social system alone, which the human relations movement emphasized, but instead it was necessary to optimize both at the same time. This approach might involve some compromises but promised an improved method of work design with the potential for superior performance and higher work-​team satisfaction. They subsequently developed what was termed the ‘composite longwall system’ which more closely reflected the socio-​technical principles and which appeared to produce superior outcomes (Trist et al. 1963). Socio-​technical systems thinking was successfully applied in the very different context of an Indian textile mill (Rice 1958) where the initial problem had been the difficulty facing supervisors in managing the complexity of the boundary conditions. These challenges were effectively resolved by restructuring the working groups. However, implementation of changes to improve socio-​technical systems was not straightforward. Trist and his colleagues conducted a later study to implement autonomous group working in American coal mines but despite initial success ran into difficulties presented by the wider system and in particular negative trade union reactions, which limited implementation (Trist, Susman and Brown, 1977). The experiences and perspectives acquired in a number of studies by members of the Tavistock Institute and other enthusiasts of their approach were brought together in a number of books and articles presenting expositions of the socio-​technical principles for work and organizational systems design (Cherns 1976; Emery and Trist 1968; Herbst 1976). Although a number of UK organizations displayed interest in the ideas emanating from the Tavistock Institute and a few, such as Shell, tried to incorporate them, the significant developments extending work redesign into a broader concern for quality of working life occurred in Norway where interest in the scope to implement forms of industrial democracy led to a research programme—​The Cooperation Project—​ jointly sponsored and initially funded by the national trade union and management associations and subsequently by the government. It was directed, over a period from 1962 to 1970, by Einar Thorsrud, Research Director of the Institute for Industrial Social Research at the University of Trondheim from which he later moved to be Director of the Work Research Institute in Oslo. The programme was intended to take the form of action research. Thorsrud had a spell as a visitor to the Tavistock Institute and, from the start, members of the Tavistock Institute, and most notably Fred Emery, were heavily involved in the Norwegian programme. An early step involved an evaluation of an existing form of industrial democracy, namely the presence of worker representatives on the boards of companies. Emery and Thorsrud (1969) reported that this led to potential role conflict and limited opportunities for workers’ representatives to exert influence, and since it had very little impact on the shop floor it therefore failed as an initiative to improve industrial democracy for the great majority of workers. They based this conclusion not only on their Norwegian findings but also on their study of the experience in Yugoslavia, West Germany and Great Britain (Emery and Thorsrud 1969). Their

THE QUALITY OF WORKING LIFE    25 findings are echoed in the evaluation of worker directors on the board of the British Steel Corporation (Brannen et al. 1976). The main Norwegian research took place in four case study organizations through programmes of action research reported by Emery and Thorsrud (1976). They set out what they described as ‘a general set of . . . principles’ (1976: 15) for job redesign as a starting point for their work. While these were the based on the Tavistock Institute’s socio-​technical systems thinking, they also acknowledged the considerable influence of Lewin’s ideas (Lewin et al. 1939) about the powerful role of groups as well as the work of Davis (1957), an engineer who subsequently came to play a leading role in the quality of working life movement. Taken together, their ideas formed the basis for the concept of autonomous work groups. In addition to the focus on the redesign of work, Emery and Thorsrud also set out what they described as ‘general psychological requirements pertaining to the content of jobs’ 1976: 14). This normative set of concerns drew on individual psychology and included the ‘need’ for work to be reasonably demanding; for some opportunity for learning on the job; for some minimal area of decision-​making; for some minimal degree of social support; for the ability ‘to relate what the worker does and what he produces to his social life’; and ‘the need to feel that the job leads to some sort of desirable future(p. 14). These provided one set of criteria against which to judge whether the job redesign had been successful. In going beyond the confines of the job, this identification of ‘needs’ begins to stray into the broader notion of the quality of working life. While acknowledging other influences, Emery and Thorsrud claimed that the Norwegian project was distinctive because of the institutional involvement of trade union and management bodies and later the government so that ‘within the Norwegian cultural framework it was possible in the 1960s to establish a set of starting conditions for large scale social change over a period of at least ten years’ (p. 3, their italics). In summary, work redesign, primarily based on socio-​technical principles and action research to develop autonomous work groups was viewed as the means of introducing industrial democracy and of improving the quality of working life. The ideas from the Norwegian industrial democracy project attracted attention in the other Scandinavian countries and most notably in Sweden, where a much more extensive series of initiatives got underway (see Hampson and Sandberg in this volume). Around the same time, a number of leading companies in Europe had identified a need to develop new ways of working, often provoked by concerns about poor industrial relations and low motivation among employees. One such company, Philips, had begun some initial work on job design in its Dutch factories in the late 1950s, and by 1963 had produced a report, ‘Work-​Structuring for Unskilled Workers’, which set out the principles for work structuring. The report also recognized that the features of the work environment that have implications for quality of working life could be improved including, for example, humidity, temperature and noise; but these concerns were downplayed in comparison to the importance of redesign of work. A summary of the principles of work restructuring, the process by which it was introduced and examples of its application were outlined in two subsequent internal reports ‘Work Structuring: A

26   david guest Summary of Experiments at Philips’—​1963 to 1968 (Philips, undated) and a later report on Participation (Philips 1975), which reviewed a range of studies within Philips’ European plants. In it was a recognition of the need to extend participation through work redesign to respond to improvements in education, rising employee expectations and indications of dissatisfaction with the limited freedom provided by most jobs. In the UK, Shell also initiated a programme of work redesign in the 1960s, recognizing from the outset that this involved a significant change of approach to management that went beyond the design of jobs. As in the Norwegian case, members of the Tavistock Institute were heavily involved as consultants, initially helping to draw up and ‘sell’ to senior managers a new philosophy of management (Hill 1971). This effort was viewed with varying degrees of enthusiasm by refinery managers but there were some notable attempts to restructure work. ICI in the UK, also driven by concerns about poor industrial relations and low employee motivation, embarked on productivity deals that focused on work redesign. They sponsored research by Herzberg and colleagues (Paul et al. 1969) to provide convincing evidence of the positive effects of job redesign based on the kind of quantitative data that might appeal to scientists and engineers. Following a productivity deal, scope to redesign work was extended throughout the company, with formal involvement of local shop stewards linking redesign of jobs to changes in pay grades via job evaluation (Roeber 1975). Despite its considerable influence, the Norwegian project attracted quite extensive academic criticism. Bolweg (1976), for example, raised a number of concerns. Firstly, he argued that the studies had not been as successful on a number of criteria as Emery and Thorsrud had claimed. Second, he suggested that the Norwegian unions expressed only lukewarm support for the programme and remained more interested in representative industrial democracy; indeed, Bolweg noted that by the early 1970s they had lost interest in work redesign and were focusing more on health and safety at work. Third, he argued that the changes ran out of steam. Although worker representatives and other staff were heavily involved in the programme to redesign jobs, once this process was complete it was unclear what came next. With management power restraining further developments in participation, there was an absence of diffusion within the case study organizations. Relatedly, Bolweg argued that there was a wider problem of diffusion in Norway. He analysed 32 other Norwegian work redesign initiatives and found that very few had actually got off the ground. He quotes the Cooperation Council in 1973 as finding that ‘[t]‌he interest in starting local cooperation projects can only be evaluated as absolutely minimal given the publicity the Cooperation Project has gotten” (1976: 51). The issue of diffusion was also a major focus of Blackler and Brown’s (1980) analysis of work redesign at Shell in their provocatively titled ‘Whatever Happened to Shell’s New Philosophy of Management?’. Based on a series of interviews with managers and staff in Shell, they concluded that despite some early action, the philosophy had failed to bed in and had gradually disappeared into the background. Underpinning their analysis was a view, also expressed in their analysis of innovations in work redesign in the motor industry (Blackler and Brown 1978) that management support for this approach was motivated primarily by a desire to overcome industrial relations problems and enhance

THE QUALITY OF WORKING LIFE    27 productivity, and in so doing reassert management control. Roeber’s (1975) account of the ICI initiatives similarly outlines how the initiatives largely failed to diffuse throughout the company, mainly as a result of management opposition. In summary, there is no doubt that industrial relations problems and the challenges of enhancing productivity were major reasons why there was support for these initiatives in the two UK companies. It is also notable that it was considered necessary to introduce work redesign in the context of productivity deals that acknowledged that pay would need to be improved as a basis for gaining union agreement for potential changes. Therefore, while there was some consideration in Shell of the need for a new management philosophy, there was no view that improvements in quality of working life or the extension of industrial democracy was an end in itself or a necessary or fruitful path to pursue.

The Context of the 1960s and the Rise of Interest in Quality of Working Life A feature of the approach adopted by members of the Tavistock Institute towards the redesign of work was a view of organizations as open systems, subject to the vicissitudes of a turbulent external environment. Emery and Thorsrud (1976) had made reference to the problems of alienation and of the reluctance of some workers to undertake traditional factory work. In the US, Blauner’s (1964) study of alienation among factory workers had been highly influential and the study had been linked to wider problems in society. Concerns about growing alienation had been instrumental in the setting up in 1971 of a Task Force by the Secretary for Health, Education and Welfare to investigate and report on the state of work in America. Early in the report, it is claimed, ‘There is now convincing evidence that some blue-​collar workers are carrying their frustrations home and displacing them in extremist social and political movements or in hostility toward the government’ (O’Toole 1973: 30). Echoing more recent political developments in the US, the report goes on to say, ‘These “alienated” workers tend to participate less in elections and, when they do vote, tend to cast their ballots for extremist or “protest” candidates’ (1973: 31). In short, the problems of alienation from work spilled over into potentially problematic behaviour outside work. The causes of alienation were attributed to the greater routinization and reduction in autonomy at work caused by advances in technology, and the pervasive effects of bureaucratization of both blue-​ collar and increasingly also white-​collar work. These trends coincided with an increasingly well-​educated workforce with rising expectations and, when these expectations were not met, a tendency to react with increased absenteeism and labour turnover. The proposed solution lay in reversing these processes by improving the quality of working life and notably by humanizing work through work redesign. The pessimistic tone of the analysis that informed the Work in America report can be contrasted with the wider political and social environment of the late 1960s and

28   david guest early 1970s. This was the era of student protests and associated demands for greater democracy as well as the rather heady optimism of ‘flower power’. It was also the beginning of the women’s liberation movement and calls for greater emancipation. These movements created a sense of optimism about the possibility of change for the better. The academic literature was also highlighting the imminent advent of what was termed ‘the post-​industrial society’ (Bell 1974) in which work was being transformed away from that dealing with machinery in factories. It was in this context that in 1972 a conference was organized at Arden House in New York, bringing together leading researchers from the US and Europe to review developments in theory and research on the quality of working life and to identify ways of promoting it more widely (Davis and Cherns 1975). A core assumption informing the conference was that ‘a body of knowledge and techniques common to, and usable by, the institutions and organizations centrally concerned with solving the problems of transition into the postindustrial era is partially available’ (Davis and Cherns 1975: 6). They go on to argue that ‘[t]‌he key opportunity before us lies in formulating a coherent body of theory and practice on how to create the conditions for a humane working life in its relevant social environments’ (p. 8). This conference, together with the Work in America report provided a high point in what might be described as the quality of working life movement. It coincided with a series of initiatives such as the setting up of a number of research institutes in the US and countries in Europe as well as in Canada and Australia. While work redesign was accepted as lying at the heart of steps to improve the quality of working life, there was still a lack of clarity about what this wider approach entailed. The most influential attempt to provide a wider definition was provided by Walton (1974, 1975). He identified eight characteristics of the quality of working life. These eight were:

1. 2. 3. 4. 5. 6. 7. 8.

Adequate and fair compensation Safe and healthy working conditions Opportunity to use and develop human capacities Opportunity for continued growth and security Social integration in the work organization Constitutionalism in the work organization Work and the total life space The social relevance of work life

Walton acknowledged that these criteria are based on his ‘personal observations, experiences, values and assumptions about human nature’ (p. 100). It was also possible for participants at the Arden House conference to claim that there was a growing evidence base to support each of these dimensions of quality of working life. They have some overlap with the ‘general psychological requirements’ identified by Emery and Thorsrud (1976) as well as criteria designed to address the problem of alienation. It is notable that these criteria go well beyond work redesign, setting it in a much wider psychological and social context.

THE QUALITY OF WORKING LIFE    29

Evaluating the Impact of the Quality of Working Life Movement External factors that had helped to provoke interest in the quality of working life also affected its subsequent impact. While many initiatives were reported, success stories were relatively few and far between. The research institutes set up mainly either by governments or by universities undertook extensive analysis but had relatively few examples of significant implementation of large-​scale change. At the end of the 1970s, NATO sponsored another conference, held this time in Thessaloniki, Greece, which again attracted a large number of American and European researchers. The mood was much less optimistic than at the Arden House conference. By this time, the economic environment had become less favourable. This change had started with the oil crisis of 1973 when, following the Yom Kippur war, oil-​producing countries had restricted output. Lupton and Tanner (1980: 234) had noted that subsequently ‘the general depression of the Western economies has reduced the incentive to remodel manufacturing systems and has stunted investment in experimental systems’. While Van Assen and Wester (1980) concluded that ‘if one takes a closer look at the results and impact of this 20 year history of QWL projects, the picture is rather dismal. A lot of these projects, though reported to be successful, never seemed to have outgrown their particular status of experiments, and little or nothing of them sparked off to their organizational surroundings’ (p. 237). In a number of countries, the unions had become more entrenched in their opposition to management initiatives. The hopes expressed at Arden House about improving the quality of working life through sound empirical evidence, action research and effective communication of the benefits had begun to fade. At the Thessaloniki conference, a paper by Guest et al. (1980) had identified a number of constraints on progress towards an improved quality of working life. An initial problem lay in agreeing a definition of quality of working life. The focus in Norway had been on extending industrial democracy to the shop floor mainly through direct participation. Yet right from the formulation of the ‘required psychological conditions’ there had been a wider concern to improve the quality of working life. However, Walton’s list of indicators suggests a much broader approach going well beyond the scope of a local work redesign project. This was recognized, for example, at Shell with the need to develop a new philosophy of management which would be more in sympathy with the values behind improving the quality of working life (Hill 1971). However, as the Shell case illustrates, any wider approach also contains significant risks as it challenges management power and control, and requires considerable efforts to convince management of its merits. In both Shell and ICI, this need proved to be too great a challenge (Blackler and Brown 1978; Roeber 1975; Pettigrew 1975). And as the experience at both of these organizations had indicated, unions viewed quality of working life as a focus of collective bargaining. Addressing this point, Hughes and Gregory (1974) had argued, ‘British unions should not allow themselves to be trapped by their recognition

30   david guest that attempts to improve the quality of working life are of great importance . . . Unions should insist that the quality of working life can only be improved by a comprehensive and coordinated concern with improvements in the total work situation, and in the total remuneration package.’ This statement is both an attack on the limitations of job redesign and confirmation that Walton may be right in identifying the need to address a wider range of issues. The statement also reflects the standard trade union view that even changes that result in significant gains for workers’ experience of work should also lead to better remuneration. Another constraint raised by Guest et al. was the lack of real interest in improving quality of working life among senior managers, particularly those who had the power to determine whether changes took place. By way of illustration, managers were willing to espouse equal opportunities for women at work and to accept the need for safe working environments, both features of quality of working life; but it took legislation to ensure action. This point is returned to further on. While this reflects lack of commitment to improve quality of working life, it also illustrates management’s concern to retain power and prioritize corporate objectives. However, this lack of support for job redesign if not for other aspects of quality of working life in many countries also extended to trade unions, first because of a concern that increases in direct participation might weaken the role of representative participation which provided their own power base, and because they suspected that job redesign in particular was viewed by management as a means of improving efficiency and of seeking to develop a unitarist approach to the employment relationship (Kelly 1978). A further limitation identified by Guest et al. was the weakness of the various main theories of job redesign. For example, Herzberg’s advocacy of job enrichment had appealed to managers, partly because he eschewed any form of worker participation in the process. But the theory underpinning his approach to job enrichment had received stinging academic criticism (House and Wigdor 1967; King 1970), and when attempts were made to apply job enrichment to shop floor workers, as in the case of ICI, it faced union opposition (Roeber 1975). The more influential socio-​technical systems approach had been extensively criticized, both conceptually (Kelly 1978), and for its failure to deliver more than superficial changes, even in the context of the initial Norwegian studies (Bolweg 1976). While it served as a useful analytic framework, Bolweg noted that it was hard for managers to grasp some of the rather ambiguous concepts associated with it. He also observed that the reported studies mostly took the technology as given and sought to mitigate its effects by redesigning the social system. It was therefore not really valid to consider it as the application of socio-​technical systems design. A further challenge that was identified for the quality of working life movement concerned the process of change. The core strategy was a combination of action research and organization development (OD). Action research invariably involved the use of external consultants who undertook specific projects, usually intended to act as demonstrations of the potential benefits of their approach. However, as the Norwegian cases illustrate, this process is invariably slow and labour intensive and too little attention was given to what happens when the consultants move on. This

THE QUALITY OF WORKING LIFE    31 omission links to an important related problem, that of diffusion. Here the strategy of change becomes, in the language of Chin and Benne (1976), a mix of empirical rational—​there is empirical evidence about the benefits of the approach and the next rational step is therefore to adopt it more widely; and normative re-​educative—​ managers and worker representatives need to be educated about the benefits of this approach and helped to change their values so that they accept that this path is clearly the right one to pursue. The evaluation by Blackler and Brown (1978) of the attempts to implement Shell’s new philosophy of management and Pettigrew’s (1975) detailed account of the long-​term attempts to introduce change using an OD strategy in ICI illustrate the problems of this approach. At the heart of the problem lies management’s power to prevent changes they do not support. When this issue was allied to union opposition or disinterest, then changes to redesign work and improve quality of working life became very difficult to achieve. If the challenges of improving the quality of working life lay beyond the capacity of enthusiasts within commercial organizations supported by external consultants and academics, maybe the answer lay in government action. In the early 1970s, governments in several countries set up organizations designed to improve the quality of working life. However, the UK had not taken any action until it was prompted by a NATO Committee on the Challenges of Modern Society that asked the government to sponsor a ‘pilot project’ into the problems of work motivation and satisfaction (Wilson 1973). The response was to set up a committee to review the evidence. The subsequent report ‘On the Quality of Working Life’ (Wilson 1973) covered many of the issues discussed at the Arden House Conference (Albert Cherns, who played a leading part in the American conference was a member of the committee), and recommended a ‘cell’ within the relevant government department to undertake and evaluate quality of working life initiatives, working closely with academics and industrial organizations. The result was the setting up within the Department of Employment of the Work Research Unit. As well as employing some core staff, the unit sponsored projects by academics to explore various aspects of job satisfaction and change. In an overview of the approach adopted by the unit, White (1980), a member of the unit staff, notes that it had as its main aims provision of information to disseminate good practice and, where invited, a role for its staff as consultants to aid projects to improve quality of working life. Although the initial remit had been to help to improve job satisfaction, by the end of the decade the focus was on the broader concept of quality of working life and on improving efficiency. There was also a shift from analysis and diagnosis to a greater focus on the more intractable problem of bringing about change. The unit became wiser about the kind of frameworks and principles for redesign of work settings but even so, the case used by White to illustrate their work failed to bring about significant change. By 1979, it was possible for White (1980: 456) to conclude, ‘Wilson’s hope of seeing well-​defined results from demonstration projects and his belief that these would persuade people in other organizations to copy their experience seem not only less easily realised but even somewhat naïve.’ Experience had greatly reduced expectations.

32   david guest The series of research projects undertaken and led mostly by university academics served mainly to confirm existing problems of work design rather than to show evidence about how they could be resolved. These consisted of a mixture of intended change projects that would be carefully evaluated to provide the demonstration cases and a number that analysed specific issues related to quality of working life though mainly based on features of work redesign. Two examples presented at the 1979 conference illustrate these approaches. Wallis and Cope (1980) reported two studies attempting to improve the quality of working life in psychiatric hospitals. The first went through the usual stages of diagnosis and collaborative identification and agreement about issues that needed to be addressed. These issues included dissatisfaction with many aspects of the working conditions, a shortage of staff and a lack of clear objectives to inform their work. However, commitment to making changes was lacking and the project was abandoned. The second project tried to learn the lessons of the first and ensured fuller commitment and participation in identifying and seeking to implement changes. As a result, some changes were introduced, although there was no evidence about their impact. The second study by Guest et al. (1980) was concerned with workers’ reactions to change, including job redesign. In one of their studies they asked an open question to identify workers’ accounts of changes at work with a following question asking for an evaluation of the changes identified. It was revealed that some of those who had experienced work redesign did not recall it or did not consider it to be a change worth mentioning. A number of workers were more likely to recall personal changes such as a promotion or development opportunity. In evaluating the changes as positive or negative the relevant dimensions seemed to be whether the changes made the work more or less demanding and more or less interesting. For example, one of the changes had been the introduction of a new computer system; some reacted negatively because they said it made the work more routine and less interesting, while others were positive because it made the job simpler and removed some hassles. Another study (Guest et al. 1978) compared evaluations of the job content among a sample of white-​collar workers doing identical jobs involving interaction with the public. Some reported the job as interesting, varied and challenging because each person they met was different; others said the work was boring because in each case the same procedure had to be adopted and the same forms had to be filled in. What this research indicates is that workers, even those doing the same job, have different orientations to work, in line with the influential findings of Goldthorpe et al. (1968), and these orientations affect their reaction to work and to changes, including changes in work design. Taken together with the analysis by White, these studies highlight the challenges facing any process of change to redesign work and improve the quality of working life, and illustrate why a body such as the Work Research Unit struggled to make any impact. Moreover, in the UK, 1979 saw a significant political change as Margaret Thatcher became prime minister, ushering in a new style of government that had little sympathy for any promotion of workers’ interests or improvements in quality of working life. The Work Research Unit gradually declined and was eventually disbanded.

THE QUALITY OF WORKING LIFE    33 Over the following decade, management interest in improving the quality of working life declined further in the face of a number of new approaches to the management of organizations that were generally perceived by managers to be more relevant and promising for improving organizational performance. These included learning from the growing success of Japanese companies (Pascale and Athos 1981), including in particular an interest in quality management (Lawler et al. 1995) and in lean management (Womack et al. 1990). New ideas about business strategy as a basis for gaining competitive advantage were emerging from business schools (Miles and Snow 1984; Porter 1985). Influenced by the writing and advocacy of Peters and Waterman (1982) as well as academic consultants such as Bennis and Nanus (1985) and Kanter (1983), management became absorbed by ideas about leadership and in particular transformational leadership (Burns 1978). All of these issues seemed to be more concrete, more relevant in addressing management priorities and more actionable than the complexities of socio-​ technical systems thinking and the idea of improving the quality of working life.

Improving the Quality of Working Life Although the analysis so far may appear to be highly pessimistic, there is good evidence that over recent decades the quality of working life for many people in advanced economies has greatly improved. Three main factors help to account for this improvement. The first concerns legislation, the second reflects changes in society and the third is action by organizations, partly in response to these changes. The preceding analysis indicates that the strategies of change used in the 1960s and 1970s to improve the quality of working life had limited success. An action-​centred approach of the sort favoured by the Tavistock Institute and reflected in the Norwegian studies often became restricted to the environment in which it was carried out. The normative re-​educative approach favoured for example in Shell and ICI, where there was an attempt to change values through an organization development (OD) approach, came up against more powerful countervailing forces and was therefore restricted to the small number of sites where there was enthusiastic management support. An empirical rational strategy of the sort advocated by Wilson for the Work Research Unit also failed to have an impact, partly because, as some realistic analysts at Arden House had recognized, the empirical case was not convincing and often not seen as relevant by the key stakeholders. For example, for those in the rapidly expanding services sector, the relevance of cases from manufacturing or mining seemed limited. Some of the changes in societal attitudes that had come to the fore in the 1960s persisted and required action. As noted earlier, issues such as equal pay and safety at work attracted widespread support from management, but too often there was a reluctance to take action. A different strategy of change was required—​what Chin and Benne (1976) refer to as a power-​coercive strategy, taking the form of legislation. The European Economic Community took a lead, reflecting a belief that employment rights are an

34   david guest extension of fundamental human rights. The European Social Charter, which reinforced pan-​European employment rights, was designed to complement the European Charter on Human Rights. There was also an acceptance of the legitimate interests of the social partners with different but potentially complementary interests. There was a desire to provide a level playing field by ensuring at least a floor of rights for workers in every country as Europe sought to move towards closer economic integration. The result was a series of legislative steps to regulate the employment relationship, to support the rights of workers and to improve various features of their quality of working life. In 1975, the European Commission set up The European Foundation for the Improvement of Living and Working Conditions to collect and provide information on progress and to identify areas requiring further attention. Over a number of years a series of European directives introduced a range of legislative requirements, including those obliging organizations to provide equal pay for work of equal value, to ensure equality of opportunity irrespective of gender, race or creed, to ensure safety at work with rights for employees to monitor this safety, to promote a healthy work environment and limit avoidable shift work, to operate works councils, to limit working hours, to protect the rights of temporary agency workers, to ensure maternity leave and to provide the right to request time off to facilitate work–​life balance and carry out family responsibilities. There is extensive evidence that the legislation had a generally positive influence on organizational policy and practice. The necessity of this strategy of change is illustrated, for example, by the Industrial Democracy in Europe (IDE) study, which had shown that without relevant legislation, management was likely to render systems of representative participation generally less effective (IDE 1993). Taken together, this legislative programme, covering a wide range of relevant issues, provided an effective means of promoting the quality of working life broadly conceived. Beyond Europe, there was not the same programme of legislation, although Canada and Australia, as two examples, had also introduced a range of legislation to promote workers’ rights. Even the US, with its preference for a liberal free market economic system had legislation to ensure a minimum wage, to guarantee a safe working environment and to prevent discrimination at work. The UK too, with its longstanding preference for a liberal free market tradition and its antipathy to industrial relations legislation, was forced to bring forward legislation in line with the European directives, but generally did so reluctantly and in as weak a form as possible. Under Margaret Thatcher, it opted out of the Social Chapter and even under the later Labour government, as Dickens and Hall (2009) note, tended to focus on codes of practice rather than substantive law whenever possible, and to justify any legislation on the grounds of its contribution to competitiveness rather than by any reference to employee rights or improving the quality of working life. Even in those countries reluctant to legislate, it was impossible to ignore some important trends within society and industry that were forcing organizations to address the quality of working life. In several cases, these trends had both positive and perverse effects. One such trend was globalization, with international competition encouraging a search for best practice and highlighting a need to attract and retain talented staff.

THE QUALITY OF WORKING LIFE    35 For example, this trend encouraged processes of employee involvement to engage in total quality management and innovation. It also required organizations to demonstrate that they were attractive places to work and to do so they had to demonstrate that they provided a high-​quality work experience. This need sparked an interest in employer branding (Edwards 2010). Indeed, an industry grew up reporting on ‘The 100 Best Companies to Work For’ (Fulmer et al. 2003), and criteria for achieving this accolade included job autonomy, involvement processes, extensive training and development, good pay and working conditions and, usually, a high level of job or at least employment security. However, globalization also led to the growth of off-​shoring to locations where efficiencies could be gained by using cheap labour and ignoring the need to provide a high-​quality work environment. A second development was the growing proportion of women in employment. This shift has encouraged greater employment flexibility, including a growth in part-​time working as well as a focus on work–​life balance. A third trend has been the advances in technology which have facilitated greater autonomy for some workers and greater flexibility of work location. The growth of the service economy has enhanced the opportunity for part-​time and flexible working. It has greatly increased the demand for professional and managerial work, while at the same time also expanding low-​quality service work, for example in retail, the hospitality industry and social care (for discussion of job polarization, see Dwyer et al. in this volume). Underpinning the demand for better quality jobs is the increasingly well-​educated workforce allied to concerns that younger generations of workers place a higher value on interesting work, work–​life balance and an ethical concern for the physical and social environment within which firms operate. Many of these developments are features that were highlighted in the 1970s’ debates about the post-​industrial society (Bell 1974). The anticipated collapse in the number of jobs has not come to pass, although the impact of robotization and digitization may yet come to the fore, not just in terms of eradicating jobs but in changing the quality of those jobs that remain (Warhurst et al. 2017). The changes just outlined required organizations to take steps to attract, motivate and retain staff in what have often been competitive labour markets. One way to address this, and at the same time improve the quality of working life, was through the application of human resource management. Within the business strategy literature, Barney (1991) had argued that the key to competitive advantage was to acquire resources that were valuable, scarce and difficult to imitate. Subsequently, he argued (Barney and Wright 1998) that these were likely to be human resources, and organizations should therefore invest in them. This argument led to an interest in what have been variously termed high-​ performance or high-​commitment human resource systems and associated policies and practices (see Boxall and Purcell in this volume). Although the list of practices often varies (Boselie et al. 2005), they typically include several features designed to ensure a high quality of working life. The evidence shows that effective implementation of these practices is associated with high performance (Jiang et al. 2012). Recent analysis suggests that where these practices are geared towards promoting employee well-​being as a route to high performance, they correspond closely to the conventional criteria for a high quality of working life (Guest 2017). There is little evidence that managements,

36   david guest particularly in countries such as the US and UK, are promoting these practices because they have embraced the values of quality of working life. They are doing so if they believe it makes good business sense. The discourse about the quality of working life has largely disappeared from policy discussions and academic research. However, Grote and Guest (2017) made a recent call to reinvigorate research on quality of working life research as an integrative focus of study. They suggested that while Walton’s list of dimensions of quality of working life still had relevance, subsequent developments pointed to the need to extend it to include opportunities for flexible working and for individual proactivity. They argued furthermore that while the quality of working life, as an integrated focus of interest had slipped off the academic agenda, research on the various topics that encompassed quality of working life had greatly expanded, providing a strong evidence base on which to ground advocacy. One example is the continuous stream of research on work design (Clegg and Spencer 2007; Parker and Wall 1998). However, research on the various topics under the umbrella of quality of working life has become highly fragmented, reflecting pressures within the academic world to publish specialist papers. What is lacking is a coherence that can provide once again a more integrated policy perspective and provide a multi-​ level framework within which it might be developed.

Conclusion: The Future of Quality of Working Life So where does this leave the quality of working life? In a follow-​up to the 1973 Work in America, O’Toole and Lawler (2006) published The New American Workplace providing an explicit consideration of developments since the influential 1973 report. They identified three types of US organization which they described as low-​cost operators, global competitor organizations and high-​involvement companies. The first group, characterized by some retail and fast-​food companies, pays little attention to quality of working life. The other two categories of organization are much more willing to consider features of quality of working life, although they acknowledge that the number that can be characterized as high-​involvement companies is low. What they missed, writing in 2006, was the development of the so-​called gig economy and the rapid growth of zero-​ hour contracts, on-​call contracts and the associated heightened insecurity. Any contemporary analysis of the state of the quality of working life reveals a divide between those workers with knowledge and skills that are in demand and who maintain security by being highly employable and for whom employers feel a need to provide a high quality of working life, and the large proportion of the workforce whose work is often highly routine and/​or insecure. For this second group, most employers are unlikely to see any case for going beyond the legal minimum requirements to provide a high quality of work. The limits of any efforts to improve work redesign are reflected in Green’s (2006)

THE QUALITY OF WORKING LIFE    37 analysis showing that levels of autonomy at work have tended to decline, and in the case of the UK this trend is reinforced in a series of studies under the title ‘Unequal Britain at Work’ (Felstead et al. 2015) which highlight the growing divide between the two groups of workers. At the outset, the genesis of the quality of working life movement in Norway was concerned with improving industrial democracy. A feature of the contemporary analysis by Grote and Guest (2017) is the decline of representative democracy and the increased emphasis on individual proactivity, perhaps reflecting a more general societal view of individual responsibility. In the US, one of the factors that provoked the Work in America report was concerns about alienation and particularly dimensions of powerlessness and anomie, caused partly by the dehumanizing experience of routine work. The evidence of recent trends in job content and in the growth of the gig economy, and outside work the growth of more extreme political movements suggest that alienation remains a major feature of contemporary industrial societies. The failure of governments and more particularly organizations to improve the quality of working lives of large swathes of the workforce undoubtedly helps to explain this alienation.

References Barney, J. (1991), ‘Firm Resources and Sustained Competitive Advantage’, Journal of Management, 17/​1, 99–​120. Barney, J., and Wright, P. (1998), ‘On Becoming a Strategic Partner: Examining the Role of Human Resources in Gaining Competitive Advantage’, Human Resource Management, 37/​ 1, 31–​46. Bell, D. (1974), The Coming of the Post-​Industrial Society (London: Heinemann). Bennis, W., and Nanus, G. (1985), Leaders (New York: Harper & Row). Blackler, F., and Brown, C. (1978), Job Redesign and Management Control (Farnborough: Saxon House). Blackler, F., and Brown, C. (1980), Whatever Happened to Shell’s New Philosophy of Management? (Farnborough: Saxon House). Blauner, R. (1964), Alienation and Freedom (Chicago: Chicago University Press). Bolweg, J. (1976), Job Design and Industrial Democracy (Leiden: Martinus Nijhoff). Boselie, P., Dietz, G., and Boon, C. (2005), ‘Commonalities and Contradictions in HRM and Performance Research’, Human Resource Management Journal, 15/​3, 67–​94. Boxall, P., and Purcell, J. (2022), ‘Human Resource Management and Job Quality’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 505–521. Brannen, P., Batstone, E., Fatchett, D., and White, P. (1976), The Worker Directors (London: Hutchinson). Burns, J. (1978), Leadership (New York: Harper & Row). Cherns, A. (1976), ‘The Principles of Sociotechnical Design’, Human Relations, 29/​8, 783–​792. Chin, R., and Benne, K. (1976), ‘General Strategies for Effecting Change in Organizations’, in W. Bennis, K. Benne, R. Chin, and K. Corey, eds., The Planning of Change. 3rd edn. (New York: Holt, Rinehart and Winston), 22–​45.

38   david guest Clegg, C. and Spencer, C. (2007), ‘A Circular and Dynamic Model of the Process of Job Design’, Journal of Occupational and Organizational Psychology, 80/​2, 321–​339. Davis, L. (1957). ‘Toward a Theory of Job Design’, Journal of Industrial Engineering, 8, 305–​309. Davis, L., and Cherns, A. (1975), The Quality of Working Life. Vol. 1 (New York: The Free Press). Dickens, L., and Hall, M. (2009), ‘Legal Regulation and the Changing Workplace’, in W. Brown, A. Bryson, J. Forth, and K. Whitfield, eds., The Evolution of the Modern Workplace (Cambridge: Cambridge University Press), 332–​352. Duncan, K., Gruneberg, M., and Wallis, D. (eds) (1980), Changes in Working Life (Chichester: Wiley). Edwards, M. (2010), ‘An Integrative Review of Employee Branding and OB Theory’, Personnel Review, 56, 965–​993. Emery, F., and Thorsrud, E. (1969), Form and Content in Industrial Democracy (London: Tavistock Publications). Emery, F., and Thorsrud, E. (1976), Democracy at Work (Leiden: Martinus Nijhoff). Emery, F., and Trist, E. (1968), ‘Socio-​Technical Systems’, in F. Emery, ed., Systems Thinking (Harmondsworth: Penguin), 281–​296. Felstead, A., Gallie, D., and Green, F. (eds.) (2015). Unequal Britain at Work (Oxford: Oxford University Press). Fulmer, I., Gerhart, B., and Scott, K. (2003), ‘Are the 100 Best Better? An Empirical Investigation of the Relationship between Being a “Great Place to Work” and Firm Performance’, Personnel Psychology, 56/​4, 965–​993. Goldthorpe, J., Lockwood, D., Bechhofer, F., and Platt, J. (1968), The Affluent Worker: Industrial Attitudes and Behaviour (Cambridge: Cambridge University Press). Green, F. (2006), Demanding Work (Princeton: Princeton University Press). Grote, G., and Guest, D. (2017), ‘The Case for Reinvigorating Quality of Working Life Research’, Human Relations, 70/​2, 149–​167. Guest, D. (2917), ‘Human Resource Management and Employee Well-​Being: Towards a New Analytic Framework’, Human Resource Management Journal, 27/​1, 22–​38. Guest, D., Williams, R., and Dewe, P. (1978), ‘Job Design and the Psychology of Boredom’, Occasional Paper No 13, Work Research Unit, London. Hampson, I., and Sandberg, Å. (2022), ‘The Swedish Contribution to Job Quality’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 41–62. Guest, D., Williams, R., and Dewe, P. (1980), ‘Workers’ Perceptions of Changes Affecting the Quality of Working Life’, in K. Duncan, M. Gruneberg, and D. Wallis, eds., Changes in Working Life (Chichester: Wiley), 499–​519. Herbst, P. G. (1976), Alternatives to Hierarchy (Leiden: Martinus Nijhoff). Hill, P. (1971), Towards a New Philosophy of Management (Epping: Gower Press). House, R., and Wigdor, L. (1967), ‘Herzberg’s Dual Factor Theory of Job Satisfaction and Motivation: A Review of Evidence and a Criticism’, Personnel Psychology, 20, 369–​389. Hughes, J., and Gregory, D. (1974), ‘Richer Jobs for Workers’, New Society, 14, 386–​387. IDE (1993), Industrial Democracy in Europe Revisited (Oxford: Oxford University Press). Jiang, K., Lepak, D., Hu, J., and Baer, J. (2012), ‘How Does Human Resource Management Influence Organizational Outcomes? A Meta-​ Analytic Investigation into Mediating Mechanisms’, Academy of Management Journal, 55/​6, 1264–​1294. Kanter, R. (1983), The Change Masters (New York: Simon and Schuster).

THE QUALITY OF WORKING LIFE    39 Kelly, J. (1978), ‘A Reappraisal of Sociotechnical Systems Theory’, Human Relations, 31/​12, 1069–​1099. King, N. (1970), ‘Clarification and Evaluation of the Two-​Factor Theory of Job Satisfaction’, Psychological Bulletin, 74, 18–​31. Lawler, E., Mohrman, S., and Ledford, G. (1995), Creating High-​Performance Organizations (San Francisco: Jossey-​Bass). Lewin, K., Lippitt, R., and White, R. (1939), ‘Patterns of Aggressive Behaviour in Experimentally Created Social Climates’, Journal of Social Psychology, 10, 271–​299. Miles, R., and Snow, C. (1984), ‘Designing Strategic Human Resource Systems’, Organizational Dynamics, 12/​2, 36–​52. O’Toole, J. (1973), Work in America. Report of a Special Task Force to the Secretary of Health, Education and Welfare (Cambridge, Mass: MIT Press). O’Toole, J. & Lawler, E. (2006). The New American Workplace. (New York: Palgrave Macmillan). Parker, S., and Wall, T. (1998), Job and Work Design: Organizing Work to Promote Well-​being and Effectiveness (Thousand Oaks, CA: Sage). Pascale, R., and Athos A. (1981), The Art of Japanese Management (New York: Simon & Schuster). Paul, W., Robertson, K., and Herzberg, F. (1969), ‘Job Enrichment Pays Off ’, Harvard Business Review, 47, 61–​78. Peters, T., and Waterman, R. (1982), In Search of Excellence (New York: Harper & Row). Pettigrew, A. (1985), The Awakening Giant: Continuity and Change in ICI (Oxford: Basil Blackwell). Philips (no date). Work Restructuring: A Summary of Experiments at Philips –​1963 –​1968. (The Silver Book) (Eindhoven: Philips). Philips (1975), Participation (Eindhoven: Philips). Porter, M. (1985), Competitive Advantage (New York: Free Press). Rice, A. (1958), Productivity and Social Organization (London: Tavistock). Roeber, J. (1975), Social Change at Work (London: Duckworth). Lupton, T., and Tanner, I. (1980), ‘Work Design in Europe’, in Duncan et al., eds., Changes in Working Life (Chichester: Wiley), 217–​236. Trist, E., and Bamforth, K. (1951), ‘Some Social and Psychological Consequences of the Longwall Method of Coal Getting’, Human Relations, 4, 21–​32. Trist, E., Higgin, G., Murray, H., and Pollock, A. (1963), Organizational Choice: Capabilities of Groups at the Coalface Under Changing Technologies (London: Tavistock). Trist, E., Susman, G., and Brown, G. (1977), ‘An Experiment in Autonomous Working in an American Underground Coal Mine’, Human Relations, 30/​3, 201–​236. Van Assen, A., and Wester, P. (1980), ‘Designing Meaningful Jobs: A Comparative Analysis of Organizational Design Practices’, in K. Duncan, M. Gruneberg, and D. Wallis, eds., Changes in Working Life (Chichester: Wiley), 237–​252. Wallis, D., and Cope, D. (1980), ‘Pay-​Off Conditions for Organizational Change in Hospitals’, in K. Duncan, M. Gruneberg, and D. Wallis, eds., Changes in Working Life (Chichester: Wiley), 459–​480. Walton, R. (1974), ‘Improving the Quality of Working Life’, Harvard Business Review, May/​ June, 12–​16. Walton, R. (1975), ‘Criteria for Quality of Working Life’, in E. Davis and A. B. Cherns, eds., Quality of Working Life Vol.1 (New York: The Free Press), 91–​104.

40   david guest Warhurst, C., Mathieu, C., and Wright, S. (2017), ‘Workplace Innovation and the Quality of Working Life in an Age of Uberisation’, in P. Oeij, F. Pot, and D. Rus, eds., Workplace Innovation: Theory, Research and Practice (Berlin: Springer). White, G. (1980), ‘Job Satisfaction and Motivation –​The Development of Practical Strategies for their Enhancement’, in K. Duncan et al., eds., Changes in Working Life (Chichester: Wiley), 443–​458. Wilson, N. A. B. (1973), ‘On The Quality of Working Life’, Manpower Papers No.7, Department of Employment. (London: H.M.S.O.). Womack, J., Jones, D., and Roos, D. (1990), The Machine That Changed the World (New York: Simon and Schuster).

CHAPTER 2

THE SWEDI SH C ONTRIBU TI ON TO JOB QUAL I T Y ian Hampson and åke sandberg

Introduction The recent flowering of job quality (JQ) literature makes surprisingly little reference to the Swedish contribution. Sweden produced a disproportionately large share of the world’s ‘work environment’ research in particular areas related to JQ and social and organizational aspects of work. Comparative research has placed Sweden at or near the top of OECD countries on key job quality indicators (Gallie 2003, 2007; OECD 2018; Sandberg 2019). Research institutions, such as the Swedish Centre for Working Life (Arbetslivscentrum) (from 1996 the National Institute for Working Life (NIWL, or Arbetslivsinstitutet), attracted international researchers during a ‘golden age’ of workplace research and reform from the mid-​1970s to the late 1980s. (The stagnation after 2009 has been interpreted as an effect of the Arbetslivsinstitutet’s gradual cutbacks and then closure in 2007 (SWEA 2015: 9–​10).) Redesigned car assembly in the Volvo Cars assembly plant at Uddevalla offered alternatives to mass production and even to lean production (Berggren 1992; Sandberg 1995; Boyer and Freyssenet 2000), before being overtaken by increasing neo-​liberal influence on Swedish work–​life and labour market policy as well as related research (Sandberg 2013a; Håkansta 2014b). Even so, we argue, exploring the Swedish experience of progressive work–​life reform yields lessons for JQ. A definitive list of JQ’s dimensions is elusive (Findlay et al. 2013, 2017). Grote and Guest (2017: 151, 156) suggest returning to Walton’s eight categories of Quality of Work Life (QWL)—​adequate and fair compensation, safe and healthy work environment, development of human capacities, growth and security, social integration, constitutionalism,

42    ian Hampson and åke sandberg total life space, and social relevance. In addition, they propose ‘individual proactivity’ and ‘flexible working’. Frege and Godard (2014: 943) have similarly proposed that JQ is ‘the attainment of civic principles at work’—​freedom, equality, fairness, justice, security, dignity, fulfilment and democracy. These principles have a distinctly Swedish and Scandinavian ring about them and they echo previous work by Emery and Thorsrud (1976). This duo of researchers (Australian and Norwegian) emphasize what they call the ‘psychological job demands’ of work. This is similar to the contemporaneous and influential ‘job characteristics’ model of Hackman and Oldham (1976), that links characteristics of the work (skill variety, task variety, task significance, autonomy and feedback) to the psychological state of the worker, in particular their motivation. Emery and Thorsrud (1976: 103–​105) also identify how workers’ psychological states are affected by work characteristics, like optimum length of work cycle, interlocking tasks and job rotation, and worker role in setting quality standards and in the design of work—​along with provision of feedback; an identifiable contribution to the job that contributes to the utility of the product for consumers; and a career path that may or may not involve promotion to management. Abrahamsson and Johansson (2008: 7, our translation) also emphasize that the worker should be able to learn and develop at work, as well as to make decisions ‘at least within a defined area that the individual can call his or her own’. The overall job should entail some respect and an identifiable connection between work activity and something considered ‘useful and valuable’ in the outside world. This is in fact a ‘normative theory of how good work should be’. Conflict over control of the firm in the 1970s–​80s set the scene for the Swedish contribution to JQ. Influential Swedish social democratic intellectuals and strategists, like Ernst Wigforss, had argued that industrial democracy was a logical extension of political democracy, to be pursued alongside economic democracy (Sandberg 1992: 57; Higgins and Dow 2013). Ideals of work–​life democratization, including ‘quality in work and product’, inspired action research, in which workers, unions and researchers cooperated to shape organizational and technical change. Employers supported work reform and improvements in job quality, but not industrial democracy, much less economic democracy. Unions sought to extend their workplace power through ‘codetermination’ legislation and action research. Government research policy was a strong influence. These developments reached a high point in the iconic and controversial Uddevalla Volvo car assembly (Ehn 1988; Sandberg 1992, 1995; Martin 1987). Behind this legislation lay several strikes, including a major one in the northern LKAB iron ore mines, which drew the attention of a leading group of sociologists of work to such issues as human values at work, employee influence on work, and opposition to detailed work control (Dahlström et al. 1971). Our first section begins by sketching the ‘Swedish Model’, and Sweden’s distinctive ‘management style’. We then argue that the Swedish contribution to JQ arose out of a class-​based contest over the control of the firm. The second section explores how union strategy problematized the control of the firm. Unions prevailed upon their labour movement partners, the Social Democrats, to legislate power-​sharing in the firm, in part

THE SWEDISH CONTRIBUTION TO JOB QUALITY     43 by making the latter an object of government research policy. We explore the ‘golden age’ of working life research in Sweden in the 1970s and 1980s in the third section. Union-​oriented ‘action research’ encouraged skill utilization and participation. At the same time workers’ capacity to influence technological and organizational change was investigated in a strategy also including researchers’ independent theoretical reflection. The whole approach was called ‘praxis’ or interactive research (Sandberg 1985; Aagard-​ Nielsen and Svensson 2006). Union strategies to increase power in the firm ultimately collided with the advance of neo-​liberalism, and associated shifts in government policy. The ‘lessons’, if that is an appropriate term, are about how JQ entails increasing workplace participation, and how this collides with capitalist control relations.

The Political–​E conomic Foundations of Work–​L ife Reform in Sweden: The Swedish Model and ‘Management Style’ In seeking an explanation for the flowering of progressive work–​life reform and research in Sweden, we begin by identifying historical and structural features of the Swedish political economy, known as the ‘Swedish Model’. A defining feature is the quest to reconcile industrial transformation and productivity with welfare, job quality (the good work) and equality. Swedish ‘management style’ was seen as innovative, participative, value-​based and ‘visionary’, but now increasingly follows international neoliberal trends towards market-​related workplace control (Movitz and Sandberg 2013: 44, 53–​54).

The Swedish Model and Swedish Management Style Swedish political economic history suggests JQ, welfare, equality and individual autonomy can be consistent with productivity, innovation and competitiveness (Movitz and Sandberg 2013: 38). For much of the twentieth century, during a long period of social democratic incumbency, Sweden enjoyed strong economic performance alongside social progress. Aspects of the ‘Swedish model’ included universal education and healthcare; a relatively low gender pay gap; high female labour force participation; high union density; cooperative management of industrial relations through highly organized representative bodies; union influence over aspects of public policy (including industrial policy); and the universalistic social democratic welfare state (Kjellberg 1998). Gallie (2003) found that employees in Sweden and Denmark enjoy higher JQ than OECD comparators. Sweden scored highest on indicators of quality of work task, security against dismissal and employee participation in decision-​making (also see Gallie 2007).

44    ian Hampson and åke sandberg Recently, the OECD (2019) has ranked Sweden fourth for job quality, behind three other Nordic countries. Reflecting high-​quality work organization, Sweden is in the OECD’s bottom quartile for job strain (OECD 2018). Sweden also has the fourth lowest gender pay gap, as a result of decades of gender-​friendly policies. It has the third highest employment rate in the OECD, while the employment gap between disadvantaged groups and prime-​age men is the second lowest in the OECD (after Iceland) despite the challenge of integrating large numbers of recent migrants (OECD 2018). We agree with Findlay et al. (2017: 6) that at a national level, ‘strong institutional environments’ shape JQ. The varieties of capitalism literature identifies Sweden as a ‘coordinated’ political economy (Hall and Soskice 2001), characterized by negotiations between highly centralized peak bodies of employers and unions. Katzenstein (1985) has argued that ‘cooperation’ and ‘consensus’ naturally arise between political interests strong enough to greatly damage each other. Other writers emphasize the class-​based nature of Swedish politics, characterized by a ‘democratic class struggle’ and political mobilization around socialist aims (Korpi 1978; Higgins 1980, 1985). The socialist origins of labour movements are too often (in our view) written out of history, and in the Swedish case they have been important to understand the strategic manoeuvring of the labour movement and the employers. This is notwithstanding that even ‘social’ versions of coordinated economies have recently undergone some degree of neo-​liberal transformation and union weakening, and so we may talk of varieties of liberalization (Streeck 2017) since some are more prone to liberalization than others (Thelen 2014; Baccaro and Howell 2017). A feature of Swedish political institutions has been the close links between the Social Democratic Party (Sveriges socialdemokratiska arbetareparti, SAP), formed in 1889, and the union peak organization, LO (Landsorganisation i Sverige) in turn formed in 1898 on the initiative of the SAP. Sweden’s employers established their central organization (Svenska Arbetsgivareföreningen, SAF, today Svenskt Näringsliv, SN) in 1902, and bitter industrial disputes followed. In the so-​called December Compromise of 1906, SAF recognized the unions’ rights to exist and to organize and collectively bargain, while LO accepted the employers’ rights to hire and fire, and to organize work. According to the so-​called ‘paragraph 23’ (eventually changing to paragraph 32) of SAF’s bylaws, SAF would not permit any member organization to sign any agreement with a union that made concessions on basic managerial prerogatives. This would become a key issue in the politics of participation in the 1970s (Higgins 1986; Victorin 1979: 113–​4; Bruhn et al. 2013). The depression years were again marked by labour market conflict, and in the 1938, ‘basic agreement’ (named for the seaside town Saltsjöbaden where it was signed) LO and SAF reaffirmed the 1906 compromise but extended its terms (Martin 1992). Labour and capital would cooperate in the pursuit of economic growth via technical change, while capital retained its workplace prerogatives—​to hire and fire and organize work. Labour retained the right to organize and bargain but would also pursue political action (via its links to the SAP) to build a welfare society (Korpi 1978; Sandberg et al. 1992: 40–​41).

THE SWEDISH CONTRIBUTION TO JOB QUALITY     45 ‘Joint regulation’ of the labour market saw the state recede into the background of industrial relations. Highly centralized political exchanges between SAF and LO set wages and other policy parameters, while white-​collar and professional union peak bodies during the 1980s bargained outside of LO’s influence. Unions supported industrial change, driving the latter through their wages policy (see later section on ‘Influences on Work–​Life Reform’), and eventually problematizing control of the firm (Kjellberg 1998; Bruhn et al. 2013). Contemporary writers emphasize tension between JQ and efficiency (e.g. Grote and Guest 2017: 153). Yet historically, this opposition has not been as stark in Sweden. The positive side of Swedish ‘value-​based and visionary management’, accepts employee participation, autonomous work groups, union influence, open dialogue, consensus and informality as Movitz and Sandberg (2013; 44, 53–​54) argue. On the other hand, in recent times, management increasingly couples work to the market, to customer interaction, tough evaluation of projects and a general pressure to be ‘better’. Work may indeed be ‘without boundaries’ (Allvin et al 2011) and loosely regulated, with no detailed prescriptions as to how, where and when to carry it out, but this now often comes with tough market-​related measurement of the final outcome linked to compensation, negating any real autonomy in the immediate work situation (Movitz and Sandberg 2019). This can result in stress, burnout and depression, the fastest growing work-​related illnesses today in Sweden, especially among women. For the immediate work situation, we may apply the Karasek-​Theorell model (1990) where ‘active’ jobs with high demands and high worker control stand out as good jobs, although those jobs can easily turn out to be not so good, and result in negative stress (‘strain’, in Karasek and Theorell’s nomenclature) when subject to an overall managerial and market-​related pressure for greater work volumes and continuously improved performance. Swedish employers historically sought to balance JQ and efficiency. The SAF manifesto for ‘New Factories’ in Sweden (Agurén and Edgren 1980: 8) identified ‘two goals—​ more efficient operations and better jobs for individuals’. Similarly, as one of the most significant employer advocates for work reorganization, Pehr G. Gyllenhammar, the CEO of Volvo, put it: Manufacturing work must be organized so that it requires knowledge and initiative, includes different types of tasks with varying content and difficulty, offers variation, gives employees influence and responsibility through participation in the planning, organization and follow-​up of the work, and meets the employees’ need to learn and grow, and of a sense of community and belonging. (Volvo 1974: 11, quoted in Boglind 2013: 191)

For Gyllenhammar, employee participation was one of the defining features of the Swedish approach to JQ. Employers actively promoted it—​until it clashed with management authority and melded with union demands for the democratization of work.

46    ian Hampson and åke sandberg

Swedish Work–​Life Reform: Participation and Democracy The Swedish work–​life reform movement built on the socio-​technical reaction to Taylorism, and the Tavistock studies of longwall mining in English coalmines. This work stressed that group work and shared responsibility were superior to conventional Taylorist methods in terms of productivity as well as JQ. Small groups with a high level of independence and autonomy, with enriched jobs and possibilities for learning performed at a superior level than those designed under Taylorist and individualist work organization principles. On this view, it was possible to ‘jointly optimize’ social and technical work requirements (Cole 1989). Working in such arrangements should produce greater degrees of job satisfaction due to the greater ‘sense of control over their everyday activities on the shopfloor’ (Cole 1989: 19). Accordingly, SAF undertook a ‘study mission’ to Norway, supported the translation into Swedish of Fred Emery’s and Einar Thorsrud’s (1969) work about industrial democracy inspired by the Tavistock studies, and attempted to ‘kick off ’ work–​life reforms by sponsoring the so-​called ‘Hallelujah conference’ in 1969, featuring Thorsrud as keynote speaker. It set up a new technical department, to examine work organization issues and make proposals (Cole 1989: 252–​256). According to SAF (today SN), assembly lines’ inherent limitations could be solved by work reorganization (Agurén and Edgren 1980). Assembly lines were inflexible, consisting of short-​c ycle workstations tightly linked together. Unless workers took the same time to complete a cycle (as in a perfectly balanced line), there would be unproductive ‘idle’ time. Workers were tied to the pace of the line, and could not work ahead to create small, personal breaks. The solution was to organize work in production teams. Line balancing would happen automatically as workers helped each other and shifted between tasks within groups. Job content could expand into longer task cycles and a greater degree of freedom from work pacing. Each individual worker could take responsibility for a larger component of production, and workers could distribute work among themselves (Agurén and Edgren 1980). ‘Self-​managing teams’ or ‘semi-​autonomous work groups’ could ‘make their own decisions regarding work allocation, recruitment, planning, budgeting, production, quality, maintenance and purchasing’. They could take responsibility for the overall organization of work. From management’s point of view, it was important to limit such initiatives to engendering a ‘feeling’ of participation (aka empowerment) rather than allowing substantive ‘power-​centered’ participation (aka the democratization of work) (Cole 1989: 19, 31). The Swedish Employers’ Federation (SAF) accordingly advocated the ‘coordinated independence of small systems’, or partially autonomous units coordinated by management (Agurén and Edgren 1980). Emery and Thorsrud (1969: vii), however, argued that involvement of workers in the decisions that affect them would solve management coordination problems, but it could also provide opportunities for employee development by work–​life participation (Emery and Thorsrud 1976). Trade unions wanted increasing control, coordination by workers’ elected representatives, and the

THE SWEDISH CONTRIBUTION TO JOB QUALITY     47 democratization of working life (Sandberg 1992; Movitz and Sandberg 2013: 58; Higgins and Dow 2013). This resonated with the interests of leading labour movement intellectuals such as Ernst Wigforss, treasurer in the SAP government (whose expansionist economic policy contributed to the SAP victory in 1932 and ushered in unbroken SAP rule till 1976). Like the English Guild Socialists, Wigforss saw the workplace as an arena for workers’ education and the development of political efficacy through participation in workplace decision-​making (Higgins and Dow 2013; also see Pateman 1977). Wigforss rejected the idea that there was necessarily a ‘great tradeoff ’ between equity and efficiency. Indeed, he argued that authoritarian relations at work wasted human capacity and denied the possibility that workers might increase workplace efficiency out of their own knowledge of the production process (Higgins and Dow 2013: 105–​107). Ultimately, the debate over the democratization of work raised far-​reaching issues of worker influence on product quality and product choice and, in relation to the latter, the ‘social usefulness’ of production (Cooley 1981; Ehn 1988).

Influences on Work–​L ife Reform: Union Strategy, Co-​Determination Legislation, and Government Research Policy Union strategy has too often limited itself to ‘distribution issues’ (like wages and conditions), but in Sweden ‘production issues’ like work design and decision-​making power at work—​an essential aspect of JQ—​increasingly came within its purview. This section traces the development of union strategy through wages policy under the Rehn-​Meidner labour market model, and unions’ response to the employers’ reforms initiatives. As to distribution issues, the union wages policy was solidaristic. It drove industrial restructuring by narrowing wages differentials across the labour market, following principles of ‘equal pay for equal work’ irrespective of profitability of the individual company, all underpinned by active labour market policies including training. As to production issues, unions sought legislative reform to improve worker’s bargaining position at work. Government research policy was a big influence in this mix, as (during the heyday of Swedish progressive work–​life research), it promoted research into workplace power-​sharing arrangements and democratization in worklife and the economy.

Wages Policy: Economic and Industrial Democracy? In the 1960s, LO strategy, articulated in the influential Rehn-​Meidner model (agreed at the LO conference of 1951), sought to modernize industry and shape labour movement

48    ian Hampson and åke sandberg structure, cohesion and political capacities (Meidner 1980). ‘Solidaristic wages policy’ aimed to narrow wage dispersion by restraining wages at the upper end of the working class, while boosting wages at the lower. This put pressure on inefficient firms, on the rationale that if a firm could not afford to pay ‘fair’ wages, it should go out of business and its resources be deployed elsewhere in the economy. Efficient companies could make extra profits, and grow. This would shift the whole industrial structure towards a high-​ wage and high-​skill model. Displaced workers would be offered income support and retraining (Meidner 1994; Martin 1992). Solidaristic wage policy pursued in a full employment economy also stimulated the movement for workplace reform. In the manufacturing sector, particularly in the automotive assembly, employers were faced with labour shortages due to low JQ in their factories, and because the solidaristic wages policy restricted their capacity to compensate workers with high wages (Higgins 1985; Martin 1992). Thus, in the 1960s the employers’ peak body began to examine the possibilities of work–​life reform to improve JQ as a means to compensate for lower pay, while solving line-​balancing problems in production. The Rehn-​Meidner model’s implementation raised strategic dilemmas resulting from the fact that some unions were stronger than others and could have gained wage increases that exceeded the centrally determined amounts (so-​called ‘wages drift’). If they had pressed these claims, weaker unions would have fallen behind, wages dispersion would have increased, and the overall cohesion of the movement would have suffered. On the other hand, if the stronger unions did not claim the potential wage increases (consistent with the solidaristic wages policy), profits would have been retained by employers, permitting conspicuous consumption and unproductive speculative investment. If its authority over its constituents was to remain intact, the union leadership had to persuade the stronger sections of the union membership to restrain their wage demands. In other words, there had to be a quid pro quo for wage restraint (Meidner 1980, 1994; Higgins 1985; Higgins and Dow 2013). The proposed solution to the strategic dilemma was for ‘excess’ profits to be paid into collectively controlled wage-​earner funds that could finance economic expansion and advance the democratization of capital (Higgins 1985; Higgins and Dow 2013). In the resulting controversy, unions, and to an extent previously hardly seen employers, mobilized in the streets; the first such employer demonstration gathered about 100,000 opponents of the funds. The political right lobbied in the media (which they powerfully influenced). Prior to national elections, the social democrats withdrew the radical proposal. The ideological backlash against economic democracy—​what the right characterized as ‘creeping socialism’—​was also a setback for the union movement, although, according to one commentator (Ingemar Göransson, former senior researcher at Metall and LO, personal communication) employers’ absolute refusal to share power over investment was balanced, in a ‘spirit of compromise’, by willingness to negotiate over improvements to JQ. By the mid-​1980s the economic democracy strategy gave way to one more centred on the workplace, with a shift from ‘solidaristic wages to solidaristic work’ (Mahon

THE SWEDISH CONTRIBUTION TO JOB QUALITY     49 1994). The strategy was developed by the Metal Workers’ Union and articulated in its seminal document, which announced nine principles of Det goda arbetet (The Good Work) (Metall 1985; Johansson and Abrahamsson 2009: 776). These were: Job security; A fair share of production earnings; Co-​determination in the company; Work organization that emphasized cooperation; Professional know-​how in all work; Training to be a normal part of work; Working hours based on social demands; Equality at the workplace; and a Working environment without risk to health and safety. In summary, ‘good work’ involved skills upgrading within a given position linked to wage development, multiskilling, ‘solidaristic’ teamwork and shared decision-​making (Mahon 1994). As was the case in solidaristic wages policy, ‘good work’ should be equally distributed across the working class. Yet as unemployment grew, and as Johansson and Abrahamsson (2009) argue, concerns about ‘good work’ (JQ) were overtaken by concerns about the quantity of employment.

The Co-​Determination Legislative Reforms Union cooperation with work reorganization posed dangers. If work reorganization harmed working conditions and undermined occupational health and safety, worker trust in their union would be undermined. For unions to exercise a truly protective role and not acquiesce to employer-​controlled participation, truly joint decision-​making power, including the power of veto, was needed. This meant challenging paragraph 32, and so much was agreed at the 1971 LO Congress (Higgins 1986: 255). Accordingly, in the 1970s, the sympathetic Social Democratic government legislated to support union influence on management decision-​making, signalling a break with the tradition of non-​ intervention in the labour market. In 1976, the government enacted the Co-​Determination Act. The Swedish term Medbestämmandelagen (abbreviated to MBL) translates to English as ‘co-​ determination act’. Expectations of MBL were high. Olof Palme hailed it as ‘perhaps the greatest democratic reform in Sweden since the introduction of universal suffrage’ (ACTU/​TDC 1987: 141). On the other hand, left critics referred to it as a futile attempt to ‘reconcile the irreconcilable’—​‘. . . to straddle two irreconcilable positions’—​and as little more than an obligation on management to ‘honk the horn’ before running over workers (Higgins 1986: 259–​260). Victorin (1979: 118) emphasizes that union influence was overstated and argues the Act was not designed ‘to change the fundamental principle of the employers’ right to manage’. While true, MBL and other related legislation did tilt the balance of power toward labour, as it gave access to information and the right, if not to exercise final power, at least to discuss and to get the employers’ arguments on the table. The Act made it an employer’s duty to bargain over joint regulation if the union demanded. The union could make demands that infringed managerial prerogative, although of course employers did not have to agree to them. Second, the employer had a ‘primary duty to negotiate regarding essential business changes’—​including

50    ian Hampson and åke sandberg ‘important alterations of business activity or of work or employment conditions. But ‘the right to make decisions remains with the employer; the unions’ negotiating rights are limited to requiring the employer to give reasons for his actions and to listen to the arguments of the union’. Third, the Act required the employer to keep the union informed of any impending new developments. Fourth, it gave primacy to the union’s interpretation in disputes. Fifth, the union could veto sub-​contracting where there was reason to suspect evasion of an employer’s obligations regarding, for example, social insurance and collective agreements. These changes were a big step forward for workers’ and unions’ rights at the firm, although they were not a fundamental change in power relations (Victorin 1979: 119–​124). In 1976 the Social Democrats lost government just before the MBL Act took effect, in 1977. This emboldened SAF, which began to withdraw from the centralized bargaining and other ‘corporatist’ arrangements. Employers refused to negotiate on the issues over which LO sought shared decision-​making, including personnel policy and administration; work organization and management; information about management matters; the introduction of computers and access to databases; employee consultants; and education in paid working hours (ACTU/​TDC 1987: 142). After the enactment of the MBL law, co-​determination agreements followed. The agreements in general softened the language of negotiation, most so in the private sector where in 1982 a co-​determination agreement was concluded that ‘integrated negotiations’ into bipartite groups or committees and even into the day-​to-​day activities in the organization.

The Influence of Research Policy As Håkansta (2014a, b: 3, 10–​12) has argued, the prominent role the social partners enjoyed under the Swedish model influenced government support for work–​life research. This resonated with trends in science policy. In general, scientific research tends to move between ‘mode 1’ research, which is characterized by an emphasis on political neutrality and scientific quality (as defined by scientists) and ‘mode 2’ research which is more concerned with the relevance and ‘application’ of science to ‘real problems’ that may have a political dimension. From 1945 to the 1960s Swedish research was predominantly applied, but oriented to defence and Cold War concerns rather than broader societal ones. According to Sörlin (2016) from 1960 to 1985 there emerged a strong belief in the political control of research, and this coincided with increasing application of research to ‘mode 2’ ‘real world’ concerns—​of which the quality of work was an important one. JQ was central to the work of the PA Council which SAF established in 1952, to conduct research into psychology and personnel administration, influenced by the Human Relations School. It was chaired by a SAF member, but from 1970 on with union (LO and TCO) representation on the governing board of the council. This

THE SWEDISH CONTRIBUTION TO JOB QUALITY     51 created somewhat more balance between the labour-​market parties by strengthening union representation. An official governmental investigation proposed that the council be transformed to a publicly funded research institute, but at the time it was politically impossible to defend public funding to an employer-​dominated institute (Håkansta 2014b; 13, 18). Instead, in 1972 the government set up the Work Environment Fund (WEF Arbetsmiljöfonden) to fund research into various aspects of the work environment—​ a concept that went beyond worker protection and occupational health and safety (Håkansta 2014b: 22). With trade unions having majority representation on its board, and with financial support from levies on employers as well as government grants, work–​life research flourished. Initially WEF’s mission was to finance research related to the prevention of injury, the improvement of the work environment and worker health and safety. But in 1977 the passage of the Work Environment Act (Arbetsmiljölagen) required it to also support research and training into the implementation of new co-​ determination laws (Håkansta 2014b: 18). Also, in that year the government set up the Swedish Centre for Working Life (Arbetslivscentrum) which received much funding from the WEF. Through the 1990s, the march of neo-​liberalism coincided with, or arguably was coterminous with, the reassertion of emphasis on ‘scientific quality’ in research. Swedish work–​life ‘mode 2’ research was attacked from the right as biased and of poor quality due to the closeness between researchers and unions (Håkansta, 2014a)—​although closeness between management and researchers did not lead to similar allegations but rather to a further growth of the ‘industry research institutes’, known as Rise (Sörlin 2016). Such allegations and the increased emphasis on journal metrics contributed to the decline of progressive work–​life research in Sweden—​and also elsewhere (cf. Hampson and Morgan 2016; see also Tourish et al. 2017). On the WEF managing board, management and labour representatives disagreed about what research to finance. In 1991, the employer representatives left the board as part of a general withdrawal from governmental boards, and instead focussed on informal lobbying. In 1995 the WEF was closed and its education and information activities became part of the new National Institute for Working Life (NIWL) (Arbetslivsinstitutet). The research funding part was undertaken through a new organization under the Labour Ministry, the Council for Working Life Research (Rådet för arbetslivsforskning (Ralf), which, to a higher extent than the WEF, emphasized scientific relevance, but still also financed change-​oriented projects with work–​life relevance. Already in 2001 Ralf was dissolved and the responsibility for work–​life research was split between a new research council for work–​life and social issues (Fas) under the Ministry of Social Affairs (Fas later changed its name to Forte). Fas moved further in the direction of a conventional research council with scientific relevance as the main goal dominating over work–​life relevance. It also emphasized social and health issues, with a focus on the individual, rather than on work organization and labour market relations;

52    ian Hampson and åke sandberg this reflects that Fas and Forte were organized under the Ministry of Health and Social Affairs rather than the Ministry of Employment. The problem-​and change-​oriented research was to be taken over by the innovation agency Vinnova. However, Vinnova returned to a sole focus on business and technological development and did not pursue the work–​life change aspects. (Håkansta, 2014a: 25–​6; Oscarsson 1997, Björkman and Sandberg 2019). This was a fundamental change, as a result of which development-​ oriented work–​life research was starved of public funding. In the same year, the Institute for Working Life Research (which followed after the Swedish Centre for Working Life) together with the Work Environment Institute, were restructured into the National Institute for Working Life (NIWL) (Arbetslivsinstitutet), gradually concentrating more on the labour market and the work environment, including the psychosocial, and less on work organization, industrial relations and ‘action research’ interacting with worker participation, workplace change processes and JQ (cf. Lansbury 2016). This redirection continued and was emphasized when in 2007 the new centre right government closed down the NIWL, heralding the end of the golden age of work–​life reform and research.

The ‘Golden Age’: Research and the Struggle over Control of the Firm The JQ and research debate in Sweden can be periodized in a useful way (Björkman and Sandberg 2019, and Sandberg 2016). From 1940 to 1960, the unions cooperated with employer initiatives aimed at improving job satisfaction, while solving ‘coordination’ and line-​balancing problems. From 1965, there emerged a critique of previous activity, as unions began to raise issues of workplace control. From 1975 to 1985 was the ‘golden age’ of ‘pluralist’ work–​life research—​of employee-​oriented research and also of major and growing union–​researcher cooperation and radical challenges to employer control of the firm. From 1985 onwards there was much research done under large programmes funded by the Work Environment Fund. But as the economy deteriorated and union power declined, unions began to focus less on job quality, and more on job quantity. This period of decline, ironically, coincides with the iconic high point of the Swedish contribution to JQ—​the Uddevalla plant. A unique combination of circumstances facilitated the ‘golden age’ of JQ research. Co-​determination legislation described in a previous section addressed the issue of workplace control, while government policy encouraged research into issues related to JQ, such as technological change, control at the workplace and the role of unions. While unions broadly accepted the need for accelerated technological change to improve overall national productivity and competitiveness, they had concerns about the effects on workers in particular industries and sectors. Employers implemented workplace reform to improve JQ, partly to compensate for their inability to pay extra wages to

THE SWEDISH CONTRIBUTION TO JOB QUALITY     53 compensate for low JQ in tight labour markets, but only as long as it did not compromise overall management control. Because demands for workplace change were often met with claims that work-​ improving changes were impossible due to technological limitations, researchers, unions and workers challenged employer control over technology. This meant not only choosing between already established technologies but also intervening in technology’s design. Employers had a natural advantage in this contest because of their greater resources, including the ability to acquire technical expertise. Unions therefore sought to build their own research and technical capacity to challenge the rule by experts—​an idea pioneered by the leading Norwegian computer scientist Kristen Nygaard (Higgins 1986; Sandberg 1992). In this development of a ‘counter’ expertise, unions and workers combined with researchers in a unique form of ‘interactive research’ (Aagard-​Nielsen and Svensson 2006; Sandberg 1985). To illustrate, two projects described in this section aimed to investigate the conditions for ‘good work’, including ‘quality in work and product’ (i.e. meaningful production) and industrial democracy.

Union Oriented Action Research: Demos and Utopia as Examples As Sandberg (1992: 30–​37, 101, 111) argues, much social science research is oriented towards problems defined by strong actors, such as policy decision makers or senior management. ‘Action research’ is oriented towards managerial and owner interests, or producing a particular kind of change. Much Swedish research was value oriented—​ towards the quality of work–​life and the democratization of work. But the cases presented here, Demos and Utopia, were not at all the only such projects and nor were they exclusively Swedish. There were several at the KTH Royal Institute of Technology, as well as for example the Due project at Aarhus University, Denmark. As well as seeking to influence the choice or implementation of new technology, objects of research included the ‘participatory design’ process itself, and its potential for ‘alternative technology’ the democratization of work. The Demos (Democratic Control and Planning in Working Life) project sought to explore the potential of cooperation between unions and researchers for shaping work organization and technology, building on the pioneering Norwegian NJMF project from the 1970s (Bansler 1989: 94). It also sought to build the capacity of local unions to develop their own organizational alternatives by enhancing cooperation and qualifications among workers. Workers acted as researchers in special, qualified ‘investigative groups’, which included both experienced union members and workers with little experience. The groups studied their own work and enterprises, using researchers as ‘resource persons’, and formulated action plans. In so doing they interacted with national unions, to get support of various kinds, in developing policy and in negotiations with management. Broader participation in reference groups in the form of ‘study circles’ added to

54    ian Hampson and åke sandberg collectivism and encouraged non-​members to join the unions. The union was called upon to develop its own internal competence to meet management in negotiations over the implementation of new technology. In this development, sympathetic external academic researchers undertook supplementary empirical studies and theoretical analysis—​this long-​term interaction was called praxis research or interactive research (Aagard-​Nielsen and Svensson 2006; Sandberg 1985). The Utopia project followed. It was based at the Arbetslivscentrum, in cooperation with the KTH Royal Institute of Technology in Stockholm; with researchers from Aarhus University in Denmark; and with the Nordic Graphic Workers Union. Utopia is an acronym of the Swedish for ‘Training, Technology, and Product in the Quality of Work Perspective’ (Lundin 2005: 2–​3). The broad aim of the project was to extend union and worker influence over new technology to the design stage, to develop participatory design processes, and to analyse, empirically and theoretically, the organizational and societal possibilities and limits to such changes. The Demos project had revealed limitations of local strategies that were essentially defensive in the sense that they tried to influence work organization around existing technology for example ‘hot metal’ type-setting in the printing industry (Ehn 1988; Bansler 1989). The new research aimed to counter the introduction of radically new ‘cold metal’ digital technologies (for text and image processing) that were drastically reducing employment in the printing industry, and deskilling print workers skilled in ‘hot metal’ pagination. Management strategies invoked the presumed inevitability of the introduction of certain forms of technology and organization. Researchers and unions sought to develop organizational and technical alternatives, with support from national and Nordic unions, as well as cooperation with adjacent unions and researchers. They aimed to develop software that could build on workers’ skills and capabilities, rather than follow a deskilling and control logic (Howard 1985; Ehn 1988). ‘Technological choice’ (going beyond the socio-​technical concept of ‘organizational choice’), required imaginative software design around graphic skills, and machine interaction skills. The project proceeded through mutual learning between workers and researchers covering page make-​up, image-​processing, work organization, work environment (ergonomics), person–​machine interface, and intensive training. This was a unique cooperation under favourable preconditions that has been difficult to replicate, although ultimately management resistance and union rivalry prevented the project achieving its full potential (Ehn 1988; Sandberg 1992: 244–​258). Even so, Lundin (2005: 4) judges Utopia to be a ‘seminal’ project on participatory design, which gave rise to the so-​called Scandinavian School of System Development—​ where ‘the users’ participation in system development [has] become a key element’. The project stands as a counter to any view of technological change as ‘inevitable’, ‘autonomous’ and ‘out of control’—​but potentially rather subject to social shaping (Lundin 2005)—​including towards democratic ends. The Demos and Utopia projects were, however, not only a cooperation between unions and researchers, but also an innovative and fruitful one between researchers from different fields like computer science, design, work–​life and organization studies.

THE SWEDISH CONTRIBUTION TO JOB QUALITY     55 Such cooperation was essential to support JQ when work was undergoing ‘computerization’ (as the process was then termed) and has turned out to be not so easy to replicate. Academic pressures have made computer scientists/​designers, and social scientists respectively choose different roads, with the former emphasizing ‘mainstream’ socio-​ technical design and learning aspects, with the latter, and some computer scientists, still emphasizing power differences, and labour and managerial control in digitalization processes in organizations (compare Bødker et al. 2000 and Greenbaum 1996; more recently e.g. Irani and Silberman 2016, Movitz and Sandberg 2013: 59–​60, and in general labour process analyses of work–​life digitalization). The vision of the mentor Kristen Nygaard was to integrate the two, and given the challenges in today’s work–​life such renewed researcher cooperation would be of great importance. It may perhaps again be possible given recent developments in research policy that we addressed in our earlier section on the Influence of Research Policy.

Uddevalla: A High Point of the Swedish Contribution to Job Quality Developments in the Uddevalla car assembly plant represent a high point of the Swedish contribution to JQ, arguably fulfilling the dream of enhancing productivity and work quality at the same time. It also exemplifies how union and worker involvement in production organization encountered limits to the democratization of work. Following the closure of the Uddevalla shipyard, Volvo was enticed into the area with heavy subsidies, partly to provide employment, and partly because it offered the possibility to build a new ‘greenfield’ plant following revolutionary design principles. In these, the principles of ‘The Good Work’ (Metall 1985) played a role, as did targets for female participation and an undersupply on the local labour market of the young, fit males who are the most common auto-​manufacturing workers. The project began in 1985, and the plant was opened in 1989 (Sandberg 1992; 1995; Ellegård 1995). The unions were heavily involved in the planning of the factory layout along with a number of researchers and experts from various disciplines at Gothenburg University, and Chalmers Technical University (Ellegård 1995; Engström and Medbo 1995; Nilsson 1995; Blomquist et al. 2013). As the implementation progressed, workers too became involved. Initially, Volvo’s production engineers put forward a moving assembly line model, although moderated by some ergonomic design principles developed in Volvo’s Kalmar plant. However, the unions in alliance with top Volvo management called for a different production design based on as yet undeveloped technical and pedagogical principles—​including a work cycle not less than 20 minutes (compared with the then existing work cycles in auto assembly which could be as low as one minute). A deciding factor in the early negotiation about design principles was an experiment performed by researchers who disassembled a Volvo car, laid out its parts in a workshop,

56    ian Hampson and åke sandberg and grouped the components into intelligible logical subsystems. An experienced and qualified car industry worker then ascertained how to combine them with assembly in mind and worked out how to assemble the whole car in one long work cycle. The subsystems became the basis for training design. Training modules to support a two-​ hour work cycle were initially set at three-​months’ duration. Completion of the training module delivered a pay rise, based on competence, thus driving competence acquisition. As more workers completed the training modules, the production design moved towards the assembly of whole cars, by autonomous teams, in parallel flow workshop design. Thus, the novel design in Volvo’s Uddevalla factory did away with the moving assembly line, and with it the problems associated with line-​balancing between individual workstations. The resulting so-​called ‘reflective production’ was based on advanced principles of pedagogics, high levels of learning, and worker capacity to deliver continuous improvement (Ellegård 1995). Workers could place all components in the car in the right place as long as they were presented for assembly in the right order. This was ‘meaningful’ work, in the sense that the fully trained worker was able to see the relation between all the parts of the car. On the basis of the learning, workers could suggest changes, not only to the production process, or improvements to product quality, but also modifications to the product (Berggren 1992; Blomgren and Karlsson 1995). This tested management prerogatives and probed the limits of the democratization of work. In the last stages of the Uddevalla plant’s operation there was practically no management hierarchy. Production teams were autonomous. There was no work pacing and workers could take personal breaks. Coordination demanded a management function between the autonomous teams and upper management—​but this person was not called a ‘supervisor’, but an ‘ombudsman’—​indicating that they would play a moderating role and would seek to balance any tensions between legitimate requests coming from ‘below’, as well as from ‘above’ (Sandberg 1992: 90; Sandberg 2013b). Customers could order a car to particular specifications in consultation with workers and watch it being made—​testifying to further boundary-​crossing of normal capitalist authority structures. To this day, opinions are divided on the question of whether the form of production at Uddevalla was superior to the prevailing production models of car assembly—​mass and lean production. By the time the plant was demonstrating high and increasing levels of productivity (including one highly skilled worker who could assemble a whole Volvo in 10 hours—​better than claimed assembly times in even the most productive of the Japanese plants of the day—​Volvo was restructuring its global operations, and it closed the Uddevalla plant. Exactly why is highly controversial. Lean production (Womack et al. 1990) advocates assert it was inferior in technical and productive terms; others beg to differ (Berggren 1992; Sandberg 1995). As Cole and Adler (1995) noted during the debate at the time, the plants were shut down before researchers had the chance to definitively investigate their potential. Since then, lean production has been further developed, and become more flexible and productive.

THE SWEDISH CONTRIBUTION TO JOB QUALITY     57 The debate over Uddevalla’s comparative productive efficiency is also about ideology, control and power (see Noble 1977). One view is that the plant was destined for closure precisely because of its success—​as a production model consistent with the democratization of work and high productivity that might threaten managerial authority. The Uddevalla model influenced ‘postlean’ production systems that emerged in Japan, as car producers sought to moderate the effects of lean production on attraction and retention of their workforce by such ‘innovations’ as reinstating production buffers, lengthening job cycles and limiting hours (Hampson 1999). Yet concerns have subsequently been expressed about a certain ‘idealization’ of the Uddevalla model (Sandberg 2013b). Car production, after all, leaves something to be desired in terms of ‘socially responsible production’ and personal development potential. The autonomy available under the ‘reflective’ production conditions in car assembly may not have been available to the workers who assembled the packs of components that fed into the production process. There is also a lack of empirical research—​of interviews with workers who can tell about what it was actually like to work there. There was a lack of learning beyond and between the individual teams. However, during the last months in operation changes were made to remedy this as a step towards making the plant more productive. In the Uddevalla model there is a need to negotiate the time allotted to the production of each model and variant. As there is no line to set the rhythm of work, this may be a complicated issue. In the model there is a partial reversal of ‘the intellectual division of labour’ in the direction of free cooperation of persons as, for example, in the development of open-​source software, and publications like Wikipedia. Such cooperation can be very productive and, at the same time, threaten established forms of production and authority. Ultimately, the lesson of the Uddevalla case may be about the limits of work democratization within a capitalist society (see several contributions by Freyssenet and others in Sandberg 2013b).

Conclusion: Revival beyond the End of the Golden Age? The Swedish approach to job quality (JQ) linked efficiency with the democratization of work—​that is, with workplace participation and autonomy. It questioned the idea that there was a necessary tension between JQ and efficiency and sought to increase both at the same time. At the theoretical leading edge of the labour movement, such writers as Ernst Wigforss argued that democratic workplace arrangements were a necessity for fully efficient enterprise. Developments in Swedish political economy aligned to produce a ‘golden age’ of work–​life research and reform, in which government research and worklife policy supported academic researchers and active union members to perform research that addressed basic challenges and problems in society, including and in particular how to improve JQ. Swedish intellectuals and unions emphasized workplace

58    ian Hampson and åke sandberg participation as an educative and political strategy and sought to challenge capitalist control not only at the level of the firm, but also at a political and societal level. As this constellation of forces passed, progressive work design encountered limitations deriving from a capitalist economy and neoliberal globalization. The SAP’s previously dominant position in Swedish political life ebbed away, and with it went much of the support for work–​life research and reform. There were corresponding shifts in research policy—​from emphasis on relevance and social usefulness, to rigour, political neutrality, and the increasing metricization of academic culture (cf. Alvesson, Gabriel and Paulsen 2017). As the Uddevalla model reached its high point, rising unemployment and deteriorating economic conditions focused unions’ attention back on employment, making JQ a secondary priority. By 2007, action-​oriented and interactive research into the democratization of work and organization had virtually ceased—​ symbolized by the closure of the Arbetslivsinstitutet. Change-​oriented research guided by the needs of industry and employers, however, remained and was strengthened within Industriforskningsinstituten, later assembled under the umbrella organization of Rise (Sörlin 2016). When it comes to research, the two sides of industry were treated very differently by the state: labour-​oriented research was closed down; employer-​oriented research expanded. More recently, as Sörlin (2016) points out, there are signs of renewal in Swedish research policy, which now emphasizes the need to address fundamental societal challenges as a guiding principle. A new national authority for the work environment commenced operation in the summer of 2018, with its main tasks the evaluation of government policy, knowledge mediation between research and working life, and coordination of international research cooperation. However, the authority has a relatively narrow focus on the individual and the work environment, and downplays the importance of work organisation, governance and power. A next step might be to strengthen work–​life research itself, by offering resources to build new pluridisciplinary research centres and departments at several universities. If, as may be the case, neoliberalism has reached its zenith and is commencing decline, there perhaps will be room again for fruitful multidisciplinary research directed towards important challenges in working life: into design of technology and work; into industrial relations and unions; and into how to redress power inequalities through the democratization of work. On the other hand, tendencies towards further neoliberalization are evident in growing inequalities and the weakening of unions. The outcome is therefore rather open.

References Aagard Nielsen, Kurt., and Svensson, Lennart (eds) (2006), Action and Interactive Research (Mastricht: Shaker Publishing). Abrahamsson, Lena, and Johansson, Jan (2008), ‘Det goda arbetet –​igår, idag och imorgon’, Framtidsfabriken Rapport 2, Kunskapsöversikt, Luleå: Luleå tekniska universitet, Institutionen för Arbetsvetenskap.

THE SWEDISH CONTRIBUTION TO JOB QUALITY     59 Adler, Paul A., and Cole, Robert E. (1995), ‘Designed for Learning: A Tale of Two Auto Plants’ (with a new postscript), in Åke Sandberg, ed., Enriching Production (Aldershot: Avebury), 157–​178. Also in Sloan Management Review, 34/​3, 85–​95. Agurén, Stefan, and Edgren, Jan (1980), New Factories: Job Design through Factory Planning in Sweden (Stockholm: Swedish Employers Federation, Technical Department). Allvin, Michael, Aronsson, Gunnar, Hagström, Tom, Johansson, Gunn, Lundberg, Ulf (2011), Work Without Boundaries –​Psychological Perspectives on the New Working Life (Oxford: Wiley-​Blackwell). Alvesson, Mats, Gabriel, Yannis, and Paulsen, Roland (2017), Return to Meaning: A Social Science with Something to Say (Oxford: Oxford University Press). Baccaro, Lucio, and Howell, Chris (2017), Trajectories of Neoliberal Transformation: European Industrial Relations Since the 1970s (Cambridge: Cambridge University Press). Bansler, Jørgen (1989), ‘Trade Unions and Alternative Technology in Scandinavia’, New Technology, Work and Employment, 4/​2, 92–​107. Berggren, Christian (1992), Alternatives to Lean Production: Work Organisation in the Swedish Auto Industry (Ithaca, New York, ILR Press). Björkman, Torsten, and Sandberg, Åke (2019), ‘Arbetslivsforskningen i Sverige –​en historia om fem perioder’, in Åke Sandberg, ed., Arbete & välfärd (Lund: Studentlitteratur), Ch. 4. Blomgren, Henrik, and Karlsson, Bo (1995), ‘Assembly Skills, Process Engineering and Engineering Design’, in Åke Sandberg, ed., Enriching Production (Aldershot: Avebury), 127–​136. Blomquist, Bo, Engström, Tomas, Jonsson, Dan, and Medbo, Lars (2013), ‘Assembly Systems and Work in the Swedish Automotive Industry’, in Åke Sandberg, ed., Nordic Lights (Stockholm: SNS Förlag), Ch. 6. Boglind, Anders (2013), ‘Volvo and a Swedish Organisation and Management Model’ in Åke Sandberg, ed., Nordic Lights (Stockholm: SNS Förlag), Ch. 5. Boyer, Robert, and Freyssenet, Michel (2000), ‘A New Approach of Productive Models: The World that Changed the Machine’, Industrielle Beziehungen, 7/​4, 385–​413. Bruhn, Anders, Kjellberg, Anders, and Sandberg, Åke (2013), ‘A New World of Work’, in Åke Sandberg, ed., Nordic Lights (Stockholm: SNS Förlag), Ch. 4. Bødker, Susanne, Ehn, Pelle, Sjögren, Dan, and Sundblad, Yngve (2000), ‘Co-​operative Design —​Perspectives on 20 Years with ‘the Scandinavian IT Design Model’. Keynote presentation, Proceedings of NordiCHI 2000, Stockholm, October 2000, http://​www.csc.kth.se/​ ~yngve/​SB-​PE-​DS-​YS-​Keyn​ote2​000.pdf . Cole, Robert (1989), Strategies for Learning: Small Group Activities in American, Japanese and Swedish Industry (Berkeley: University of California Press). Cooley, Mike (1981), Architect or Bee? (Slough: Langley Technical Services, Berks, UK). Dahlström, Edmund, Eriksson, Kjell, Gardell, Bertil, Hammarström, Olle, and Hammarström, Rut (1971), LKAB och demokratin, Rapport om en strejk och ett forskningsprojekt (Stockholm: Wahlström & Widstrand). Ehn, Pelle (1988), Work-​Oriented Design of Computer Artefacts, Erlbaum Download at https://​ www.diva-​por​tal.org/​smash/​get/​diva2:580​037/​FUL​LTEX​T02.pdf . Ellegard, Kajsa (1995), ‘The Creation of a New Production System at the Volvo Automobile Assembly Plant in Uddevalla, Sweden’, in Åke Sandberg, ed., Enriching Production (Aldershot: Avebury), 37–​59. Emery, Fred E., and Thorsrud, Einar [1964] (1969), Form and Content in Industrial Democracy (London: Tavistock and Assen: Van Gorcum).

60    ian Hampson and åke sandberg Emery, Fred E., and Thorsrud, Einar (1976), Democracy at Work (Martinus Nijhoff Social Sciences Division, Leiden). Engström, Tomas, and Medbo, Lars (1995), ‘Production System Design –​A Brief Summary of Some Swedish Design Efforts’, in Åke Sandberg, ed., Enriching Production (Aldershot: Avebury), 61–​73. Findlay, Patricia, Kalleberg, Arne L., and Warhurst, Chris (2013), ‘The Challenge of Job Quality’, Human Relations, 66 (4), 441–​451. Findlay, Patricia, Warhurst, Chris, Keep, Ewart, and Lloyd, Caroline (2017), ‘Opportunity Knocks? The Possibilities and Levers for Improving Job Quality’, Work and Occupations, 44/​ 1, 2–​22. Frege, Carola, and Godard, John (2014), ‘Varieties of Capitalism and Job Quality: The Attainment of Civic Principles at Work in the United States and Germany’, American Sociological Review, 79/​5, 942–​65. Gallie, Duncan (2003), ‘The Quality of Working Life: Is Scandinavia Different?’ European Sociological Review, 19/​1, 61–​79. Gallie, Duncan (2007), ‘The Quality of Work in Comparative Perspective’, in D. Gallie, ed., Employment Regimes and the Quality of Work (Oxford: Oxford University Press), Ch. 7. Greenbaum, Joan (1996), ‘Back to Labour’, Proceedings of the 1996 Conference on CSCW, ACM, 229–​237. Grote, Gudela, and Guest, David (2017), ‘The Case for Reinvigorating Quality of Working Life Research’, Human Relations, 70/​2, 149–​167. Hackman, Richard, and Oldham, Greg (1976), ‘Motivation through the Design of Work: Test of a Theory’, Organizational Behavior and Human Performance, 16, 250–​279. Hampson, Ian (1999), ‘Lean Production and the Toyota Production System: Or, the Case of the Forgotten Production Concepts’, Economic and Industrial Democracy, 20/​3, 369–​391. Håkansta, Carin (2014a), ‘Swedish Working Life Research. Formation and Conceptual Development of a Research Field in Transition’, Doctoral Thesis: Luleå University of Technology. Håkansta, Carin (2014b), ‘Between Science and Politics: Swedish Work Environment Research in a Historical Perspective, Article One’, Chapter in Doctoral Thesis, Luleå University of Technology. Hall, Peter, and David Soskice (2001), Varieties of Capitalism: The Institutional Foundations of Comparative Advantage (Oxford: Oxford University Press). Hampson, Ian, and Morgan, David E. (2016), ‘Trends in Critical “Work Life Research”: The Case of Australia’, in Åke Sandberg, ed., På jakt efter framtidens arbete –​utmaningar i arbetets organisering och forskning (Stockholm: Tankesmedjan Tiden), Ch. 16. Higgins, Winton (1980), ‘Working Class Mobilisation and Socialism in Sweden’, in Paul Boreham and Geoff Dow, eds., Work and Inequality. Vol. 1 (South Melbourne: MacMillan Press). Higgins, Winton (1985), ‘Political Unionism and the Corporatist Thesis’, Economic and Industrial Democracy, 6, 349–​381. Higgins, Winton (1986), ‘Industrial Democracy and the Control Issue in Sweden’, in Ed Davis, and Russell Lansbury, eds., Democracy and Control at the Workplace (Melbourne: Longman Cheshire.), Ch.14. Higgins, Winton, and Dow, Geoff (2013), Politics Against Pessimism: Social Democratic Possibilities since Ernst Wigforss (Bern: Peter Lang).

THE SWEDISH CONTRIBUTION TO JOB QUALITY     61 Howard, Robert (1985), ‘Utopia: Where Workers Craft New Technology’, Technology Review, 88/​3, 341–​349. Irani, Lily, and Silberman, Six (2016), ‘Stories We Tell about Labour: Turkopticon and the Trouble with Design’, Proceedings CHI’16, https://​esch​olar​shp.org/​uc/​item/​8nm27​3g3. Johansson, Jan, and Abrahamsson, Lena (2009), ‘The Good Work –​A Swedish Trade Union Vision in the Shadow of Lean Production’, Applied Ergonomics, 40, 775–​780. Karasek, Robert, and Theorell, Töres (1990), Healthy Work: Stress, Productivity and the Reconstruction of Working Life (United States of America: Basic Books). Katzenstein, Peter (1985), Small States in World Markets (Ithaca: Cornell University Press). Kjellberg, Anders (1998), ‘Sweden: Restoring the Model?’, in Anthony Ferner and Richard Hyman, eds., Changing Industrial Relations in Europe (Oxford: Blackwell Business), 74–​117. Korpi, Walter (1978), The Working Class in Welfare Capitalism (London: Routledge and Kegan Paul). Lansbury, Russell (2016), ‘Renewing Swedish Leadership in Work and Employment Relations Research’, in Åke Sandberg, ed., På jakt efter framtidens arbete –​utmaningar i arbetets organisering och forskning (Stockholm: Tankesmedjan Tiden), Ch. 3. Lundin, Per (2005), ‘Designing Democracy: The Utopia-​Project and the Role of the Labour Movement in Technological Change, 1981-​1986’, Paper No. 52, Royal Institute of Technology, Centre of Excellence for studies in Science and Innovation. Mahon, Rianne (1994), ‘From Solidaristic Wages to Soldaristic Work: A Post-​Fordist Historic Compromise for Sweden?’, in William Clement and Rianne Mahon, eds., Swedish Social Democracy: A Model in Transition (Canada: Canadian Scholars Press), 285–​314. Martin, Andrew (1992), ‘Foreword, in Åke Sandberg, with Gunnar Broms, Arne Grip, Lars Sunström, Jesper Steen, and Peter Ullmark, (1992), Technological Change and Co-​ Determination in Sweden (Philadelphia: Temple University Press), vii–​xxiv. Martin, Andrew (1987), ‘Unions, the Quality of Work, and Technological Change in Sweden’, in Carmen Sirianni, ed., Worker Participation and the Politics of Reform (Philadelphia: Temple University Press), 95–​139. Meidner, Rudolf (1980), ‘On the Concept of the Third Way: Some Remarks on the Socio-​Political Tenets of the Swedish Labour Movement’, Economic and Industrial Democracy, 1/​3, 343–​370. Meidner, Rudolf (1994), ‘The Rise and Fall of the Swedish Model’, in Wallace Clement and Rianne Mahon, eds., Swedish Social Democracy: A Model in Transition (Toronto: Canadian Scholars Press), Ch. 14. Metall/​Svenska metallindustriarbetareförbundet (1985), Det goda arbetet (Tiba Tryck). Movitz, Fredrik, and Sandberg, Åke (2013), ‘Contested Models: Productive Welfare and Solidaristic Individualism’, in Åke Sandberg, ed, Nordic Lights (Stockholm: SNS Förlag), Ch. 2. Movitz, Fredrik, and Sandberg, Åke (2019) ‘Ledning i Sverige och konstruktionen av svensk ledning’, in Åke Sandberg, ed., Arbete & välfärd (Lund: Studentlitteratur), Ch. 13. Nilssson, Lennart (1995), ‘The Uddevalla Plant: Why Did it Succeed with a Holistic Approach and Why Did it Come to an End’? in Åke Sandberg, ed., Enriching Production: Perspectives on Volvo’s Uddevalla Plant as an Alternative to Lean Production (Aldershot, England: Avebury), 75–​86. Noble, David (1977), America by Design (Oxford: Oxford University Press). OECD (2018), How Does SWEDEN Compare? http://​www.oecd.org/​swe​den/​jobs-​strat​egy-​ SWE​DEN-​EN.pdf . OECD (2019), OECD database, https://​stats.oecd.org/​Index.aspx?Data​SetC​ode=​JOBQ.

62    ian Hampson and åke sandberg Oscarsson, Bo (1997), 25 år av arbetslivets förnyelse –​forskning och utveckling på arbetslivsområdet 1972 –​1997 (Stockholm: Rådet för arbetslivsforskning). Pateman, Carole (1977), Participation and Democratic Theory (Cambridge: Cambridge University Press). Sandberg, Åke (1985), ‘Socio-​Technical Design, Trade Union Strategies and Action Research’, in Enid Mumford et al., eds, Research Methods in Information Systems (Amsterdam: Elsevier/​ North Holland), Ch. 5. Sandberg, Åke (ed.) (1995), Enriching Production: Perspectives on Volvo’s Uddevalla Plant as an Alternative to Lean Production (Aldershot, England: Avebury). May be downloaded at https://​akes​andb​erg.se/​downl​oad/​4508 /​ and at http://​www.fre​ysse​net.com/​?q=​en/​ node/​748. Sandberg, Åke (ed.) (2013a) Nordic Lights: Work, Management and Welfare in Scandinavia (Stockholm: SNS Förlag). Sandberg, Åke (ed.) (2013b), Contested Nordic Models of Work and Employment. Volvo Uddevalla and Welfare Capitalism. Stockholm: Arena idé. May be downloaded at http://​ www.akes​andb​erg.se/​contes​ted-​nor​dic-​mod​els-​of-​work/​ and http://​arena​ide.se/​rappor​ ter/​contes​ted-​nor​dic-​mod​els-​of-​work-​and-​emp​loym​ent-​volvo-​uddeva​lla-​and-​welf​are-​cap​ ital​ism/​. Sandberg, Åke (2016), ‘Arbetslivsforskning i Sverige –​skiss till historik’, in Åke Sandberg, ed, På jakt efter framtidens arbete –​utmaningar i arbetets organisering och forskning (Stockholm: Tankesmedjan Tiden), Ch. 2. Book in Swedish and Abstracts in English may be downloaded here: http://​www.akes​andb​erg.se/​cha​llen​ges-​for-​work-​resea​rch/​. Sandberg, Åke, with Broms, Gunnar, Grip, Arne, Sundström, Lars, Steen, Jester, and Ullmark, Peter (1992), Technological Change and Co-​Determination in Sweden (Philadelphia: Temple University Press). Streeck, Wolfgang (2017), How Will Capitalism End? Essays on a Failing System (London: Verso). Sörlin, Sverker (2016), ‘Blinda fläckar i samhällsförståelsen –​och hur vi kan undvika dem’, in Åke Sandberg, ed., På jakt efter framtidens arbete –​utmaningar i arbetets organisering och forskning (Stockholm: Tankesmedjan Tiden), Ch. 4. SWEA, Swedish Work Environment Authority (Arbetsmiljöverket, AV), (2015), Swedish Work Environment Research, 2017-​ 2027: Input to the Government’s Research Policy (Vällingby: Elanders Sverige). Thelen, Kathleen (2014), Varieties of Liberalization and the New Politics of Social Solidarity (New York: Cambridge University Press). Tourish, Dennis, Craig, Rusell, and Amernic, Joel (2017), How an Audit Culture Undermines the Purpose of Universities, in Tony Huzzard, Mats Benner, and Dan Kärreman, eds., The Corporatization of the Business School: When Minerva Meets the Market (Abingdon, Oxon: Routledge), 34–​54. Victorin, Anders (1979), ‘Co-​ Determination in Sweden: The Union Way’, Journal of Comparative Corporate Law and Securities Regulation, 2, 111–​140. Womack, J. P., Jones, D. T., and Roos, D. (1990), The Machine that Changed the World (New York: Ranson Associates).

Chapter 3

JOB QUAL I T Y A family affair? chris warhurst, sally wright And chris mathieu

Introduction This chapter outlines the parameters and range of concepts that can be attributed to current understandings of job quality. At its heart is a deceptively simple question: what is job quality? Most studies to date only partially attempt to answer this question. They tend to focus on measures and, typically, measures that are driven by data availability. Whatever datasets contain provides the basis for those studies in terms of what is considered to be job quality. As what is contained in datasets varies, so too then does what is presented as job quality. Such studies start with what can be measured rather than what needs to be measured, and what can be measured determines what is conceived of as job quality. Rarely do researchers start by trying to define job quality and then develop or identify datasets from which measurement can be made. As a consequence, a myriad of measures currently exist, which makes the evidence base messy. This chapter provides structured understanding of the variety of approaches to the study of the quality of jobs by outlining a ‘family of concepts’ related to the quality of jobs. Multiple terms and concepts have been developed, each seeking to capture dimensions or the entirety of the quality of jobs. The chapter starts by setting out various parameters applicable in analysing the concepts found in the ‘family’. It then surveys the more commonly used concepts in the family, highlighting their scope, content and origin, before turning to job quality itself. This section on job quality highlights core dimensions of job quality based on existing research and which might provide the basis of a future research consensus to the study of the quality of jobs. Unless otherwise stated, such as with the work of the ILO which has a global ambit, this discussion is limited to advanced developed economies.

64    chris warhurst, sally wright And chris mathieu

The Parameters In this section we outline four parameters for analysing the purpose and content of the family of concepts. A common issue runs through these parameters—​can and should jobs be assessed with or without reference to jobholders? Or, put another way, are jobs ‘artefacts’ of socio-​economic processes that can and should be treated unto themselves in an isolated, restricted and neutral manner, or are jobs inherently and inextricably associated with their actual or presumptive incumbents? This fundamental question permeates the discussion that follows.

Analytic versus Programmatic A basic differentiation can be made between conceptualizations that are analytical versus programmatic. Analytical conceptualizations seek to discern, define and clarify basic aspects of jobs for investigatory or academic purposes. This approach often provides the foundations for, but stops short of, a normative agenda or standard. Programmatic conceptualizations on the other hand seek to group advantageous and disadvantageous aspects of jobs and reflect normative standards, seeking to maximize advantageous dimensions and minimize disadvantageous dimensions. However, there is a link between analytical and programmatic conceptualizations—​in order to build a programmatic agenda, the basic elements or variables need to be defined before any decision can be made about them being advantageous or disadvantageous. Sometimes the same aspect can be advantageous to some and disadvantageous to other workers. An example is part-​time jobs, which can be advantageous for those who only can or wish to work part time, while disadvantageous for those workers who must or want to work full time. Similarly, temporary project-​based employment may be amenable to those who enjoy moving from project to project, affording an opportunity to choose between and negotiate with regard to various intrinsic and extrinsic rewards on offer in different projects. However such employment can be detrimental to those workers seeking durable employment security or lack the capacity to choose and negotiate from a position of strength (Barley and Kunda 2004). Furthermore, programmatic approaches often seek to summarize a broad range of factors that are taken to apply universally to all workers as well as setting a normative ‘bar’. In this respect, terms such as ‘fair’, ‘good’ and ‘decent’ are often used as prefixes to programmatic formulations regarding jobs, but often without clarifying what these prefixes mean, while still offering indicators and measures.

Advantageous for Whom? Explicitly embedded in programmatic approaches are the salience of jobholders or other beneficiaries and degrees of job qualities. Although many programmatic agendas

JOB QUALITY   65 seek to define and build upon more or less universally advantageous and disadvantageous aspects of jobs, it is pertinent to pose the question for whom specific aspects of jobs are advantageous or disadvantageous. There is a pervasive tendency to discuss the quality of jobs from the worker or employee perspective. However, the ‘for whom?’ question also invites analysis from the perspective of other actors, for example from that of different types of employees, as well as at least two other categories of actors or interested parties—​employers and ‘society’, the latter from the perspective of society’s primary representative—​the state. Thus, three main categories of interested parties can be discerned regarding the quality of jobs—​ employees, employers and society.1 These actors have sets of generally identifiable interests. Employees are primarily interested in jobs that promote various types of physical, material, social and mental (the latter two often now rolled up into what is called ‘psycho-​social’) forms of utility or well-​ being. Employers have an interest in jobs that are productive and efficient with direct positive economic outcomes, though they can also have a non-​instrumental interest in the well-​being of their employees. Society also has an interest in productive jobs resulting in economic growth. However, narrow economic productivity regimes often carry with them negative externalities—​such as employee ill-​health—​that are shunted onto the state, civil society institutions, individuals and their families. States thus seek either to curb or prevent these externalities being exported or force firms to bear or share the economic burdens of such externalities (see Albin et al. in their chapter, ‘The Cornerstone of Job Quality: Occupational Safety and Health’ in this volume). Society is also interested in jobs that promote workers’ interests in and abilities to participate in public life and the private-​sphere activities that public life builds upon, such as family, leisure, cultural and civic activities (see Befort et al. in their chapter, ‘Using Efficiency, Equity and Voice for Defining Job Quality, and Legal Regulation for Achieving It’ in this volume). The interests of actors within each category can diverge, with some aspects of the quality of jobs having differential effects on each category. Furthermore, advantages and disadvantages are not necessarily mutually exclusive to each category. The existence of so-​called ‘win-​win-​win’ situations display how certain aspects of jobs, such as training and education, can have positive effects across category groups. Likewise, workplace accidents can generate ‘lose-​lose-​lose’ situations for employees, employers and society.

Objective versus Subjective A further basic parameter is whether the quality of jobs can and should be assessed objectively or subjectively (see also Muñoz de Bustillo et al., ‘Quantitative Approaches to

1  We note that in the US there is debate about ‘communities’ as additional interested actors, see Bernhardt and Osterman (2017).

66    chris warhurst, sally wright And chris mathieu Assessing Job Quality’ and Knox and Wright ‘Understanding Job Quality Using Qualitative Research’ in this volume). One way of understanding the objective–​ subjective dichotomy is whether dimensions should be restricted to the characteristics of the jobs themselves or reflect the perspectives and/​or circumstances of the worker. Studies conducted from the objective tradition restrict their analysis to the (objective) constituent elements of a job per se, using standards or metrics affording factual measurement with neither individual normative nor perceptual bias or limitations (Popper 1972). The aim is to obtain a measure of the quality of jobs independent of workers’ personal perspectives or circumstances. Subjective approaches can be divided into two basic types. The first is through the circumstances of the worker. Here, the quality of the job is assessed through the preferences or expectations of the individual or categories of employees. From this perspective, the quality of a job is contingent on the worker and his/​her specific circumstances. The second type of subjective approach simply asks employees themselves to make the assessments. These assessments can be either of a factual or normative nature. The former comprises cognitive subjectivity, the latter normative subjectivity. A cognitively subjective approach asks individual employees to report on factual states from a particular positional vantage point, sometimes reporting on personal assessments (such as feelings of job insecurity, see below). However, the most common way of using the term ‘subjective’ is associated with the normative ‘fit’ of aspects of a job to the personal circumstances of an individual employee. The objective/​subjective distinction is not necessarily dichotomous—​jobs, and specific aspects of jobs, can be simultaneously objectively and/​or subjectively good or bad, satisfactory or unsatisfactory, as the Figure 3.1 below shows. Some objectively bad aspects of jobs can be perceived positively by some individual workers. Likewise, some aspects of jobs perceived negatively by some categories of workers can be objectively good (see Knox and Wright in this volume). Taking job insecurity as an example, Klandermans et al. (2010) show how subjective and objective assessments of the same phenomenon can be generated and crossed for analytical purposes. Their study builds upon the observation that persons objectively having high job security can feel job insecurity and vice versa. Moreover, perceptual beliefs can have behavioural consequences. In the case at hand, subjective beliefs and objective circumstance can not only contradict each other, objective circumstances

Subjective job quality

Objective job quality

+

-

+

Fulfilling good jobs

Unfulfilling good jobs

-

Fulfilling bad jobs

Unfulfilling bad jobs

Figure 3.1  Job quality categorization using subjective and objective approaches Source: Knox et al., 2015, p.1563.

JOB QUALITY   67 can also precondition subjective beliefs. Rather than associating this discrepancy with erroneous beliefs, Klandermans et al. associate subjective beliefs with a two-​fold risk assessment of the subjective probability of losing one’s job and the severity of the consequences of losing one’s job, which in turn are associated with objective contractual conditions—​holding permanent versus temporary contracts. As they argue, ‘Although there is no complete match between objective and subjective insecurity, it would be a mistake to look at job insecurity as subjective appraisal only’ (p. 559).

Intrinsic versus Extrinsic A further parameter that is sometimes conflated with the subjective–​objective distinction is that of intrinsic and extrinsic approaches. The reason for this conflation is the assumption that extrinsic dimensions of job quality are readily measured objectively, while intrinsic dimensions can only be measured subjectively. It is, however, important to keep these parameters—​the subjective–​objective and intrinsic–​extrinsic—​distinct, as they pertain to essentially different things. Intrinsic dimensions are inherent aspects of the job or work process itself. They are indivisible from the work and job and cannot be extracted out of the workplace. Extrinsic dimensions on the other hand are outcomes from carrying out a job that are deployed and capitalized upon outside the workplace. Some aspects of jobs almost exclusively belong to one category or the other. On the extrinsic side, wages for example are considered extrinsic as they are not deployed within the workplace (though wage-​differentials can give variable status and be the heart of organizational justice and fairness discussions and conflicts, see Tyler 2012). On the intrinsic side, the satisfaction derived from being able to spend time with patients and provide high-​quality patient care is something that can only be derived from carrying out the job in a particular way and does not produce anything that is deployable outside the workplace, though the feeling of meaning and accomplishment can be carried by the employee through time and space (Smart et al. 2014). Still other aspects have ‘dual natures’ in that they can be deployed in both arenas. Some occupational training fits this example if it simultaneously leads to improvement of the work process and satiates curiosity, giving a heightened sense of accomplishment by providing new knowledge, skill and understanding (intrinsic value), and if this new knowledge, skill and understanding has a market value in terms of making it easier to obtain a new or better job or higher wage (extrinsic value), that is, it can be converted outside the workplace into something desired by employees. Another example, is customer-​facing employees being provided with training to learn new language skills that can be deployed in work and outside the workplace, the latter when employees are on holiday for example (Findlay et al. 2006). A related distinction to the intrinsic–​extrinsic parameter is the differentiation between instrumental versus eudaimonia approaches to jobs and work (Turban and Yan 2016). Extrinsic dimensions map well onto the instrumental approach, as both highlight the external utility of the aspects of jobs that can be converted outside the workplace

68    chris warhurst, sally wright And chris mathieu for other things deemed valuable. Eudaimonia, or living well or ‘flourishing’ at and through work, maps well onto the intrinsic dimensions of work, if these are positive, as both predicate an inherent experience in the workplace that is of direct benefit, and not transferable, capitalizable or capable of being experienced outside of the workplace and work process. Some eudaimonic approaches make a single but broad factor such as ethical agency the core of potentially obtaining the state of flourishing through work (Hinchliffe 2004), whereas other approaches differentiate between multiple factors as the sources of two different forms of intrinsically derived happiness—​eudaimonic and hedonic (Straume and Vittersø 2012; Frenkel 2015). The instrumental–​fulfilment distinction is fairly well established, though the label ‘eudaimonia’ has been relatively recently affixed. The four parameters surface implicitly or explicitly and can be used to analytically differentiate the family of concepts to which we now turn.

The Family of Concepts An obvious reason why so many different terms have been used in research on the quality of jobs stems from no definitional agreement existing in the field. One result is that different terms and conceptual components overlap, sometimes are used interchangeably and sometimes conflated by researchers and policymakers. Others are simply overstretched. These problems are especially acute with regard to programmatic formulations that build upon common normative terms. What may be considered ‘decent’ or ‘good quality’ work can mean different things to different workers, social partners or policymakers. For example, in reference to decent work, Anker et al. (2003: 147) note that ‘every person at work or looking for work . . . has a notion of what “decency” at work stands for’. However, they acknowledged the difficulty in translating the general concept into easily understandable characteristics of work, including the difficulty in identifying statistical indicators with which those characteristics can be measured with an acceptable degree of consistency, accuracy and cross-​country comparability. As Muñoz de Bustillo et al. (2011) similarly note about job quality, it is ‘elusive’ because ‘it is one of those concepts . . . which everyone understands yet it is difficult to define precisely’ (p. 4). Another reason for the breadth of the conceptualizations regarding the quality of jobs is that this is an interdisciplinary research field and researchers bring their orthodox and heterodox theories, methodologies and vocabularies to the field. Hurley et al. (2012) identified seven disciplinary traditions of research on the quality of jobs, each with a different focus and different measures or indicators. For example, orthodox economists tend to focus on pay while traditional sociologists typically consider issues associated with skill, autonomy and job content. Researchers from the institutional tradition tend to focus on contract status and stability of employment, and opportunities for skills development and career progression. Researchers from the occupational medicine and

JOB QUALITY   69 health and safety approach focus on the physical and psychosocial risks of work. To this list, geography ought now to be attached, bringing spatiality to the analysis (see Weller and Campbell 2015). These starting points tend to have quite different scientific and normative implications.2 To unpack some of the ‘messiness’, it is instructive to consider the scope and origin of seven often used conceptualizations. Several concepts fall within the programmatic genre of approaches to the quality of jobs, while others are largely analytical. Some of the latter, though, also have a normative orientation. The dissection then suggests that these terms fall within the broader concept of job quality, arguing that a family of concepts exists which are all forms of job quality.

Decent Work Decent work is a concept popularized by the UN’s International Labour Organization (ILO). The ILO defines decent work as meaning ‘all women and men should work in conditions of freedom, equity, security and human dignity’ (ILO 1999, 2001). It focuses as much on labour markets as workplaces. The concept is connected to poverty reduction and reducing inequality in developing countries in the context of globalization and sustainable economic development. As part of its Sustainable Development Goal 8, the UN wants to achieve decent work for men and women by 2030. The ILO’s Decent Work Agenda has 12 dimensions: 1) social dialogue; 2) workers’ and employee representation; 3) employment opportunities; 4) adequate earnings and productive work; 5) decent working time; 6) combining work, family and personal life; 7) abolition of certain types of work (e.g. child labour, forced labour); 8) stability and security of work; 9) equal opportunity and treatment in employment; 10) safe work environment; 11) social security; and, 12) economic and social context for decent work. From the early 2000s, researchers attempted to operationalize the ILO’s concept of decent work using a variety of indicators and methods. The result was more a shopping list than coherent measurement (e.g. Standing 2002; Ghai 2003; Anker et al. 2003; Bescond et al. 2003; Bonnet et al. 2003). More recently, with its origins in development economics, decent work has been picked up by civil society organizations as a means to reduce in-​work poverty. For example, Oxfam Scotland applied the concept of decent work to its study of low-​wage workers in Scotland. Out of 26 decent work factors it identified, the five indicators rated the highest by these workers were: sufficient pay to cover basic needs; job security; paid holidays and sick leave; a safe working environment; and a supportive line manager (Stuart et al. 2016). Despite many researchers trying to operationalize this concept of decent work, it is probably best understood as a programmatic, aspirational political and social agenda (Munoz de Bustillo et al. 2011).

2 

For further exposition of these disciplinary traditions and foci, see Wright (2019).

70    chris warhurst, sally wright And chris mathieu

Fair Work Fair has become a vogue term and one interpreted differently in different countries by their respective governments. In each case there is a different focus and emphasis. Two current government-​led initiatives to create fair work occur in Australia and within the UK. Fair work in Australia centres on the provision of minimum stands of employment. The notion of a ‘fair go’ has historically underpinned Australian labour law and workplace relations. The 2000s in Australia were marked by industrial relations rancour. To settle this rancour in 2009, the Australian federal government’s concept of fair work was enshrined in ten National Employment Standards (NES) plus a national minimum wage (NMW). The NES are legislated minimum terms and conditions of employment covering the following areas: maximum weekly hours of work; the right to request flexible working arrangements; parental leave and related entitlements; annual leave; personal carers’ leave and compassionate leave; community service leave; long-​service leave; public holidays; notice of termination and redundancy pay; provision by employer of an information statement; plus the NMW (Cooper and Ellem 2009). Within the UK, fair work is promoted by both the Scottish and Welsh governments. In both cases fair work centres on the promotion of social partnership between employers and trade unions both as a feature of, and to support, progressive workplaces. The Scottish Fair Work Convention was established against the background of the Scottish government wanting to create inclusive workplaces that deliver economic growth. For the Fair Work Convention (2016), the indictors of fair work are: effective voice, opportunity, security, fulfilment and respect. Following the Scottish government’s footsteps, and similarly wanting to address in-​work poverty and inequality, the Welsh Government’s Fair Work Commission (2019) defined fair work as consisting of a series of characteristics: fair reward; employee voice and collective representation; security and flexibility; opportunity for access, growth and progression; safe, healthy and inclusive working environment; and legal rights respected and given substantive effect. However, neither country has legislative power over employment, which instead rests with the UK government. As a consequence, whist firmly embedded in policy statements, its translation into practice has been limited and fair work remains an aspiration in both countries.

Good Work Three versions of ‘good work’ emerge from the literature: narrow, broad and bridging. The narrow version centres around three indicators: workers’ rights and participation; equal opportunities, safety and health protection at work; and family-​friendly organization of work. It originated with EU policymakers wanting to encourage greater labour

JOB QUALITY   71 market participation and is promoted by the Council of the European Union to ensure that workplaces are inclusive. Prior to the Global Financial Crisis (GFC) in the 2000s, it was concerned with promoting the social dimension of the EU as a way to strengthen economic and social cohesion, harmonizing and improving working conditions. In this sense, it hopes to solve wider socioeconomic problems through work. More specifically, it wants to use work to lever the social inclusion and cohesion of marginalized groups. Measures centre on ‘flexicurity pathways’ tailored to labour market needs and reinforcing a lifecycle approach to work, with getting into and staying longer in work being the desired outcome. A tightening of labour markets across the EU followed the GFC. Although the European Commission maintained its position, the language of ‘good work’ lessened and even transmuted into ‘decent work’ (Council of the European Union 2007). The broad version seeks to build out from the workplace into society. In this version of good work, workers are assumed to be capable of self-​governance and their work needing to be socially responsible and meaningful (Overell 2008; Overell et al. 2010). It emerged out of civic-​minded organizations wanting a new paradigm, linking workplace stakeholders and work and society, and includes seven indicators: 1) safety and job security; 2) autonomy, choice, control, discretion and influence in work; 3) effort and reward that is fair, with reward systems that are open, transparent and equitable; 4) training that provides opportunities to develop and use skills, making work more satisfying and offering progression; 5) being treated fairly in a trusting environment; 6) relationships that build trust and resilience, respectful of diversity; and 7) rights to voice and association. The broad version continues to be promoted by civil society organizations, loosely defined. The UK’s Good Work Commission, for example, called for a reconfigured social settlement based on a ‘new tripartism’ that delivered ‘wider social gains’, and argued that ‘good work is an attempt to advance [and] arrive at a new settlement for work in the 21st century that breaks out of the straightjacket of interest-​ group gains and losses and instead aims at work that is able to deliver performance, engagement and fairness’ (pp. 12–​13). What may be considered a bridging version of good work is concerned with changes to the labour market and the quality of work and wanting to create better working lives. It emerged out of the UK government’s 2017 Taylor Review of Modern Working Practices (Taylor 2017a). It argues a dual case for creating good work: normative and instrumental. As to the first, it wants to prescribe what should be: in the twenty-​first century: ‘Bad work . . . just feels wrong’ stated Taylor (2017b). As to the second, it also recognizes the benefits to be gained, not just for workers but also employers and society. It notes the extent of in-​work poverty; the negative impact of bad jobs on health and well-​being; its association with low firm and national productivity; and the potential for automation to shape the future of work. The Review noted the problems of defining and measuring good work. Instead, it adopts an inductive approach—​letting people voice what good work means. It did, however, suggest a way to resolve the measurement problem, recommending indicators that only focus on the workplace, with a working

72    chris warhurst, sally wright And chris mathieu group (MJQWG 2019) established to develop measures. The resulting measures incorporate the six dimensions outlined for job quality later in this chapter.

Well-​being and Work Well-​being and work is another popular focus, gaining much academic and policy attention (e.g. Blanchflower and Oswald 2004; CIPD 2016; Green et al. 2015; Jahoda 1981; Muñoz de Bustillo et al. 2011; OECD 2013). Addressing and/​or improving employee well-​being is used liberally as a justification for measuring job quality. However, ‘well-​being’ is often poorly conceived and operationalized. Some analyses skip any definition as if the meaning of well-​being is self-​evident. Others suggest that it is related to employees’ mental health and sometimes both mental and physical health. In other studies, its operationalization is proxied by job satisfaction. Whilst there are varying characterizations and measures of ‘well-​being’ across these studies, they can be differentiated by two main approaches: ‘well-​being from work’ and ‘well-​being in work’. The well-​being from work approach emerged from concerns about the effect of unemployment on individual psychology as unemployment levels rose over the 1970s–​ 1980s. It was concerned with the (mainly) social psychological effects of not being in work through unemployment. More recent initiatives, however, are predicated on a need to extend from existing measures of economic performance to broader measures of social progress. The origins of well-​being in work also lie in psychology but there have been many different routes to and spinouts from it. For example, it has been adopted as a field of study by economics and sociology and is also a feature of human resource management. The well-​being in work approach is concerned with re-​invigorating debates about the link between job satisfaction (as a proxy for well-​being) and productivity in the workplace and economic growth nationally (a statistical link is accepted between employee satisfaction and performance, although causality, and the direction of causality, has yet to be firmly established). Not only is ‘well-​being’ variously conceived, but different studies also use different measures of well-​being. Measurements of well-​being and work include morale, labour turnover and absence; physical and mental health; work demands and environment; organizational values/​principles; collective/​social relations; and personal growth. The well-​being from work approach tends to focus on self-​esteem and depressive affect scales; internal–​external locus of control scale, and mood scales. The OECD (2013) ‘model’ focuses on three dimensions: life evaluation; affect—​a person’s feelings or emotional state; and eudemonia—​a sense of meaning and purpose in life. The well-​being in work approach focuses on ‘job-​related well-​being’ with two core dimensions: enthusiasm for and contentment with the job, both of which cover job satisfaction. Indeed, job satisfaction has become a proxy for well-​being and is the most commonly investigated aspect of well-​being by academics and policymakers. However, it should be noted there are problems with using job satisfaction as a measure: firstly, it is not synonymous with job

JOB QUALITY   73 quality (Green. 2009; Kalleberg 2011); secondly, it is difficult to determine if it is an input or outcome of job quality (see Muñoz de Bustillo et al. 2011).

Fulfilling Work There are also two versions of fulfilling work: the first seeks to extend from well-​being and the second focuses on the drivers of well-​being. The origins of this concept lie in psychology. The first version is vague on dimensions but includes pay, achievement, recognition, responsibility, advancement and the nature of the work. For example, UK recruiter Randstad assessed the relationship between jobs and the level of satisfaction expected by the British public (fulfilment was defined here as ‘the achievement of something desired, promised or predicted’). Results showed that pay was the primary reason for British workers to feel fulfilled and stay in their current job. Workers also valued achievement, recognition, responsibility, advancement and the nature of the work (Randstad 2014). The second version covers three themes: availability of work; quality of work; and work and well-​being. Each theme has a number of factors, with job satisfaction featuring as part of the work and well-​being theme (see also earlier in this chapter). It is sometimes difficult to disentangle fulfilling work and well-​being, as well as cause and outcome. For example, the civil society organization Carnegie UK Trust positioned fulfilling work as ‘a significant determinant of people’s wellbeing’ (White, 2016: 2).Building up a significant body of work promoting fulfilling work, the Carnegie UK Trust elsewhere suggests that work that is ‘motivating and life-​enhancing’ makes a significant contribution to this well-​being (Irvine 2018: 4 and 5).

Meaningful Work Originating again in psychology, the concept of meaningful work draws on 1970s’ concerns with worker motivation being addressed through work redesign (e.g. Hackman and Oldham 1976) but continues today through the Laboratory for the Study of Meaning and Quality of Work (Steger et al. 2012). It has three core components: work is perceived to have significance and purpose; work’s contribution to broader meaning in life; and work’s positive contribution to the greater good. There is an attempt here to displace job satisfaction with the concept of meaningful work: meaningful work can be derived from the individual or the organization and is argued to be a good predictor of desirable work attitudes (and a better predictor of absenteeism from work than job satisfaction). In the context of the ‘confusing workplace’ of the twenty-​first century, it is one part of the answer to the ‘who am I?’ question (Overell 2008: 6). However, meaning and meaningfulness has been a central optic in the analysis of work from Marx to the more contemporary studies of Casey (2002), Veltman (2016) and others studying ‘job-​crafting’, wherein employees mobilize autonomous capabilities to

74    chris warhurst, sally wright And chris mathieu pursue meaningful but not always sanctioned goals at work (Wrzesniewski et al. 2013). From a philosophical perspective, Veltman emphasizes four components of meaningfulness: exercising, developing and obtaining recognition for using human capabilities; the production of virtues such as self-​respect, dignity and pride through work; providing a sense of personal purpose and the sense of producing things of enduring value; and the integration of elements of a worker’s life into wider relationships, environments and activities with which the worker identifies. Similarly, for Casey, meaningfulness at work is a combination of individual biographical elements and pre-​structured social interpretations that are subjectively meshed in interaction with colleagues in the work setting. Though some elements are socially structured, the content of meaningfulness for Casey and other similar perspectives is variable, subjective and largely contingent upon workplace relationships.

Quality of Working Life (QWL) The concept of QWL emerged from the late 1940s and centred on making the technical and social systems in workplaces fit together better. It is a ‘needs-​based approach’ aimed at improving the quality of working life in the context of poor industrial relations and worker motivation. The concept emerged from psychology, then migrated to sociology and industrial relations. Initially, the QWL movement sought to address worker alienation, then extended to a drive for industrial democracy and wider social change (Trist and Bamforth 1951; Walton 1974, see also Guest in ‘The Quality of Working Life’ in this volume). The QWL movement resulted in job redesign experiments that restructured work groups and enhanced task participation. The initial concept had a limited focus on job design to improve what is now referred to as ‘employee engagement’ with work. The later, more expansive conceptualization of QWL evolved to incorporate adequate and fair compensation; safe and healthy working conditions; opportunity to use and develop human capacities; opportunity for continued growth and security; social integration in the work organization; constitutionalism in the work organization; work and the total life space; and the social relevance of work–​life. As theory and practice diffused across countries, variations in operational intent and disagreements about definition proliferated and the QWL movement began to lose its way and was dropped from its very prominent public policy position with the onset of economic recession in the mid-​1970s as companies and government looked less to redesigning work and more to creating work. Nevertheless, as work in the early twenty-​first century increasingly transcends or took place outside traditional bilateral employment relationships and on-​site work, QWL, in its more pervasive scope, received renewed attention. For example, there has been a recent call to reinvigorate QWL research in the context of the gig economy (Grote and Guest 2017; Guest in this volume) and non-standard employment (Warhurst and Knox 2022). All of the above seven concepts are concerned in some way with the quality of jobs and can be ‘umbrellaed’ by the term ‘job quality’. Indeed, many use their own conceptual

JOB QUALITY   75 label interchangeably with that of job quality. The ILO, for example, has used the terms ‘decent work’ and ‘job quality’ interchangeably in recent years (ILO 2015). Fair work in Scotland too is sometimes articulated as ‘job quality+​’ (Findlay 2018).

Job Quality Given that many if not most of these terms often interchangeably refer to ‘job quality’ in their articulation, it is not unreasonable to suggest that they are part of the family of concepts of job quality. Indeed the term ‘job quality’ has itself recently become a shorthand in debate and discussion amongst academics and policymakers, both in reference to these other terms and in itself (e.g. Cazes et al. 2015; Davoine et al. 2008; EC 2012; ILO 2015; Muñoz de Bustillo 2011; OECD 2013, 2016; Knox and Warhurst 2015; Leschke et al. 2012). The origins of concerns with job quality arguably lie with eighteenth-​ century proto-​social theorists and economists (the latter’s approach developed into the twentieth century, around the theory of compensating wage differentials, see Bryan and Rafferty 2015). Since then, it has become of interest to a range of disciplines as we have noted. It is important to understand what a ‘job’ is in two different contexts. At its most basic level, a job is a bundle of tasks carried out within an employment setting. However, in order to measure and compare the quality of jobs—​all of which are unique at the basic level as we have named, they need to be aggregated. This leads to ‘jobs’ becoming aggregated into an occupation—​an aggregation of roughly similar jobs (in the sense of bundles of tasks) within an industry. Confusion can arise when work, employment and jobs are conflated. Jobs comprise both work (the application of labour to tasks using technologies to produce use or exchange value) and the terms and conditions of employment through which that work is contracted, importantly, that it is paid. Sometimes the quality of work and the quality of employment are analysed separately. There are seven dimensions to the quality of employment, each with sub-​ dimensions: safety and ethics of employment; income and benefits from employment; working time and work–​life balance; security of employment and social protection; social dialogue; skills development and training; employment-​related relationships and work motivation. An example of a measure that is weighted towards employment is the UN Expert Group of Measuring Quality of Employment (2015). There are five dimensions of the quality of work: ergonomic; complexity; autonomy; control; and economic dimensions. An example of an approach weighted towards the organization and nature of work is provided by Centra and Gualitieri (2014). Comprehensive measures of job quality cover both aspects—​work and employment. Some measures purporting to be job quality extend beyond any given worker’s work and employment and into labour market issues. The European Commission’s Laeken Indicators have been criticized for not being confined to the attributes of jobs. The Laeken Indicators have 10 aggregate dimensions (only some of these

76    chris warhurst, sally wright And chris mathieu dimensions relate to the job, whereas others to the wider labour market context): intrinsic job quality; skills, lifelong learning and career development; gender equality; health and safety at work; flexicurity and security; inclusion and access to the labour market; work organization and work–​life balance; social dialogue and workers’ involvement; diversity and non-​discrimination; and overall economic performance and productivity. More narrowly focused approaches restrict themselves to particular aspects of work and employment within job quality. The obvious example is Osterman and Shulman’s (2011) use of only pay to define and measure job quality. Their argument for this highly restricted focus is that pay data is readily available, usually longitudinally and often in a form that allows international comparison. An example of a slightly more expanded but still restricted focus comes from the OECD (2016). This measure has three main dimensions: earnings quality; labour market security; and quality of the working environment, capturing non-​economic aspects of work. The OECD argues that these dimensions are important for worker well-​being and are constituents of a ‘good job’. In reality, it is probably better to think of such restricted-​scope approaches as weighted towards employment or work. Some measurement approaches scope across both the work and employment of jobs and only focus on aspects of the job. An example of such as an approach is that of Muñoz de Bustillo et al. (2011), who complain that some studies of job quality include dimensions that are not strictly properties of jobs. For example, certain dimensions are concerned with labour-​market conditions, organizational context and with outcomes at the individual level (e.g. job satisfaction, engagement, life satisfaction) and at the organizational level (e.g. economic performance, productivity). Muñoz de Bustillo et al. (2009: 25) call for a ‘purging’ of variables that do not directly affect the quality of the job, arguing that ‘the practice of “anything goes” in constructing job quality indicators has proven to be extremely detrimental to the relevance and usefulness of indicators’. Muñoz de Bustillo et al. (2011: 2) argue strongly in favour of strictly limiting indicators to those aspects of the job that have a clear and direct impact on the well-​being of workers. Their concept of job quality is linked on the one hand with the characteristics of the work performed and its environment (which they call the ‘work’ dimension, including among other things the level of autonomy at work, as well as its social and physical environment), and on the other hand with the characteristics of the contractual conditions under which the job is performed (which they call the ‘employment’ dimension, and includes pay, contractual stability and development opportunities, among other things). Their concept excludes issues which may be related to the well-​being of workers but which are not characteristics of the jobs they perform (such as their psychological states or the social support they have outside work) as well as concepts which concern the labour market (such as the level of unemployment). Their resulting measure has five dimensions: pay, intrinsic characteristics of work, quality of employment, work–​life balance, and health and safety. This measure is operationalized via a job quality index (JQI) that has gained some traction with the European Commission and its agencies.

JOB QUALITY   77 Taken together, the family of concepts is multifocal and multilevel. Its scope ranges from a focus on the individual worker (e.g. his or her well-​being) to the immediate worksite (e.g. versions of QWL) to the enterprise (e.g. Fair Work) to the economy and society (e.g. Good Work). Because the concept of job quality encompasses both work and employment, it covers the essential components of jobs while, in accordance with the approach of Muñoz de Bustillo et al. (2009, 2011), can also be delimited to focus on job-​specific factors while excluding peripheral features. Agreeing job quality as a concept that combines both work and employment factors with a core focus on job-​specific aspects, the following section offers a refined conceptualization of job quality that can provide a promising avenue to study the quality of jobs that can clear the hitherto messiness.

Towards a Consensus-​Derived Measure of Job Quality While there is growing consensus that job quality is a multidimensional concept, there is variance in the type and number of dimensions across work and employment that have been proposed as comprising job quality (Gallie 2007; Muñoz de Bustillo et al. 2011; Kalleberg 2011; see also Hauff and Kirchner in ‘Understanding Differences and Trends’ in this volume). This section focuses on the development of a measure of job quality that can obtain sufficient consensus to operate as a standard. Based in part of the terms outlined earlier to help underpin a systematic mapping of recent literature in the field (see Warhurst et al. 2017 and Wright et al. 2018), this measure is based on the identification of core dimensions of job quality that appear in recent research. This mapping revealed considerable variation in the number and combination of dimensions used by researchers. Typically, between four and ten dimensions are used. It also revealed that, while there is no agreed definition and measure, a relatively high degree of overlap exists in the number and type of dimensions used by various researchers. When extraneous characteristics are removed, overlapping dimensions are disentangled and regrouped, and semantic proliferation is simplified and consolidated, six core dimensions of job quality emerge. Below we list these six categories by topic heading and give examples of the components within each dimension. • Pay and other rewards: including objective aspects such as wage level, type of payment (e.g. fixed salary, performance pay) and non-​wage fringe benefits (e.g. employer-​provided pension and health cover) and subjective aspects (such as satisfaction with pay). • Intrinsic characteristics of work: including objective aspects (such as skills, autonomy, control, variety, work effort) and subjective aspects (such as meaningfulness, fulfilment, mastery, social support and powerfulness).

78    chris warhurst, sally wright And chris mathieu • Terms of employment: including objective aspects (such as contractual stability and opportunities for training, development and progression) and subjective aspects (such as perception of job security and mobility) • Health and safety: including physical and psycho-​social risks. • Work–​life balance: including working-​ time arrangements such as duration, scheduling and flexibility, as well as work intensity. • Representation and voice: including employee consultation, trade union representation and employee involvement in decision-​making. Jobs can be evaluated in terms of their contributions to organizational and economic performance, societal integration/​inclusion and participation, public health and welfare. While these job effects are unquestionably important issues, we contend that they should be dealt with as supplementary analyses whilst yet building upon a clearer and focused definition of job quality. We contend furthermore that the six dimensions named above adequately cover the quality of jobs, without overextending the concept of job quality. With this headway made, energies can turn to operationalizing the dimensions and their components. A number of authors have highlighted the difficulty in translating notions of job quality into a set of specific markers and indicators (e.g. Muñoz de Bustillo et al. 2011; Sengupta et al. 2009). Operationalization and measurement can be both pragmatic in terms of matching these to existing empirical datasets but endeavours should also be made in developing new variables, instruments and measurement strategies, especially as the landscape of jobs continues to change.

Concluding Remarks This chapter has highlighted the various popular terms used to capture job quality in existing studies, articulating these terms as part of a family of concepts. As part of doing so, it has also examined various parameters applicable for analysing these conceptualizations of the quality of jobs. As an encompassing yet focused analytical concept, with operationalizable components, the concept of job quality is argued to be the conceptual cornerstone upon which to proceed. While there are other ways of approaching the quality of jobs and their effects, job quality plays the pivotal conceptual role in providing an essential foundation of precision from which other perspectives can build. That a thousand conceptual and measurement flowers have bloomed is a strong indication of the importance of job quality in both scientific and policy communities. That a key reason for this interest is the well-​being of employees (no matter how ill-​defined) is to be welcomed. However, this floral explosion does create analytical problems. The differences in conceptual delineation and measurements of job quality create incommensurability, resulting in in international and even, in some cases such as the UK,

JOB QUALITY   79 sub-​national comparisons being disabled. What is essential to understanding job quality can be lost, as Hauff and Kirchner note in this volume. To solve these problems, we would argue for the adoption of the six dimensions of job quality that have been outlined, if only because they emerge out of reviews of existing multidisciplinary research, and, for this reason, both provide a basis for building scientific and policymaker consensus and can be populated with sub-​dimension indictors that have already been tested analytically. However, as two of us have argued elsewhere (Mathieu and Warhurst 2018) an international governmental body, such as the European Commission, OECD or UN through its ILO, and working with statisticians, needs to take the lead and adopt and apply a single, coherent, comprehensive, standardized and measurable framework for job quality for its member countries. This adoption will enable standardized measurement of job quality and encourage researchers and policymakers to develop supportive datasets if they don’t already exist. Thereafter the framework would facilitate comparative analysis and, over time, longitudinal analysis. It would help governments and researchers monitor and evaluate trends and developments and progress in improving job quality. Drawing on the research base, this chapter has suggested such a standard measure. Whilst it is important that new concepts emerge, develop and are tested, the six dimensions suggested offer an underpinning analytical use for research into the quality of jobs for the future.

Acknowledgement This chapter has been developed from research conducted for the UK’s Chartered Institute for Personnel and Development (CIPD) and the Horizon 2020 project QuInnE, (project no. 649497) and we thank the CIPD and the European Commission for their support.

References Albin, M., Mathieu, C., Takala, E-​S., and Theorell, T. (2022), ‘The Cornerstone of Job Quality –​Occupational Safety and Health’, in C. Warhurst, C. Mathieu, and R. Dwyer, eds, Oxford Handbook of Job Quality (Oxford: Oxford University Press), 220–243. Anker, R., Chernyshev, I., Egger, P., Mehran, F., and Ritter, J. (2003), ‘Measuring Decent Work with Indicators’, International Labour Review, 142/​2, 147–​177. Barley, S., and Kunda, G. (2004), Gurus, Hired Guns, and Warm Bodies (Princeton: Princeton University Press). Befort, S. F. and Budd, J. W. (2022), ‘Using Efficiency, Equity and Voice for Defining Job Quality, and Legal Regulation for Achieving It’, in C. Warhurst, C. Mathieu, and R. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 522–541. Bernhardt, A., and Osterman, P. (2017), ‘Organizing for Good Jobs: Recent Developments and New Challenges’, Work & Occupations, 44/​1, 89–​112.

80    chris warhurst, sally wright And chris mathieu Bescond, D., Châtaignier, A., and Farhad Mehran, F. (2003), ‘Seven Indicators to Measure Decent Work: An International Comparison’, International Labour Review 142/​2, 179–​211. Blanchflower, D., and Oswald, A. (2004), ‘Well-​Being Over Time in Britain and the USA’, Journal of Public Economics, 88/​7–​8, 1359–​1386. Bonnet, F., Figueiredo, J. B., and Standing, G. (2003), ‘A Family of Decent Work Indexes’, International Labour Review, 142/​2, 213–​238. Bryan, D., and Rafferty, M. (2015), ‘Beyond the Necessary (Utilitarian) and Mundane: The Economics of Job Quality’, in A. Knox and C. Warhurst, eds., Job Quality in Australia (Annandale, NSW: Federation Press), 127–​149. Casey, C. (2002), Critical Analysis of Organizations: Theory, Practice, Revitalization (London: Sage). Cazes, S., A. Hijzen, and A. Saint-​Martin (2015), ‘Measuring and Assessing Job Quality: The OECD Job Quality Framework’, OECD Social, Employment and Migration Working Papers No. 174. (Paris: OECD). Centra, M., and Gualtieri, V. (2014), ‘The Quality of Work during the Employment Crisis: The Italian Case’, International Review of Sociology, 24/​2, 241–​258. Chartered Institute for Personnel & Development (CIPD) (2016), Growing the Health and Well-​Being Agenda (London: CIPD). Cooper, R., and Ellem, B. (2009), ‘Fair Work and the Re-​Regulation of Collective Bargaining’, Australian Journal of Labour Law, 22/​3, 284–​305. Council of the European Union (2007), Presidency Conclusions of the Brussels European Council (8/​9 March 2007), 7224/​1/​07 REV 1, http://​www.consil​ium.eur​opa.eu/​ueD​ocs/​ cms_​D​ata/​docs/​pressD​ata/​en/​ec/​93135.pdf. Davoine, L., Erhel, C., and Guergoat-​ Larivière, M. (2008), ‘Monitoring Quality in Work: European Employment Strategy Indicators and Beyond’, International Labour Review, 147/​2–​3, 163–​198. European Commission (EC) (2012), Europe 2020, http://​ec.europa.eu/​europe2020/​ Fair Work Commission (2019), Fair Work Wales (Welsh Government), https://​gov.wales/​sites/​ defa​ult/​files/​publi​cati​ons/​2019-​05/​fair-​work-​wales.pdf. Fair Work Convention (2016), Fair Work Framework 2016, www.fairworkconvention.scot. Findlay, P. (2018), ‘Fair Work in Scotland’, Making Good Work: Policy, Practice and Research Workshop. (London: Science Gallery). Findlay, P., Stewart, R., Dutton, E., and Warhurst, C. (2006), Evaluation of the Scottish Union Learning Fund (SULF) (2000-​2005), Research Findings No. 39, Social Research, (Glasgow: Scottish Executive). Frenkel, S. (2015), ‘Integrating Sociological and Psychological Perspectives in Studying Job Quality’, in A. Knox and C. Warhurst, eds., Job Quality in Australia (Annandale, NSW: Federation Press), 150–​171. Gallie, D. (2007), ‘Production Regimes, Employment Regimes, and the Quality of Work’, in D. Gallie, ed., Employment Regimes and the Quality of Work (Oxford: Oxford University Press), 1–​34. Ghai, D. (2003), ‘Decent Work: Concept and Indicators’, International Labour Review 142/​2, 113–​145. Green, F. (2009), Demanding Work: The Paradox of Job Quality in the Affluent Economy. (Princeton and Oxford: Princeton University Press). Green, F., Mostafa, T., Parentthirion, A., Vermeylen, G., van Houten, G., Biletta, I., and Lylyyrjanainen, J. (2015), Is Job Quality Becoming More Unequal? Industrial and Labor Relations Review, 66/​4, 753–​784.

JOB QUALITY   81 Grote, G., and Guest, D. (2017), ‘The Case for Reinvigorating Quality of Working Life Research’, Human Relations, 70/​2, 149–​167. Guest, D. (2022), ‘The Quality of Working Life’, in C. Warhurst, C. Mathieu, and R. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 23–40. Hackman, J. R., and Oldham G. R. (1976), ‘Motivation through the Design of Work: Test of a Theory’, Organisation Behaviour and Human Decision Process, 16, 250–​279. Hauff, S., and Kirchner, S. (2022), ‘Understanding Differences and Trends in Job Quality: Perspectives from Cross-​National Research’, in C. Warhurst, C. Mathieu, and R. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 87–106. Hinchliff, G. (2004), ‘Work and Human Flourishing’, Educational Philosophy and Theory, 36, 535–​547. Hurley, J., Fernández-​ Macías, E., and Muñoz de Bustillo, R. (2012), ‘Assessing Recent Employment Shifts in Europe Using a Multidimensional Job Quality Indicator’, in E. Fernández-​Macías, J. Hurley, and D. Storrie, eds., Transformation of the Employment Structure in the EU and USA, 1995-​2007 (London: Palgrave), 147–​179. ILO (1999), Decent Work: Report of the Director General, International Labour Conference, 87th Session (Geneva: International Labour Office). ILO (2001), ‘Reducing the Decent Work Deficit: A Global Challenge’, Report of the Director-​ General to the 89th session of the International Labour Conference. ( Geneva: ILO). ILO (2015), A Market Systems Approach to Decent Work (Geveva: ILO). Irvine, G. (2018), Fulfilling Work in Ireland: Discussion Paper (Dunfermline: Carnegie UK Trust). Jahoda, M. (1981), ‘Work, Employment and Unemployment: Values, Theories and Approaches in Social Research’, American Psychologist, 36, 184–​191. Kalleberg, A. (2011), Good Jobs, Bad Jobs (New York: Russell Sage Foundation). Klandermans, B., Hasselink, J., and van Vuuren, T. (2010), ‘Employment Status and Job Insecurity: On the Subjective Appraisal of an Objective Status’, Economic and Industrial Democracy, 31/​4, 557–​577. Knox, A., and Warhurst, C. (2015), Job Quality: Perspectives, Problems and Proposals (Annandale, NSW: Federation Press). Knox, A., and Wright, S. (2022), ‘Understanding Job Quality Using Qualitative Research’, in C. Warhurst, C. Mathieu, and R. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 107–125. Leschke, J., Watt, A., and Finn, M., (2012), ‘Job Quality in the Crisis –​an Update of the Job Quality Index (JQI)’, ETUI-​REHS Research Department Working Paper No. 2012.07. (Brussels: European Trade Union Institute). Mathieu, C., and Warhurst, C. (2018), ‘QuInnE Policy Recommendations’, QuInnE Working Paper No. 15, www.quinne.eu. Measuring Job Quality Working Group (2018), ‘Measuring Good Work’, Carnegie UK Trust/​ RSA, https://​www.carn​egie​uktr​ust.org.uk/​publi​cati​ons/​measur​ing-​good-​work-​the-​final-​ rep​ort-​of-​the-​measur​ing-​job-​qual​ity-​work​ing-​group/​ . Muñoz de Bustillo, R., Fernández-​Macías, E., Antón, J. and Esteve, F. (2009), Indicators of Job Quality in the European Union (Brussels: European Parliament, Department of Employment and Social Affairs). Muñoz de Bustillo, R., Fernández-​Macías, E., and Antón, J. (2022), ‘Quantitative Approaches to Assess Job Quality’, in C. Warhurst, C. Mathieu, and R. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press), 126–147.

82    chris warhurst, sally wright And chris mathieu Muñoz de Bustillo, R., Fernández-​Macías, E., Esteve, F., and Antón, J. (2011), Measuring More than Money (Cheltenham: Edward Elgar). OECD (2013), ‘Well-​ Being in the Workplace: Measuring Job Quality’, in How’s Life 2013: Measuring Well-​being (Paris: OECD), 147–​171. OECD (2016), How Good Is Your Job? Measuring and Assessing Job Quality (Paris: OECD). Osterman, P., and Shulman, B. (2011), Good Jobs in America (New York: Russell Sage Foundation). Overell, S. (2008), Inwardness: The Rise of Meaningful Work (London: Work Foundation). Overell, S., Mills, T., Roberts, S., Lekhi, R., and Blaug, R. (2010), ‘The Employment Relationship and the Quality of Work’, Provocation Paper 7, Good Work Commission. (London: Work Foundation). Popper, K. (1972), Objective Knowledge (Oxford: Clarendon Press). Randstad UK (2014), Fulfilment@Workreport2014. Randstad, www.randstad.co.uk/​about-​us/​ press-​releases/​randstadnews/​10-​million-​peopleprofessionally-​unfulfilled-​acrossuk/​. Sengupta, S., Edwards, P. K., and Tsai, C.-​J. (2009), ‘The Good, the Bad, and the Ordinary’, Work and Occupations, 36/​1, 26–​55. Smart, D., English, A., and James, J. (2014), ‘Compassion Fatigue and Satisfaction: A Cross-​ Sectional Survey among US Healthcare Workers’, Nursing and Health Sciences 16/​1, 3–​10. Standing, G. (2002). ‘From People’s Security Surveys to a Decent Work Index’, International Labour Review, 141(4): 441–​454. Steger, M., Dik, B., and Duffy, R. (2012), ‘Measuring Meaningful Work: The Work and Meaning Inventory (WAMI)’, Journal of Career Assessment, 20/​3, 322–​337. Straume, L.V., and Vittersø, J. (2012), ‘Happiness, Inspiration and the Fully Functioning Person: Separating Hedonic and Eudaimonic Well-​Being in the Workplace’, Journal of Positive Psychology, 7/​5, 387–​398. Stuart, F., Pautz, H., and Wright, S. (2016), Decent Work for Scotland’s Low-​Paid Workers: A Job To Be Done (Glasgow: Oxfam Scotland). Taylor, M. (2017a), Good Work: The Taylor Review of Modern Working Practices (London: HM https://​www.gov.uk/​gov​ernm​ent/​publi​cati​ons/​good-​work-​the-​tay​lor-​rev​ Government). iew-​of-​mod​ern-​work​ing-​practi​ces. Taylor, M. (2017b), What Does Good Work Mean For You? https://​www.the​rsa.org/​disco​ver/​ publi​cati​ons-​and-​artic​les/​matt​hew-​tay​lor-​blog/​2017/​05/​what-​does-​good-​work-​mean-​for-​ you?utm​_​med​ium=​email&utm​_​sou​rce=​fel​lows​hip-​new​slet​ter&utm_​c​ampa​ign=​Good-​ work-​lau​nch. Trist, E., and Bamforth, K. (1951), ‘Some Social and Psychological Consequences of the Longwall Method of Coal Getting’, Human Relations, 4, 21–​32. Turban, D. B., and Yan, W. (2016), ‘Relationship of Eudaimonia and Hedonia with Work Outcomes’, Journal of Managerial Psychology, 31/​6, 1006–​1020. Tyler, T. (2012), ‘Justice and Effective Cooperation’, Social Justice Research, 25/​4, 355–​375. UN Expert Group of Measuring Quality of Employment (2015), Handbook on Measuring Quality of Employment (Geneva: UN). Veltman, A. (2016), Meaningful Work (New York: Oxford University Press). Walton, R. (1974), ‘Improving the Quality of Working Life’, Harvard Business Review, May/​ June, 12–​16. Warhurst, C., Wright, S., and Lyonette, C. (2017), Understanding and Measuring Job Quality: Part 1 –​Thematic Literature Review (London: CIPD).

JOB QUALITY   83 Warhurst, C., and Knox, A. (2022), ‘Manifesto for a new Quality of Working Life’, Human Relations, 75/2, 304–321. Weller, S., and Campbell, I. (2015), ‘The Contextualized Geographies of Job Quality’, in A. Knox and C. Warhurst, eds., Job Quality in Australia (Annandale, NSW: Federation Press), 84–​102. White, D. (2016), ‘Fulfilling Work’, Work and Wellbeing Discussion Paper. (Dunfermline: Carnegie UK Trust). Wright, S. (2019), ‘Conceptualising and Operationalising Job Quality in Australia’, Doctoral thesis, University of Warwick. Wright, S., Warhurst, Lyonette, C., and Sarkar, S. (2018), Understanding and Measuring Job Quality: Part 2 –​Indicators of Job Quality (London: CIPD). Wrzesniewski, A., LoBuglio, N., Dutton, J. E., and Berg, J. M. (2013), ‘Job Crafting and Cultivating Positive Meaning and Identity in Work’, in A. Bakker, ed., Advances in Positive Organizational Psychology (Bingley: Emerald Publishing), 281–​302.

PA RT I I

U N DE R STA N DI N G J OB QUA L I T Y

CHAPTER 4

understanding differences and trends in job quality Perspectives from cross-​national research sven hauff and stefan kirchner

Introduction In order to understand job quality and its development in a country it is useful to look beyond national borders. Accordingly, an increasing number of studies are focusing on cross-​national differences and trends in job quality. These studies offer plural conceptual perspectives regarding the fundamental question of which work and employment conditions are relevant for defining a job as ‘good’ or ‘bad’. They also use different empirical approaches to analysing cross-​national differences and/​or global trends: some studies look at differences and trends regarding specific job quality dimensions, others seek to empirically identify interrelationships between different dimensions or construct indices of overall job quality. Even though the plurality of conceptual perspectives and empirical approaches are useful in developing a comprehensive picture of cross-​national differences and trends in job quality, the question arises as to why there is no international consensus about the conceptualization and analysis of job quality. This chapter addresses this issue. We first provide an overview of the current state of research, showing how job quality is conceptualized in cross-​national research and which analytical approaches are undertaken to study cross-​ national differences and trends in job quality. We understand job quality as the sum of work and employment conditions related to a job that influence the employees’ well-​being (including, for instance, physical and psychological health, job satisfaction) as well as job-​related attitudes (e.g. commitment and turnover intentions) (Green 2006; Muñoz de Bustillo et al. 2011b; Holman, 2013b). Accordingly, our review does not include an analysis of studies of labour market conditions, since aspects such as national unemployment levels do not refer to actual work and employment conditions. Similarly, we do not include studies that only address the outcomes of job quality, such as job satisfaction. Since our

88    sven hauff and stefan kirchner understanding of job quality embraces various work and employment conditions, we limit our review to studies that explicitly address cross-​national differences and/​or trends in job quality as a multidimensional construct using several variables. Building on our review, we introduce the idea of job quality as a relational concept. We argue that the evaluation of work and employment conditions is always embedded in a specific context, that is the specific institutional and cultural conditions in a country at a specific time. This perspective not only fills the gap in the lack of consensus in cross-​national research about the conceptualization and analysis of job quality, it also provides a deeper understanding of the differences and trends in job quality across countries.

Conceptual Perspectives on Job Quality A fundamental issue in job quality research regards the question of what defines a job as ‘good’ or ‘bad’. Several theoretical traditions provide distinct perspectives on this question by emphasizing different dimensions of job quality (Muñoz de Bustillo et al. 2011a; see also Knox and Wright, ‘Understanding Job Quality Using Qualitative Research’ in this volume, and Muñoz de Bustillo et al., ‘Quantitative Approaches to Assessing Jobs’ in this volume). For instance, the theory of compensating differentials highlights the importance of wages, while behavioural theories focus on hierarchy and possibilities for employee participation. Traditional sociological approaches emphasize the relevance of self-​determination potential as well as an occupation’s meaningfulness for intrinsic job quality. Institutional approaches refer to the importance of contract status and employment period. Reports on the relationships between physical and psychological risks and job quality can be found in the literature on occupational medicine, and occupational health and safety. Finally, studies on work–​life balance concentrate on aspects such as working hours, work duration and work intensity. In light of these diverse perspectives, current research offers various conceptuali­ zations of job quality, in which different work and employment conditions are integrated as dimensions of job quality. For example, the International Labour Organization (ILO) provides a framework of Decent Work Indicators that includes employment opportunities, adequate earnings and productive work, decent working time, combining work, family and personal life, work that should be abolished, work stability and security, equal opportunity and treatment in employment, a safe work environment, social security, social dialogue and employers’ and workers’ representation (ILO 2012). Likewise, the European Foundation for the Improvement of Living and Working Conditions (Eurofound) has developed and implemented a conceptual framework for measuring job quality that includes four main dimensions of job quality, namely earnings, prospects, working time quality and intrinsic job quality. The latter dimension is further differentiated into four sub-​dimensions: skills use and discretion, social

understanding differences and trends in job quality    89 environment, physical environment and work intensity (Eurofound 2012). Similarly, the OECD focuses on three key dimensions of job quality: earnings quality, labour market security and quality of the working environment (OECD 2014). Beyond these policy-​oriented initiatives, many researchers are analysing cross-​ national differences and trends in job quality by using different conceptualizations of job quality. The plurality of perspectives becomes most apparent when we compare different studies using the same datasets. Various key studies in job quality research have used the European Working Condition Survey (EWCS) data from 2005 or the work orientation modules from the International Social Survey Program (ISSP), which both provide valuable information about the work and employment conditions in several countries. Using the same datasets, different authors have empirically analysed different dimensions of job quality (Table 4.1). A dimension accounted for in almost all studies is job autonomy. Often used variables are task variety, physical and ambient demands, wage level, work duration and scheduling, training and development opportunities and perceived job security. Other variables—​such as dependency of work, clear formulation of tasks and requirements, work–​life balance and the boundaries between work and private life, or aspects concerning fairness at work—​are used rather seldom. Interestingly, there are different perspectives regarding the measurement of single dimensions. For instance, the EWCS data from 2005 contains several questions about job autonomy: Respondents provided information on the ability to change task order, work methods, work speed or work rate as well as on the influence on the choice of working partner, the ability to apply own ideas, the assessment of the quality of own work, the influence on the division of rotating tasks and team autonomy. Tangian (2009) used all of these variables. In contrast, Smith et al. (2008) as well as Leschke and Watt (2008) referred only to the ability to change task order, work methods, work speed or work rate. An even more general approach can be found in studies that use the ISSP data (Clark 2010; Olsen et al. 2010) because, in the ISSP data, autonomy is measured by a single item that captures the respondent’s agreement or disagreement with the statement ‘I can work independently’. Furthermore, some studies integrate additional aspects. For instance, Leschke and Watt (2008) also used trade union density to make statements about overall job quality in European countries. Davoine et al. (2008) further integrated aspects such as unemployment rates, the employment gap between men and women, productivity or length of maternity leave. Here, we argue that these aspects are not necessarily dimensions of job quality: If job quality concerns actual work and employment conditions that determine employee well-​being, these variables either represent possible influencing factors of the work and employment conditions (e.g. union density) or general labour market conditions. Summing up, these studies do not draw on a common concept of job quality. This pattern has continued with more recent studies, for example Green et al. (2013), Greenan et al. (2014) and Holman and Rafferty (2017). Differences amongst these studies include their choice of job quality dimension as well as the measurement of single dimensions. Regarding the latter, one could argue that a detailed measurement (e.g. a detailed

90    sven hauff and stefan kirchner Table 4.1 Dimensions of job quality captured in EWCS 2005 and ISSP data and their uses in different studies

✓

Variety (non-​ monotonous work, interesting job)

✓

✓

✓

✓

Olsen et al. (2010)

Clear formulation of tasks and requirements

✓

✓

Clark (2010)

✓

✓

Availability in ISSP (1989, 1997, 2005)

Involvement (e.g. employees are consulted about changes in work organization)

✓

Holman (2013b)

✓

✓

Muñoz de Bustillo et al. (2011b)

Dependency (e.g. work pace depends on colleagues and machines)

Tangian (2009)

✓

Leschke and Watt (2008)

Gallie (2007a)

✓

Smith et al. (2008)

Availability in EWCS (2005) Job content

Autonomy (e.g. ability to change task order, work methods, work independently)

Davoine et al. (2008)

Studies using EWCS (2005) data

Dimensions of job quality

Autonomy and involvement

Studies using ISSP data

✓ ✓

✓

✓

✓ ✓

(✓)

✓

✓

✓

✓

✓ ✓

✓

✓

Intellectual demands

✓

✓

✓

✓

Emotional demands

✓

✓

✓

✓

Contact with others (colleagues, customers)

✓

✓

✓

✓

Meaningfulness (useful to society)

✓

✓

Social Relationships to relationships and support from management

✓

✓

Relationships to and support from colleagues

✓

✓

✓

✓

✓

✓ ✓

✓

✓

✓ ✓

✓

✓

✓ ✓

understanding differences and trends in job quality    91 Table 4.1  (Continued)

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

Ambient demands (e.g. noise)

✓

Workload

✓

Work intensity (high speed, tight deadlines)

✓

✓

Wage level

✓

✓

Payment type (e.g. fixed salary, performance pay)

✓

Full-​time vs. part-​time Working time Duration (exact hours and work-​life per week) balance Scheduling (work at night and over weekends)

✓ ✓

✓

✓

✓ ✓

✓

✓ ✓

✓

✓

✓ ✓

✓

✓

✓

✓ ✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

(✓)

✓

✓

(✓)

✓

✓

✓

Work–​life balance

✓

Clear boundaries between work and private life

✓

Skills utilization

✓

✓

Flexibility (possibility to adapt working hours, to take an hour off or to take breaks)

Skills and Training development Development opportunities

✓

✓ ✓ ✓

✓

✓

✓

Olsen et al. (2010)

✓

Clark (2010)

Tangian (2009)

✓

Availability in ISSP (1989, 1997, 2005)

Leschke and Watt (2008)

✓

✓

Holman (2013b)

Smith et al. (2008)

✓

Physical demands (e.g. carrying or moving heavy loads)

Muñoz de Bustillo et al. (2011b)

Davoine et al. (2008)

Availability in EWCS (2005) Wages and payment

Gallie (2007a)

Studies using EWCS (2005) data

Dimensions of job quality

Working conditions

Studies using ISSP data

✓

✓

✓

✓

✓

✓

(✓)

✓

✓

✓

✓

✓

✓

✓

✓

✓

(✓)

✓ ✓

(continued)

92    sven hauff and stefan kirchner Table 4.1  (Continued)

Fairness

✓

(✓)

✓

✓

✓

✓

Perceived job security

✓

Workplace (e.g. working at company premises, working at home)

✓

Distance between home and work

✓

✓

Physical violence

✓

✓

✓

Bullying or harassment

✓

✓

✓

Discrimination

✓

✓

✓

✓

Olsen et al. (2010)

Availability in ISSP (1989, 1997, 2005)

✓

Contract type

Clark (2010)

Holman (2013b)

Tangian (2009)

✓

Muñoz de Bustillo et al. (2011b)

✓

Leschke and Watt (2008)

Smith et al. (2008)

Davoine et al. (2008)

Availability in EWCS (2005) Workplace

Gallie (2007a)

Studies using EWCS (2005) data

Dimensions of job quality

Contractual status and stability

Studies using ISSP data

✓ ✓

(✓)

✓

✓

✓

Note: Analysis of Davoine et al. (2008) refers to their second taxonomy; checkmarks in parenthesis: not available in all three waves. Source: Authors’ own analysis

measure of the different aspects of work autonomy) might be preferable in order to generate a comprehensive picture of the work and employment conditions of jobs. However, several work and employment conditions might not be encountered in all jobs. For instance, not all jobs include the need to choose a working partner or to work in teams and to rotate tasks. Thus, more general measures (e.g. a broad measure if employees can work independently) that apply to a wide range of jobs might be more useful in order to compare the work and employment conditions in a broad variety of jobs and jobs across sectors. Regarding the choice of job quality dimensions, our analysis of these studies reveals that there are very different concepts of job quality. The question of which dimensions need to be integrated to properly address differences and trends in job quality is not an easy one. On the one hand, it can be argued that we need a comprehensive picture

understanding differences and trends in job quality    93 that includes all possible dimensions. The risk here is that one gets lost in detail. On the other hand, it might be reasonable to focus on key or primary dimensions of job quality. However, this raises the question: What are the most important and relevant dimensions of job quality? Some authors argue that single aspects such as pay (Osterman and Shulman 2011) or employee involvement (Eurofound 2013) are decisive, while others concentrate on a few key aspects (e.g. Gallie 2007a; Eurofound 2012; OECD 2014). Thus, it is difficult to decide what the most relevant dimensions of job quality are, and there is the risk of missing important aspects. In fact, we argue that it is very difficult to have a common concept of job quality since the definition of what makes a job ‘good’ or ‘bad’ is always context dependent. Before we delve deeper into this issue, we will expand our analysis of these studies by considering the analytical approaches undertaken in cross-​ nations’ research on differences and trends in job quality.

Analytical Approaches In addition to the aforementioned conceptual differences, there are different approaches to analysing cross-​national differences and trends in job quality (see Table 4.2). We will briefly introduce these approaches and highlight key findings. We will also discuss the main advantages and disadvantages associated with the different approaches.

Analysis of Discrete Dimensions of Job Quality A basic approach to analysing cross-​national differences and trends in job quality is to look at discrete dimensions of job quality (e.g. autonomy, job security, etc.) and to describe the differences between countries as well as developments within countries or general trends. This approach is most often used in cross-​national research on job quality, and the available studies provide very detailed and useful insights. Since it would go beyond the constraints of this chapter to compare all of these findings, we present only one study—​that of Olsen et al. (2010) which included four countries (US, Great Britain, West Germany and Norway), three points in time (1989, 1997 and 2005) and a wide range of job aspects. Their results indicate that job security, the ability to work independently and the quality of physical working conditions and interpersonal relationships tend to be better in Norway and West Germany than in the US and Great Britain. In contrast, advancement opportunities, the possibility to help others, utility to society and overall job satisfaction tend to be higher in the US. Results indicate an increase of job insecurity and work intensity in all countries in at least one period. However, there is also evidence of stability and improvement. For example, intrinsic job quality improved in the US and West Germany. Olsen et al. also found evidence for convergence of job security, work intensity, working conditions and social relationships across the four countries, with convergence of job security and work intensity strongest.

94    sven hauff and stefan kirchner Table 4.2 Approaches to analysing cross-​national differences and trends in job quality Approach

Exemplary studies

Analysis of discrete dimensions

Clark, 2005, 2010; Green, 2006; De Grip and Wolbers, 2006; Gallie, 2007a; 2007b; European Commission, 2008; Leschke and Watt, 2008; Smith et al., 2008; Peña-​Casas and Pochet, 2009; Drobnič, Beham, and Präg, 2010; Greenan, Kalugina, and Walkowiak, 2010; Olsen et al., 2010; Erhel, Guergoat-​Larivière, Leschke, and Watt, 2012; Esser and Olsen, 2012; Eurofound, 2012; Leschke, Watt, and Finn, 2012; Green et al., 2013; Greenan et al., 2014; Leschke and Watt, 2014; OECD, 2014; Eurofound, 2015; Holman and Rafferty, 2017

Analysis of patterns at the job level

Lorenz and Valeyre, 2005; Valeyre et al., 2009; Eurofound, 2012; Holman, 2013b

Analysis of patterns at the country level

Davoine and Erhel, 2006; Davoine et al., 2008; European Commission, 2008

Composite indices approaches

European Commission, 2008; Leschke and Watt, 2008; Tangian, 2007; 2009; Muñoz de Bustillo, Fernández-​Macías, Antón et al., 2011, 2011; Antón, Fernández-​Macías, and Muñoz de Bustillo, 2012; Erhel et al., 2012; Leschke et al., 2012; Holman, 2013a; Leschke and Watt, 2014

Source: Authors’ own analysis

This example underlines the point that analysis of single dimensions can provide detailed information on cross-​national differences and trends for specific job aspects. This approach allows, for instance, targeted intervention by policymakers. A disadvantage of such an approach is that important interrelationships between discrete dimensions of job quality are not accounted for. Combined effects of separate aspects (e.g. higher autonomy together with lower job security) are not considered, rendering a comprehensive evaluation of overall job quality difficult. A further problematic aspect of such an approach is that it can attain high complexity. With a large number of dimensions, time periods and countries integrated into the analysis, the overall picture becomes fuzzy and general explanations become increasingly difficult. For instance, Olsen et al. link their analysis to the institutionalist literature, namely the varieties of capitalism (Hall and Soskice 2001) and the employment regime (Gallie 2007b) theories. However, these two theoretical approaches cannot fully explain their reported findings because the cross-​national differences and developments across time are too heterogeneous.

Analysis of Patterns at the Job Level There are good theoretical reasons to assume that discrete dimensions of job quality are interrelated, forming different job types (Kalleberg 2011). For instance, labour market

understanding differences and trends in job quality    95 segmentation theories assume that certain aspects of job quality connect, resulting in ‘particular sets of characteristics or governing rules’ (Tilly 1997: 269). Likewise, industrial relations research distinguishes between different employment systems (e.g. Arthur 1994; Osterman 1987), and human resource management (HRM) literature assumes that human resource practices are not used in isolation but rather form HRM systems—​as distinct bundles or configurations of different HRM practices (e.g. Hauff et al. 2014; Jackson et al. 2014). A comprehensive analysis in this direction was conducted by Holman (2013b). Using the 2005 5th Wave of the EWCS, Holman developed a taxonomy of job types in Europe that not only provides insights into the interrelationships between different job quality dimensions, but also these different job types’ qualities and their prevalence in different institutional regimes. Holman identified six job types in detail: active jobs (characterized for instance by high job discretion and high social support), saturated jobs (similar to active jobs but with higher demands such as longer working hours), team-​based jobs (e.g. team-​based work with high autonomy levels), passive independent jobs (e.g. low work intensity, little contact with others, and low intellectual and emotional demands), insecure jobs (e.g. a non-​permanent contract and low pay), and high-​strain jobs (e.g. high work intensity, physical and ambient demands). He further showed that these job types differ in relation to job quality. Measured against job satisfaction as well as psychological and physical well-​being, active jobs are characterized by high job quality, and saturated and team-​based jobs by moderate quality, passive jobs by low to moderate quality, and insecure and high-​strain jobs by low quality. Finally, he could show that the distribution of high-​quality and low-​quality jobs across Europe is largely in line with expectations based on the employment regime approach (Gallie 2007b). Accordingly, the proportion of higher-​quality jobs is highest in social democratic regimes (Denmark, Finland and Sweden). Continental regimes (e.g. Germany and Austria) have a higher proportion of higher-​quality jobs than liberal regimes (UK, Ireland) and southern European regimes (e.g. Italy, Spain and Portugal). Liberal regimes have a higher proportion of higher-​quality jobs than southern European regimes. However, the proportion of higher-​quality jobs in transitional regimes (e.g. the Czech Republic and Hungary) is not significantly different from that in continental and liberal regimes, but is significantly higher than in southern European regimes. Analysis of the interrelationships between single job quality dimensions at the job level provides valuable insights into different forms of work organization, their distribution across countries, and the related overall job quality. It also reduces complexity by analysing and describing differences and trends at the aggregate job level. However, the relatively pronounced degree of abstraction naturally limits the possibilities for detailed assessments, for instance concerning distribution and variance of single job quality dimensions within countries or their contributions to overall job quality. In addition, such analysis does not allow a disentangling of whether different job types’ overall effects are caused by single work and employment conditions or the specific combinations of these conditions.

96    sven hauff and stefan kirchner

Analysis of Patterns at Country Level The notion of interrelationships between different job quality dimensions can also be transferred to a macroeconomic perspective. Institutional regimes may influence not only the prevalence of a single dimension of job quality (e.g. the degree of autonomy) but rather affect the typical combination of work and employment conditions, leading to national employment systems (Dobbin and Boychuk 1999). The shaping of these national employment systems will affect job quality in a given country. Such an approach has been presented by Davoine et al. (2008), who sought to identify different national models of job quality in Europe. The authors used multiple sources with mainly macroeconomic indicators but also self-​reported data from employees. Applying a principal component analysis, they showed that different job and labour market features co-​vary. For instance, relatively bad working conditions (health at risk and long working days) tend to occur with bad labour market performance (long-​term unemployment, involuntary part-​time work). In contrast, high levels of mean wages, training, productivity and job satisfaction manifest simultaneously. Through a subsequent cluster analysis, Davoine et al. (2008) identified four models of job quality across Europe. The northern cluster (e.g. Finland, Sweden and the UK) is characterized by high wages, good working conditions and high participation in training and education, but also by a comparatively high work intensity. The continental cluster (e.g. Germany, France and Ireland) displays high wages and productivity but average rates of training and education. The southern cluster (e.g. Italy, Spain and Portugal) is characterized by low employee involvement and narrow pay gaps between genders. The new member states cluster (e.g. Hungary, the Czech Republic, Poland and the Slovak Republic) is characterized by low socio-​economic security (low wages and high long-​term unemployment) and bad working conditions (long working days and health risks), but work intensity is much lower compared to other clusters. The findings of Davoine et al. (2008) highlight the interrelationships between certain work and employment conditions at the country level. However, interpretation of the results is not always easy because positive and negative features occur together (e.g. good working conditions vs. high intensity in the northern cluster). Generally, however, this approach is promising in terms of analysing patterns at the country level in order to compare different settings of work and employment conditions and their effects on job quality within countries. However, such an approach leads to a loss of information because it is lacks sensitivity to possible variety within countries.

Composite Indices Approaches A further possibility to analyse cross-​national differences and trends in job quality is to design an overall index of job quality that comprises various work and employment conditions. This approach has been pursued in several studies but, to date, there

understanding differences and trends in job quality    97 is agreement neither about the dimensions to be used nor about the weighting of single dimensions within the indices (for an overview and discussion on various index approaches, see Muñoz de Bustillo et al. 2011b). Consequently, results differ depending on the construction of index being used. To illustrate this approach, Table 4.3 presents the findings of three index approaches that all refer to data from the EWCS from 2005. The comparison shows various similarities amongst the studies. First, overall job quality is highest in the northern European countries, together with the UK, Ireland, the Netherlands, Luxemburg and Belgium. Second, average values are mostly achieved in continental European countries, followed by southern European countries. Third, job quality is lowest in most of the East European countries. However, there are also some remarkable differences in the assessments of the overall job quality within the EU. For instance, Slovenia is ranked twelfth by Leschke and Watt (2008), which is slightly above average of the EU27. However, the same country is ranked twenty-​fourth by Tangian (2009). Index approaches can also be used to analyse trends across time. For instance, Leschke, Watt and Finn (2012) analysed developments between 2005 and 2010 and found a small decline in overall job quality between these years in the EU. Within this general decline, there are countries with pronounced deteriorations (Ireland, France, UK and Sweden) but also countries with improvements in overall job quality (Poland, the Czech Republic, Belgium and Denmark). Index approaches seek to analyse cross-​national differences and trends in job quality at a highly aggregated level. While the results are useful because countries can be compared and positioned unambiguously according to their overall job quality, there is also a loss of information at a less aggregated level. Thus, index approaches neglect specific differences in particular dimensions across and within countries. As a consequence, they provide little basis for targeted policy interventions at that level. Accordingly, studies that seek to construct an overall job quality index often provide information at a less aggregated level. For example, Leschke and Watt (2008) analysed country profiles on the basis of sub-​indices and revealed heterogeneity within and between countries. Leschke, Finn and Watt (2012) used the same sub-​indices to show developments regarding specific job aspects. In addition, based on the approach of Muñoz de Bustillo et al. (2011a), Antón et al. (2012) showed that overall job quality differs across sectors and occupations. Since some countries have a larger proportion of jobs with low job quality, the average job quality in these countries is lower. Thus, differences in job quality not only result from different work and employment conditions in the same job type but also from the distribution of jobs within a country.

Job Quality as a Relational Concept Initially, we defined job quality as the work and employment conditions that influence employee well-​being. Building on this definition, it could be assumed that

98    SVEN HAUFF AND STEFAN KIRCHNER Table 4.3 Index approaches based on the EWCS 2005 data Leschke and Watt (2008)

Tangian (2009)

Muñoz de Bustillo et al. (2011b)

Max. difference in ranking places

Denmark

1

1

2

1

Netherlands

2

2

5

3

UK

3

4

3

1

Sweden

4

8

6

4

Finland

5

7

9

4

Luxemburg

6

6

1

5

Belgium

7

5

7

2

Ireland

8

3

4

5

Austria

9

9

10

1

Malta

10

12

15

5

France

11

17

8

9

Slovenia

12

24

17

12

Cyprus

13

11

13

2

Germany

14

10

11

4

Italy

15

18

12

6

Estonia

16

21

20

5

Czech Republic

17

23

19

6

Portugal

18

20

16

4

Latvia

19

15

24

5

Lithuania

20

26

25

9

Spain

21

13

14

8

Hungary

22

19

23

4

Slovakia

23

16

21

7

Bulgaria

24

14

26

12

Greece

25

27

18

9

Romania

26

22

27

5

Poland

27

25

22

5

Notes: 1) =​top 20%, =​upper 20 to upper 40%, =​middle 20%, =​lower 20 to 40%, =​lowest 20%. 2) Leschke and Watt (2008) used several other sources of data in addition to that of EWCS 2005. Source: Own depiction

understanding differences and trends in job quality    99 there is a defined set of economic (pay, benefits, etc.) and non-​economic factors (autonomy, involvement, etc.) that constitute job quality. However, our analyses show that current studies integrate different work and employment conditions as dimensions of job quality. The question that arises is why there is no international consensus when defining and analysing job quality (Findlay et al. 2013). Here, we argue that the question of what defines a job as ‘good’ or ‘bad’ depends on the specific situation in a country at a specific time. More specifically, since the evaluation of work and employment conditions is always embedded in a specific context, that is, the particular socio-​ economic conditions of a country at a specific time, it is very difficult to generate a common concept of job quality. This point is made by others. For example, Muñoz de Bustillo et al. (2011a: 69) state that ‘the characteristics of employment interact with the features of social systems in ways that can make similar employment characteristics have very different implications for the well-​being of the worker in different countries’. Similarly, Carré et al. (2012: 9) point out that ‘[j]‌ob quality can be “read” differently in different countries at different times’. Thus, job quality cannot be a fixed construct. Several theoretical approaches underline this point. In particular, the varieties of capitalism (Hall and Soskice 2001) and employment regime (Gallie 2007b, 2009; Korpi 1978, 2006) approaches point out that institutional context has major implications for job quality in a given country (see also Amable 2003). For example, proponents of the varieties of capitalism approach argue that the institutional arrangements of different national market economies foster or impede particular economic activities. In an ideal typical distinction, liberal market economies (LMEs) (e.g. the US and UK), rely on market mechanisms, a low level of labour regulation and relaxed employment protection. In contrast, Germany and the Scandinavian countries are examples of co-​ordinated market economies (CMEs), characterized by a high level of regulation and higher employment protection. Employment regime theory focuses on the institutionalized role of organized labour. Within this framework, employment regimes—​that is, social democratic (e.g. Scandinavian countries), continental (e.g. Germany and France), liberal (e.g. the US and UK), southern European (e.g. Spain and Greece), and transitional regimes (e.g. Hungary)—​differ systematically by, for example, organized labour’s participation in decision-​making, the principles underlying employment policy, the salience of quality of work–​life programmes, and the level of welfare protection offered to the unemployed. Together, institutional theories highlight that different institutional contexts generate specific national employment systems (Dobbin and Boychuk 1999) with distinct (combinations of) work and employment conditions. Therewith, we argue that national institutions also define which work and employment conditions are seen as important aspects of job quality. In other words, national institutional conditions shape specific work and employment conditions by what March (1994) calls dominant ‘logics of appropriateness’. These logics guide actors’ behaviours by establishing general identities and providing rules to interact at the workplace (see also Frege and Godard 2014; Kirchner and Hauff 2019). Thus, the evaluation of work and employment conditions is

100    sven hauff and stefan kirchner always influenced by and depended on the specific institutional context. Importantly, the relative influence of different work and employment conditions might change over time since institutions change (Streeck and Thelen 2005). Static country classifications of national models have long sparked controversies (Hancké et al. 2007; Crouch 2009; Lane and Wood 2012). Also, initiatives at the EU level (e.g. Rubery et al. 2008) as well as ubiquitous national government liberalization policies (e.g. Thelen 2012) increasingly undercut the notion of well-​defined, static country categories. Accordingly, there is also a temporal aspect that needs to be considered. A similar argument can be made for national culture. Culture refers to ‘acquired knowledge that shapes values, originates attitudes and affects behaviour, and which members of a society (or a social group) use to interpret experience and generate social behaviour’ (Vaiman and Brewster 2014: 152). Accordingly, national culture is a key determinant in how work and employment conditions are perceived by individuals (Hauff et al. 2015; Reiche et al. 2012; Richter et al. 2022). Indeed, according to Triandis (1994), national culture forms the ‘base rate’ that defines what employees expect from work. Similarly, Erez (1994: 568) states that ‘culture shapes the cognitive schemas that ascribe meaning and values to motivational variables and guide our choices, commitments, and standards of behavior. Culture serves as a criterion for evaluating the meaning of various managerial techniques and the valences of their behavioral outcomes’. Thus, culture also defines which work and employment conditions are seen as important aspects of job quality or what is ‘appropriate’. Again, culture does not lead to static views on what defines a job as ‘good’ or ‘bad’ since national cultures change over time (Taras et al. 2012). The outlined relational perspective on job quality helps explain the differences in defining and analysing job quality and is also useful for cross-​national research. In particular, it points out that the relationships between work and employment conditions and employee well-​being might not be universal. Rather, together with individual employee characteristics (see Knox and Warhurst 2015), national characteristics moderate the influences of work and employment conditions on employee well-​being (see Figure 4.1). Accordingly, findings from national surveys at specific times should not be simply transferred to different countries and different periods. Instead, job quality always has to be seen in context. This perspective should be incorporated into scientific research as well as policymaker thinking. Nationally or internationally developed (e.g. at the level of the EU or the OECD) models of job quality always reflect a specific territorial focus that might not be applicable to other territories. Whilst necessary and useful, integrating the relational perspective also provides a new challenge for studies of job quality. It questions for instance the common construction of composite indices, which are usually built using equal weights across countries and time. Using equal weights does not correspond with the idea that the influence of work and employment conditions on job quality depends on the specific situation in a country at a specific time. To date, the relative influences of different work and employment conditions have received little attention in studies of job quality. Tangian (2009) and Drobnič et al. (2010) for instance have shown that different job aspects’ effects on

understanding differences and trends in job quality    101

NATIONAL CHARACTERISTICS Institutions and Socio-Economic Factors (e.g., social security, wealth) National Culture (e.g., power distance, individualism)

Country Level Individual Level

WORK AND EMPLOYMENT CONDITIONS

EMPLOYEE WELL-BEING

INDIVIDUAL CHARACTERISTICS (e.g., age, gender, education, occupation)

Figure 4.1  Job quality as a relational concept Source: Adapted from Drabe, Hauff, and Richter (2014)

overall job satisfaction differs across countries. However, job satisfaction is just one outcome of job quality and further well-​being dimensions, for example physical and psychological health, eudemonic happiness and social well-​being (Grant et al. 2007; Guerci et al. 2022), must be considered.

Conclusion Our analyses have shown that current research on cross-​national differences and trends provides multiple views regarding the conceptualization and empirical analysis of job quality. Policy-​oriented initiatives and research projects include very different perspectives regarding the question of what job quality is and how it is constituted. This does not only refer to the question of which work and employment conditions should be considered as a dimension of job quality, but also to the question of how the single dimensions should be measured in order to get a comprehensive picture. Besides these different perspectives on how to conceptualize job quality, there are several empirical approaches to analysing cross-​national differences and trends. In particular, we identified studies focusing on the analysis of discrete dimensions, studies analysing patterns at the job level and the country level, and studies using composite indices. Each of these approaches has its merits and its weaknesses.

102    sven hauff and stefan kirchner The diverse perspectives are partly a result of different theoretical views and partly a result of the choice of data and specific variables. Beyond that, we argue that the evaluation of work and employment conditions is always embedded in a specific context, that is, the specific institutional and cultural conditions in a country at a specific time. Therefore, it is very difficult to have a common concept of job quality. Job quality is constituted through simultaneous effects of different work and employment conditions on employee well-​being. However, single work and employment conditions have unequal influences, and the effect of a certain job aspect on job quality might differ across countries and over time. Going forward, new studies should pay greater attention to this relational perspective on job quality. We still need to fully understand different dimensions’ relative influence on employee well-​being. A comprehensive analysis that incorporates the relational perspective would provide valuable insights, especially regarding the question of which are the most influential dimensions of job quality across countries. Thus, we encourage future research to further determine the roles of specific contexts when analysing the question of what defines a job as ‘good’ or ‘bad’.

References Amable, B. (2003), The Diversity of Modern Capitalism (Oxford: Oxford University Press). Antón, J-​I., Fernández-​Macías, E., and Muñoz de Bustillo, R. (2012), ‘Identifying Bad Jobs across Europe’, in C. Warhust, F. F. P. Carré, and C. Tilly, eds., Are Bad Jobs Inevitable? Trends, Determinants and Responses to Job Quality in the Twenty-​First Century (New York: Palgrave Macmillan), 25–​44. Arthur, J. B. (1994), ‘Effects of Human Resource Systems on Manufacturing Performance and Turnover’, The Academy of Management Journal, 37, 670–​687. Carré, F., Findlay, P., Tilly, C., and Warhurst, C. (2012), ‘Job Quality: Scenarios, Analysis and Interventions’, in C. Warhust, F. F. P. Carré, and C. Tilly, eds., Are Bad Jobs Inevitable? Trends, Determinants and Responses to Job Quality in the Twenty-​First Century (New York: Palgrave Macmillan), 1–​22. Clark, A. E. (2005), ‘Your Money or Your Life: Changing Job Quality in OECD Countries’, British Journal of Industrial Relations, 43, 377–​400. Clark, A. E. (2010), ‘Work, Jobs, and Well-​Being across the Millennium’, in E. Diener, J. F. Helliwell, and D. Kahneman, eds., International Differences in Well-​Being. Series in Positive Psychology (Oxford: Oxford University Press), 436–​468. Crouch, C. (2009), ‘Typologies of Capitalism’, in B. Hancké, ed., Debating Varieties of Capitalism: A Reader (Oxford: Oxford University Press), 75–​94. Davoine, L., and Erhel, C. (2006), ‘Monitoring Employment Quality in Europe: European Employment Strategy Indicators and Beyond’, Document de Travail No. 66. Noisy-​le-​ Grand: Centre d’Etudes de l’Emploi. Davoine, L., Erhel, C., and Guergoat-​ Larivière, M. (2008), ‘Monitoring Quality in Work: European Employment Strategy Indicators and Beyond’, International Labour Review, 147, 163–​168. De Grip, A., and Wolbers, M. H. J. (2006), ‘Cross-​National Differences in Job Quality among Low-​Skilled Young Workers in Europe’, International Journal of Manpower, 27, 420–​433.

understanding differences and trends in job quality    103 Dobbin, F., and Boychuk, T. (1999), ‘National Employment Systems and Job Autonomy: Why Job Autonomy is High in the Nordic Countries and Low in the United States, Canada, and Australia’, Organization Studies, 20, 257–​291. Drabe, D., Hauff, S., and Richter, N. F. (2014), ‘Job Satisfaction in Aging Workforces: An Analysis of the USA, Japan and Germany’, The International Journal of Human Resource Management, 26, 783–​805. Drobnič, S., Beham, B., and Präg, P. (2010), ‘Good Job, Good Life? Working Conditions and Quality of Life in Europe’, Social Indicators Research, 99, 205–​225. Erez, M. (1994), ‘Toward a Model of Cross-​Cultural Industrial and Organizational Psychology’, in M. D. Dunnette, H. C. Triandis, and L. M. Hough, eds., Handbook of Industrial and Organizational Psychology (Palo Alto, Calif.: Consulting Psychologists Press), 559–​608. Erhel, C., Guergoat-​Larivière, M., Leschke, J., and Watt, A. (2012), ‘Trends in Job Quality during the Great Recession: A Comparative Approach for the EU’, Document de Travail No. 161-​1. Noisy-​le-​Grand: Centre d’Etudes de l’Emploi. Esser, I., and Olsen, K. M. (2012), ‘Perceived Job Quality: Autonomy and Job Security within a Multi-​Level Framework’, European Sociological Review, 28, 443–​454. Eurofound (2012), Trends in Job Quality in Europe (Luxembourg: Publications Office of the European Union). Eurofound (2013), Work Organisation and Employee Involvement in Europe (Luxembourg: Publications Office of the European Union). 2010 Eurofound (2015), Convergence and Divergence of Job Quality in Europe 1995–​ (Luxembourg: Publications Office of the European Union). European Commission (2008), Employment in Europe 2008 (Luxembourg: Office for Official Publications of the European Communities). Findlay, P., Kalleberg, A. L., and Warhurst, C. (2013), ‘The Challenge of Job Quality’, Human Relations, 66, 441–​452. Frege, C., and Godard, J. (2014), ‘Varieties of Capitalism and Job Quality: The Attainment of Civic Principles at Work in the United States and Germany’, American Sociological Review, 79, 942–​965. Gallie, D. (2007a), ‘Production Regimes and the Quality of Employment in Europe’, Annual Review of Sociology, 33, 85–​104. Gallie, D. (2007b), ‘Production Regimes, Employment Regimes, and the Quality of Work’, in D. Gallie, ed., Employment Regimes and the Quality of Work (Oxford: Oxford University Press), 1–​33. Gallie, D. (2009), ‘Institutional Regimes and Employee Influence at Work: A European Comparison’, Cambridge Journal of Regions, Economy and Society, 2, 379–​393. Grant, A. M., Christianson, M. K., and Price, R. H. (2007), ‘Happiness, Health, or Relationships? Managerial Practices and Employee Well-​Being Tradeoffs’, Academy of Management Perspectives, 21, 51–​63. Green, F. (2006), Demanding Work: The Paradox of Job Quality in the Affluent Economy (Princeton: Princeton University Press). Green, F., Mostafa, T., Parent-​Thirion, A., Vermeylen, G., van Houten, G., Biletta, I., and Lyly-​Yrjanainen, M. (2013), ‘Is Job Quality Becoming More Unequal?’ Industrial and Labor Relations Review, 66, 753–​784. Greenan, N., Kalugina, E., and Walkowiak, E. (2010), ‘Trends in Quality of Work in EU-​ 15: Evidence from the European Working Conditions Survey (1995-​2005)’, Document de Travail No. 133. Noisy-​le-​Grand: Centre d’Etudes de l’Emploi.

104    sven hauff and stefan kirchner Greenan, N., Kalugina, E., and Walkowiak, E. (2014), ‘Has the Quality of Working Life Improved in the EU-​15 between 1995 and 2005?’ Industrial and Corporate Change, 23, 399–​428. Guerci, M., Hauff, S., and Gilardi, S. (2022), ‘High Performance Work Practices and Their Associations with Health, Happiness and Relational Well-​being: Are There Any Tradeoffs?’ International Journal of Human Resource Management, 33, 329–​359. Hall, P. A., and Soskice, D.,W. (eds) (2001), Varieties of Capitalism: The Institutional Foundations of Comparative Advantage (Oxford: Oxford University Press). Hancké, B., Rhodes, M., and Thatcher, M. (2007), ‘Introduction’, in B. Hancké, M. Rhodes, and M. Thatcher, eds., Beyond Varieties of Capitalism: Conflict, Contradictions, and Complementarities in the European Economy (Oxford: Oxford University. Press), 1–​20. Hauff, S., Alewell, D., and Hansen, N. K. (2014), ‘HRM Systems between Control and Commitment: Occurrence, Characteristics and Effects on HRM Outcomes and Firm Performance’, Human Resource Management Journal, 24, 424–​441. Hauff, S., Richter, N. F., and Tressin, T. (2015), ‘Situational Job Characteristics and Job Satisfaction: The Moderating Role of National Culture’, International Business Review, 24, 710–​723. Holman, D. (2013a), ‘An Explanation of Cross-​National Variation in Call Centre Job Quality Using Institutional Theory’, Work, Employment and Society, 27, 21–​38. Holman, D. (2013b), ‘Job Types and Job Quality in Europe’, Human Relations, 66, 475–​502. Holman, D., and Rafferty, A. (2017), ‘The Convergence and Divergence of Job Discretion Between Occupations and Institutional Regimes in Europe from 1995 to 2010’, Journal of Management Studies, 55, 619–​647. International Labour Organization (ILO) (2012), Decent Work Indicators: Concepts and Definitions (ILO manual. Geneva: ILO). Jackson, S. E., Schuler, R. S. and Jiang, K. (2014), ‘An Aspirational Framework for Strategic Human Resource Management’, The Academy of Management Annals, 8, 1–​56. Kalleberg, A. L. (2011), Good Jobs, Bad Jobs: The Rise of Polarized and Precarious Employment Systems in the United States, 1970s to 2000s (New York: Russell Sage Foundation). Kirchner, S. and Hauff, S. (2019), ‘How National Employment Systems Relate to Employee Involvement: A Decomposition Analysis of Germany, the UK, and Sweden’, Socio-​Economic Review, 17, 627–​650. Knox, A., and Warhurst, C. (2015). ‘More than a Feeling: Using Hotel Room Attendants to Improve Understanding of Job Quality’, International Journal of Human Resource Management, 26, 1547–​1567. Knox, A., and Wright, S. (2022), ‘Understanding Job Quality Using Qualitative Research’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press). Korpi, W. (1978), ‘Workplace Bargaining, the Law and Unofficial Strikes: The Case of Sweden’, British Journal of Industrial Relations, 16, 355–​368. Korpi, W. (2006), ‘Power Resources and Employer-​Centered Approaches in Explanations of Welfare States and Varieties of Capitalism: Protagonists, Consenters, and Antagonists’, World Politics, 58, 167–​206. Lane, C., and Wood, G. (2012), Capitalist Diversity and Diversity within Capitalism (London: Routledge). Leschke, J., and Watt, A. (2008), ‘Job Quality in Europe’, Working Paper No. 2008.07. Brussels: ETUI.

understanding differences and trends in job quality    105 Leschke, J., and Watt, A. (2014), ‘Challenges in Constructing a Multi-​Dimensional European Job Quality Index’, Social Indicators Research, 118, 1–​31. Leschke, J., Watt, A., and Finn, M. (2012), ‘Job Quality in the Crisis –​An Update of the Job Quality Index (JQI)’, Working Paper No. 2012.07. Brussels: ETUI. Lorenz, E., and Valeyre, A. (2005), ‘Organisational Innovation, Human Resource Management and Labour Market Structure: A Comparison of the EU-​15’, Journal of Industrial Relations, 47, 424–​442. March, J. G. (1994), Primer on Decision Making: How Decisions Happen (New York: Free Press). Muñoz-​ de-​ Bustillo, R., Fernández-​ Macías, E., and Antón, J.-​ I. (2022),‘Quantitative Approaches to Assessing Jobs’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press). Muñoz de Bustillo, R., Fernández-​Macías, E., Antón, J.-​I., and Esteve, F. (2011a), Measuring More Than Money: The Social Economics of Job Quality (Cheltenham: Edward Elgar). Muñoz de Bustillo, R., Fernández-​Macías, E., Esteve, F., and Antón, J.-​I. (2011b), ‘E pluribus unum? A Critical Survey of Job Quality Indicators’, Socio-​Economic Review, 9, 447–​475. OECD (2014), OECD Employment Outlook 2014 (Paris: OECD Publishing). Olsen, K. M., Kalleberg, A. L., and Nesheim, T. (2010), ‘Perceived Job Quality in the United States, Great Britain, Norway and West Germany, 1989-​2005’, European Journal of Industrial Relations, 16, 221–​240. Osterman, P. (1987), ‘Choice of Employment Systems in Internal Labor Markets’, Industrial Relations, 26, 46–​67. Osterman, P., and Shulman, B. (2011), Good Jobs America: Making Work Better for Everyone (New York: Russell Sage Foundation). Peña-​Casas, R., and Pochet, P. (2009), Convergence and Divergence of Working Conditions in Europe: 1990-​ 2005 (Luxembourg: Office for Official Publications of the European Communities). Reiche, B. S., Lee, Y., and Quintanilla, J. (2012), ‘Cultural Perspectives on Comparative HRM’, in C. Brewster and W. Mayrhofer, eds., Handbook of Research on Comparative Human Resource Management (Cheltenham: Edward Elgar), 51–​68. Richter, N.F., Strandskov, J., Hauff, S., and Taras, V. (2022), International Business Strategy and Cross-​cultural Management: An Applied Approach (Cheltenham: Edward Elgar). Rubery, J., Bosch, G., and Lehndorff, S. (2008), ‘Surviving the EU?: The Future for National Employment Models in Europe’, Industrial Relations Journal, 39, 488–​509. Smith, M., Burchell, B., Fagan, C., and O’Brien, C. (2008), ‘Job Quality in Europe’, Industrial Relations Journal, 39, 586–​603. Streeck, W., and Thelen, K. (2005), ‘Introduction: Institutional Change in Advanced Political Economies’, in W. Streeck and K. Thelen, eds., Beyond Continuity. Institutional Change in Advanced Political Economies (Oxford: Oxford University Press), 1–​39. Tangian, A. (2007), ‘Is Work in Europe Decent? A Study Based on the 4th European Survey of Working Conditions 2005’, WSI-​Diskussionspapiere No. 157. Düsseldorf: WSI. Tangian, A. (2009), ‘Decent Work: Indexing European Working Conditions and Imposing Workplace Tax’, Transfer: European Review of Labour and Research, 15, 527–​556. Taras, V., Steel, P., and Kirkman, B. L. (2012), ‘Improving National Cultural Indices Using a Longitudinal Meta-​Analysis of Hofstede’s Dimensions’, Journal of World Business, 47, 329–​341. Thelen, K. (2012), ‘Varieties of Capitalism: Trajectories of Liberalization and the New Politics of Social Solidarity’, Annual Review of Political Science, 15, 137–​159.

106    sven hauff and stefan kirchner Tilly, C. (1997), ‘Arresting the Decline of Good Jobs in the USA?’ Industrial Relations Journal, 28, 269–​274. Triandis, H. C. (1994), ‘Cross-​Cultural Industrial and Organizational Psychology’, in M. D. Dunnette, H. C. Triandis, and L. M. Hough, eds., Handbook of Industrial and Organizational Psychology (Palo Alto, Calif.: Consulting Psychologists Press), 103–​172. Vaiman, V., and Brewster, C. (2014), (How Far Do Cultural Differences Explain the Differences between Nations? Implications for HRM’, The International Journal of Human Resource Management, 26, 151–​164. Valeyre, A., Lorenz, E., Cartron, D., Csizmadia, P., Gollac, M., Illéssy, M., and Makó, C. (2009), Working Conditions in the European Union: Work Organisation (Luxembourg: Office for Official Publications of the European Communities).

chapter 5

understanding job quality using qualitative research angela knox and sally wright

Introduction Globally, the ILO is promoting the importance of ‘decent work’ in achieving sustainable development. For example, during the UN General Assembly in September 2015, decent work and the four pillars of the ILO’s Decent Work Agenda—​employment creation, social protection, rights at work, and social dialogue—​were incorporated into the new 2030 Agenda for Sustainable Development. Promoting more and better jobs throughout the European Union is also the main aim of the European Employment Strategy; with this aim constituting part of the European Union’s Europe 2020 strategy for smart, sustainable and inclusive growth. In addition to the focus on job quality among policymakers, there is a long history of research and debate on job quality among academics. Over recent decades, much of the academic debate on job quality has centred on whether recent changes in the employment structure due to such factors as deindustrialization coupled with growth in services, increased female participation and technological change have generated more or fewer good jobs. In this chapter, we examine how qualitative research has progressed our understanding of job quality and how it might be used to improve the conceptualization and subsequent operationalization of this important phenomenon. Existing qualitative research has produced critical insights regarding key aspects of job quality, including the consequences of job quality, changes in job quality and the concept of job quality. Yet, we believe that completing the key tasks of conceptualization and operationalization will require additional qualitative research along with subsequent quantitative research. Moreover, we argue that these tasks will require distinguishing explicitly between the objective, subjective and integrated components of job quality. Distinct reference to and articulation of these three separate constructs will facilitate more precise conceptualization of job quality and enable more effective operationalization.

108    angela knox and sally wright

Job Quality: What is it and Why does it Matter? Broadly, job quality refers to the work and employment characteristics of a job, including: pay and benefits, work organization, training and skill development opportunities, career progression opportunities and worker autonomy, participation and representation (Knox et al. 2011). Extrapolating on these broad notions, researchers tend to consider high-​quality jobs, or so called ‘good jobs’, to be those offering financial and employment security while allowing workers to develop and utilize their skills with a degree of autonomy and control in addition to giving workers opportunities to have their voices heard, to participate in decision-​making and to progress their careers (Findlay et al. 2013). In contrast, low-​quality jobs, or so called ‘bad jobs’, are thought to offer little or no financial or employment security while restricting workers’ ability to develop and utilize skills, providing workers with limited autonomy and control and few opportunities to have their voices heard, or to participate in decision-​making or to progress their careers. Labour market analysis usually focuses on the number of jobs: that is, job quantity. However, not all jobs are the same, so extending research to consider the quality of jobs is important because it helps better understand the process of creative destruction and how the quality, not just the quantity, of jobs changes (Muñoz de Bustillo et al. 2011a). Job quality is important because it affects attitudes, behaviour and outcomes at the individual, organizational and national level (Clark 2005; Gallie 2013; Loughlin and Murray 2013; Okay-​Sommerville and Scholarios 2013). Indeed, it is increasingly clear that job quality affects individual’s job satisfaction and commitment as well as physical and mental health and well-​being (Bohle et al. 2001; Burgard et al. 2009; Butterworth et al. 2011; Clark 2005; Emberland and Rundmo 2010; Gallie 2013; Loughlin and Murray 2013, Okay-​Summerville and Scholarios 2013, Quinlan and Bohle 2015). While an association between unemployment and ill health is well established, the impact of job quality on physical and mental health and well-​being is less well understood. As workforce downsizing and outsourcing are both ongoing and pervasive features of modern organizations and because a growing number of countries require the unemployed to accept any job, a salient question arises of whether there are some jobs that are so low in quality that remaining unemployed may be better, or at least no worse for the health and well-​being of an individual. Studies have found, for example, that job insecurity associated with poor quality jobs can diminish worker health and well-​being (Quinlan and Bohle 2015); an individual’s decision latitude is important not only for their psychological well-​being but also for their physical health (Gallie 2013). The individual outcomes of job quality also affect organizational performance along with social and national economic outcomes (European Commission 2002, Nolan and Whelan 2014, Siebern-​Thomas 2005). Evidence of this nature illustrates the specific benefits that can be reaped from efforts to improve job quality. Studies highlight, for

understanding job quality using qualitative research    109 example, that job satisfaction is negatively correlated with absenteeism (Clegg 1983) and turnover (Freeman 1978). Thus, enhancing employee job satisfaction, through the development of better-​quality jobs, can reduce the organizational costs associated with absenteeism and turnover substantially. Moreover, job quality and job satisfaction have been linked to labour productivity (Siebern-​Thomas 2005), such that improving job quality and job satisfaction will typically yield higher rates of labour productivity. Such thinking has found its way into management literature on ‘high performance’, ‘high involvement’ or ‘high commitment’ management (Gallie 2013), as improving the way that work is organized and providing employees with voice in the organization are both likely to bring about higher rates of labour productivity. The above findings correspond with research indicating that sectors and countries with above average job quality are more productive and innovative (Green 2006; Erhel and Guergoat-​Larivière 2016; Patterson et al. 1997, Muñoz de Bustillo et al. 2016, Toner 2011). As Green (2006) notes, better-​performing economies have better jobs, and increasing the proportion of better jobs can have individual, social and organizational benefits. Siebern-​Thomas (2005) has also highlighted how countries with improved job quality increased their national employment rate from 60 to 64 per cent and decreased their unemployment rate from 10 per cent to 6 per cent over a 10-​year period. From this perspective, job quality has the potential to deliver ‘win-​win-​win’ outcomes, producing private gains for individuals (through increased wages and well-​being), and firms (through increased productivity, greater innovation and reduced costs), as well as improving national economies (through increased employment participation and reduced unemployment) (Clark 2005, European Commission 2002, Siebern-​ Thomas 2005).

Contested Conceptual Terrain While the potential benefits associated with improving job quality are increasingly recognized, the precise definition of job quality along with its conceptualization and operationalization remain the subject of debate (e.g. Appelbaum et al. 2003, Batt et al. 2003, Clark 2005, Handel 2005; Findlay et al. 2013, Green 2006; Kalleberg 2011; Osterman and Shulman 2011, Muñoz de Bustillo et al. 2011a, 2011b; Warhurst et al. 2017). As Muñoz de Bustillo and his colleagues (2011b) note, job quality is ‘elusive’ because ‘it is one of those concepts . . . which everyone understands yet it is difficult to define precisely’ (p. 4). Loughlin and Murray (2013) observe that while there is revitalized interest in job quality, integration across disciplines has been lacking, due to differences regarding how to conceptualize and operationalize job quality and so measure it (Warhurst and Knox 2015; Wright 2015). Conceptual problems also remain because a plethora of sometimes distinct, sometimes overlapping terms are used to encompass and label job quality. For example, Warhurst and his colleagues (2017) identified ten commonly used terms in the job

110    angela knox and sally wright quality literature: decent work, fair work, the quality of working life (QWL), good work, well-​being and work, fulfilling work, meaningful work and job quality (within which the quality of employment and the quality of work are sometimes analysed separately). They observe that the scope of this ‘family of concepts’ ranges from a focus on workers (e.g. his/​her well-​being) to the immediate worksite (e.g. QWL), to the enterprise (e.g. fair work) and to the economy and society (e.g. good work) (see also Warhurst et al., ‘Job Quality: A Family Affair’ in this volume). Another reason for conceptual problems stems from the range of different disciplines which have been active in this field. While the concept of job quality has its origins in the social sciences, it has also been the focus of research among academics coming from a range of different fields and disciplinary traditions, including economics, political economy, psychology, gender studies, labour relations, geography and health and safety (Wright 2015). The background of researchers influences the focus of their research as well as the methods they employ. For example, orthodox economists tend to focus on pay and use quantitative research methods. In contrast, traditional sociologists typically consider issues associated with skill, autonomy and job content and may use qualitative, quantitative or mixed methods of research. A third reason for the lack of conceptual clarity relates to tension between the use of individualized measures of job quality versus more generalizable measures of ‘core’ job characteristics (Loughin and Murray 2013). Factors including the increase in female participation in the labour force as well as growth in new forms of work (such as outsourcing, freelancing, contracting, portfolio work, ‘gig’ work, telework, homework) have resulted in calls for more ‘individualized’ measures of job quality that take into account variations in scheduling, part-​time options and remote working. On the other hand, the desirability of comparing and contrasting variations in job quality among and between countries has led to calls for a ‘core’ set of dimensions of job quality. While it is not surprising to find that the background of researchers influences the focus of their research as well as the type of methods they employ, researchers’ and policymakers’ ability to take advantage of the potential benefits that flow from ‘good jobs’ is hampered by the continuing absence of a unified, accepted definition of job quality. Analysis of existing literature reveals that job quality is conceptualized in at least three different ways, though few researchers refer explicitly to the approach taken. The most common approach relies on conceptualization of job quality by way of objective dimensions. This approach involves focusing on the characteristics of the job, in the form of economic and/​or non-​economic indicators. For example, Osterman and Shulman (2011) rely on the economic indicators/​characteristics of the job by measuring pay, primarily because of the availability and ease of obtaining such data. Other studies rely on non-​economic (but still objective) indicators of the job. For example, Hackman and Oldham’s (1975) seminal job diagnostic survey measures skill variety, task identity, task significance, autonomy and feedback. In comparison, Kalleberg (2011) includes both economic (pay, benefits, etc.) and non-​economic indicators (autonomy, control over work, scheduling, job termination, etc.) in order to provide a more

understanding job quality using qualitative research    111 holistic assessment of the objective aspects of job quality. An ongoing problem with this approach is the inconsistency in indicators used by different researchers. A second approach relies on gathering workers’ subjective accounts of job quality. Within this approach, researchers accept that subjectivity is affected by workers’ characteristics such as sex, age, ethnicity, qualifications and socio-​economic background (Brown et al. 2007; Sledge et al. 2008). Thus, perceptions of good and bad jobs may vary in accordance with job-​holder characteristics. Two main sub-​categories of the subjective approach are found in the literature: using job satisfaction as an indicator of the quality of jobs and asking workers about various aspects of their jobs (Muñoz de Bustillo et al. 2011b). With the former (using job satisfaction as an indicator of job quality), it is assumed that high-​quality jobs will lead to high job satisfaction (Brown et al. 2012, Clark 2011, Green et al. 2010, Kalleberg 2011). This approach is often criticized. Muñoz de Bustillo and his colleagues (2005, 2011a), for example, identify four problems with using (subjective) job satisfaction as a global measure of job quality. First, the lack of variation in job satisfaction within and between countries is low and does not appear congruent with conditions of work. Secondly, important differences in job quality do not exert a significant difference on reported job satisfaction. Thirdly, while job satisfaction might be related to job quality, there are many unrelated and endogenous variables affecting the level of job satisfaction, undermining the potential role of this variable as an overall indicator of job quality. Finally, using job satisfaction to measure job quality has limited utility for policy and scientific purposes because it is not possible to decompose the multidimensional construct of job quality to understand what attributes of work are responsible for overall job quality. With this sub-​category, the problem lies not in subjectivity, but rather in the use of a single, global output measure. The other sub-​category of the subjective approach involves asking workers about various aspects of their jobs. Research using this approach has focused on the reported attitudes and experiences of the job-​holder and the extent to which the job meets the job-​holder’s needs (Brown et al. 2007; Eurofound 2012; Sutherland, 2011). Information gathered about which attributes are considered important by workers when defining a ‘good job’ helps in the understanding of job quality as a multidimensional construct (Muñoz de Bustillo et al. 2011b), but while this approach gives workers a voice in the definition of what makes a good job, disagreement often results about the relative importance given to the different aspects of a job. A third approach relies on combining or integrating both objective and subjective dimensions in order to conceptualize and operationalize job quality. For example, job quality research by Holman (2013a) was based on five objective—​economic and non-​ economic—​indicators (work organization, wages and payment system, security and flexibility, skills and development, and engagement and representation), which he then validated using subjective indicators. Other studies have also encompassed objective (economic and non-​economic) and subjective dimensions. Green et al. (2010), for instance, used (economic and non-​economic) objective dimensions of job quality, including pay and training opportunities, and subjective dimensions, including satisfaction with pay, hours and the job. Such measures were used to assess the ‘subjective

112    angela knox and sally wright characteristics of the employment environment’, thereby reflecting job-​holders’ subjective attitudes and experiences of the job (Eurofound 2012; Green et al. 2010: 617; also Brown et al. 2007). Also using what could be termed an integrated approach, some researchers have used the subjective ratings to evaluate the objective attributes of jobs. For instance, Jencks et al. (1988) asked US workers to rate monetary and non-​monetary job characteristics and then used the results to weight all job characteristics according to their effects on workers’ judgements about how ‘good’ their current jobs were compared with an average job. Relatedly, the potential importance of job-​holder preferences has also been noted (Brown et al. 2007, Knox et al. 2015, Loughlin and Murray 2013; Sutherland 2011), though further empirical analysis/​support remains absent. Although an integrated approach clearly has merits, such as taking account of jobholders’ perceptions, it remains contested, both in terms of its content and its validity. Lack of agreement regarding the content and validity of these differing approaches continues to be a significant problem, as emphasized by Green et al. (2010) and others (e.g. Holman 2013a, 2013b; Knox et al. 2015). Similarly, Findlay et al. (2013) noted the inconsistent and inadequate conceptualization and use of objective and/​or subjective dimensions or indicators of job quality. This unresolved issue impedes our understanding of what constitutes job quality and, with it, ‘good’ and ‘bad’ jobs (Green et al. 2010; Holman 2013a, 2013b; Knox et al. 2015). According to Knox et al. (2015: 1550), resolving this fundamental problem is essential to working out the ‘tension between the articulation of objective and subjective dimensions of job quality that hinders better understanding of job quality’. We argue that existing problems related to the conceptualization and operationalization of job quality will only be resolved as further qualitative research is conducted, as outlined in the next section.

Reasons for and Benefits of Using Qualitative Research in Studying Job Quality While job quality research has deployed quantitative and/​ or qualitative research methods, the majority has relied on quantitative methods (e.g. Brown et al. 2007, Handel 2005, Holman 2013a, 2013b, Kalleberg 2011, Muñoz de Bustillo et al. 2011a, 2011b). From a methodological standpoint, quantitative and qualitative research methods tend to be utilised for different, though often complementary, purposes as quantitative methods such as surveys, elicit different responses than qualitative methods such as interviews (Brown et al. 2012, Walters 2005). Quantitative research methods tend to be used in a more confirmatory and positivistic manner, and typically deployed for deductive purposes involving theory testing (Creswell and Clark 2007, Phillimore and Goodson 2004). In contrast, qualitative research tends to be used in a more exploratory and

understanding job quality using qualitative research    113 interpretive manner and deployed for inductive purposes as it lends itself to theory building and development (Yin 2003). In studying job quality, researchers have highlighted various reasons for and benefits of using quantitative and/​or qualitative methods, as well as their associated limitations. For example, Green et al. (2015) emphasize that quantitative approaches enable the analysis of group-​wide and/​or nation-​wide trends while acknowledging that no one survey series can capture all aspects of job quality; for some the span of time is limited, for others key dimensions/​indicators of job quality (such as work-​life balance) are missing. Extending beyond these points, Sengupta et al. (2009: 27) highlight that quantitative methods involve ‘clear measures of job quality and representative data sets, but involve the limitation of not connecting directly with the concrete work experience of workers’, an issue echoed by Weaver (2009: 581) in noting the failure of quantitative research to ‘capture the richness of individuals’ experiences’. In contrast, qualitative researchers will often explore a relatively new and complex phenomenon (such as job quality) using qualitative research methods, enabling them to build theory, and they will subsequently deploy quantitative research methods in order to test such theories (Creswell and Clark 2007). The complementary nature of quantitative and qualitative methods has been noted by Green et al. (2015: 10): ‘The advantage of a quantitative approach is that we are able . . . to test whether hypotheses otherwise established for small well-​defined groups or in case studies are valid when generalised to the population at large.’ One of the strengths of qualitative research is that it is typically conducted under ‘real world’ conditions where contextual conditions are important (Yin 2011). Close researcher involvement means researchers are often able to gain an insider’s view of the phenomenon. Before the advent of large-​scale quantitative surveys, sociologists, in particular those from within the tradition of labour process theory, used qualitative research methods to gain an insider’s view of job quality (albeit they did not use the term ‘job quality’). For example, sociologist Lupton (1963) undertook case studies using real work groups (rather than experimental groups) to research working conditions on the shop floor. Lupton described his method of data collection as ‘open participant observation’. That is, the workers with whom he worked knew he was a researcher and were told of his purpose. In contrast, while undertaking an ethnographic study of work in a British motor components factory, Glucksmann (1982) went ‘undercover’ (aka Ruth Cavendish) to gather an insider’s view of the lived experience of low-​paid, mainly migrant women (along similar lines, was Pollert’s Girls, Wives, Factory Lives). More recently, Masterman-​ Smith and Pocock (2008) gathered personal work and life histories from 92 low-​paid workers in Australia, shedding light on what they term a ‘darker side’ of the economic prosperity reported by other studies. These kinds of research methods allow participants to choose their own words and respond in detail, thus helping the researcher to develop a real sense of how workers perceive the quality of their jobs and how they sit within the broader context of the organization, wider labour market and society. Another strength of qualitative research is that rather than relying on a single source alone, the research involves collecting data from multiple sources, such as structured,

114    angela knox and sally wright semi-​structured or unstructured interviews, focus groups, documents, participant observation, diary accounts and ethnography. The collection of evidence from different sources is commonly used to combat threats to validity (Yin 2011). Triangulation—​ seeking at least three ways of verifying or corroborating information—​is used to strengthen the validity of qualitative research. For instance, triangulation was used by Oxfam Scotland, where the researchers combined one-​to-​one interviews, focus groups, street stalls and an on-​line survey to answer the research question of ‘what makes decent work’ for low-​paid workers in Scotland (Stuart et al. 2016). While the design of quantitative research is largely determined before it begins, with qualitative research it is usually possible to change or adjust the design as the research progresses and understanding of the subject matter evolves. Moreover, while quantitative research tends to be useful in obtaining data on ‘what’, it is not so good at obtaining data on ‘why’. For example, similar to other surveys, the European Working Conditions Survey (EWCS) includes a range of questions in which respondents are asked about various aspects of their job using closed questions and rating scales. However, the survey does not allow respondents to explain their choices or the meaning the questions may have for them. In contrast, qualitative research methods such as in-​depth interviews and focus groups allow researchers to probe participants about any important underlying factors that may influence their assessments of job quality. For instance, the flexible, informal, open-​ended interviews as used by Weaver (2009) and others, allow the exploration of more in-​depth perceptions of job quality, and provide the opportunity for interviewees to explain their views. According to Weaver, such approaches can also produce unexpected perspectives and insights, and open-​up new avenues for theory and research. Qualitative methods such as these enable greater understanding of subjective and contextual factors, Weaver states, but they face issues in terms of generalization (Sengupta et al. 2009).

How Can Qualitative Research Enhance Our Understanding of Job Quality? Job quality research involving qualitative methods has progressed our understanding of job quality in three key areas. These key insights relate to the consequences of job quality; changes in job quality; and, perhaps most importantly, the concept of job quality. We discuss these insights in turn.

Consequences of Job Quality Findings regarding the consequences of job quality are important because they enhance our understanding of outcomes associated with job quality, allowing us to assess and

understanding job quality using qualitative research    115 compare the outcomes associated with good and bad jobs, for example. In addition, such findings provide valuable evidence to compel key parties to focus on and enhance job quality. As pointed out by Weaver (2009: 579), for example: How employees perceive the quality of their jobs can influence their decision to pursue, accept, and sustain certain types of employment. Knowledge of employees’ perceptions of job quality has value to employers who seek to attract and retain competent employees.

More specifically, Weaver’s findings highlight that some tourism employees expressed feeling more inclined to seek alternative employment if their job was not ‘good’—​ defined by employees as jobs providing faster promotion prospects and career progression opportunities. Extending on this idea, Clarke’s (2015) research examining community aged-​care workers, similarly reveals how job quality impacts attraction, retention, job satisfaction and, in turn, firm performance. Similarly, within the healthcare sector, Morgan et al. (2013) used a mixed-​methods approach to examine frontline workers across a variety of healthcare settings. Their findings highlight how the intrinsic (e.g. meaningful tasks) and extrinsic (e.g. wages and benefits) elements of job quality shape job satisfaction, while the extrinsic elements (i.e. wages) determine employees’ intent to stay with their employer, helping us to understand why frontline workers can report high job satisfaction while, at the same time, intending to leave their jobs. Other research has gone further and illustrated how job quality impacts workers’ families and communities as well as the wider social fabric (Masterman-​Smith and Pocock 2008). Examining employees in the Australian mining industry, Peetz and Murray (2011) reported that involuntary long working hours and lack of employee control over hours produced adverse consequences for work–​ family balance; detrimentally affecting family life and personal relationships. Another important insight relates to understanding how workers cope with the consequences of job quality. For instance, Brown et al.’s (2012) research develops our understanding in relation to how workers cope with jobs that are of poor (objective) quality. In this case, Brown et al. (2012) refer to ethnographic studies by Roy (1959) and Burawoy (1979) that demonstrate how workers can generate job satisfaction despite poor (objective) job quality by developing systems and conventions of informal interaction, often in breach of work rules. Moreover, research by MacKenzie et al. (2006) indicates that physically demanding and dangerous work can engender a strong emotional attachment that defuses otherwise negative assessments made by workers.

Changes in Job Quality The second key insight relates to understanding changes in job quality and the factors that trigger or shape such changes. Research by Vallas and Prener (2012) used discourse analysis to highlight the structural and cultural forces that underpin changes

116    angela knox and sally wright in employment systems and the associated quality of jobs, for example. According to Kalleberg (2012: 432), such research helps to explain the growth and increased legitimacy of temporary jobs through the expansion of the temporary agency industry. Similarly, in developing his taxonomy of jobs and job quality, Vidal (2013) emphasized how qualitative research can be used to understand how the logic of externalization, or other employment logics, shapes managerial strategy and the related quality of jobs. Other research has pointed to the value of qualitative methods as a means of understanding how contextual factors shape and potentially change job quality. Using a combination of interviews and non-​participant observation of work processes, Sengupta et al. (2009) revealed how market (including product and labour markets) and work situations (including firm size) shape job quality. For example, within the manufacturing firms studied, relatively secure product market conditions engendered high levels of job security and little pressure to reduce labour costs. Yet, the labour market provided a strong supply of workers, keeping wages low. In addition, small firms were associated with the development of positive attitudes towards managers and loyalty to the firm. Critically, these findings enabled Sengupta et al. (2009: 51) ‘to show how and why particular types of jobs have combinations of good and bad features, and thus why polarization has not occurred’. Isolating how market and work situations (as well as other potential factors) impact job quality enables levers to be identified, which can be adopted by policymakers interested in improving job quality. Relatedly, Osterman and Shulman’s (2011) research examining weatherization jobs isolated key levers and choices made by the parties who ultimately determine the quality of these relatively new jobs. In doing so, Osterman and Shulman conclude that strong public commitment and political leadership is essential to creating and changing job quality. While local politics and government played an important role in shaping and enhancing job quality, federal government was said to play the most critical role in ‘pushing up job quality’ (p. 132) by developing a clear position and strong federal standards, primarily through regulation and enforcement. A key issue that remains to be redressed however, as Felstead et al. (2015) highlight, relates to the prevalence of neo-​liberal thinking over recent decades, discouraging governments from recognizing potential or actual problems and hindering their intervention in the activities of private property owners, businesses or multinationals. A shift in ideology may therefore be needed in order to spur the required regulation and enforcement necessary to improve job quality. In the UK, Simms’ (2017) research examining cleaners and healthcare assistants demonstrated the dynamic interaction between institutions and interests, and illustrated how trade unions could produce successes in relation to job quality issues. Based on these findings, she argued that ‘unions can and do respond to the interests of workers in bad jobs’ (p. 5) such that unions can subsequently influence institutions to represent new interests. Rather than possessing little power or agency, unions can create and pursue collective interests directed towards improving (and regulating) job quality by extending institutions of job regulation. Yet, Simms is also quick to highlight that the overall effectiveness of unions in affecting job quality changes depends at least in part on employer behaviour.

understanding job quality using qualitative research    117 In the US and Mexico, research studying immigrant construction workers has gone a step further and illustrated how workers ‘concerted engagement with, and modification of, the labour process at the worksite . . . enabled them to negotiate for job quality improvements’ (Iskander and Lowe 2013: 802). Critically, workers’ ability to wrest improvements depended on how production was organized locally, where it was more malleable, workers were able to gain more latitude to alter their working conditions and prospects for advancement, for example. This research usefully highlights the potential importance of worker power and agency in affecting job quality changes.

The Concept of Job Quality The third key insight relates to the concept of job quality. While the concept of job quality remains contested, as discussed previously, extant qualitative studies contribute important insights and developments. According to Brown et al. (2007, 2012), job quality can only be adequately conceptualized by integrating both objective and subjective dimensions. Brown et al. (2012) argue that purely objective conceptualizations of job quality lack reference to the formation and impact of workers’ norms and expectations, said to be vital to understanding ‘the lived experience of work’ (p. 1011). This view echoes that of Brown et al.’s (2007) earlier work in which they argue that changes in objective job quality through time can only be understood when the subjective dimensions of job quality—​primarily the norms and expectations of job-​holders—​are integrated. The importance of integrating objective and subjective dimensions of job quality is also demonstrated by other research. Watson’s (2005) research examining job quality among casual/​temporary workers in Australia, illustrates how a reliance on assessing subjective dimensions of job quality, to the exclusion of objective dimensions, can conceal problematic—​objective—​job characteristics. In this case, the positive subjective views portrayed by casual workers masked ‘inferior’ (p. 371) objective characteristics of their jobs, primarily a significant pay deficit. This effect is also evident in Rose’s (1988) insights in relation to Goldthorpe et al.’s (1968) ‘affluent worker’ studies. Here, car assembly workers expressed broad satisfaction with their work while, at the same time, indicating that it was fragmented, monotonous, machine-​paced and lacking social interaction. Similarly, Brown et al. (2007) highlight how many workers reporting strong subjective perceptions of their jobs are, objectively, in low-​quality jobs. Apparently, these workers express satisfaction because of low norms and expectations. Conversely, jobs entailing high-​quality objective characteristics, including wages, job security and skill, can produce low subjective assessments, including low job satisfaction (e.g. Perlow 1999). Critically, as Green (2006) highlights, subjective indicators reflect two factors: (1) the facet of job quality itself; and (2) job-​holders’ norms and expectations regarding that facet of job quality. For example, asking a job-​holder to assess the level of effort that the job demands will involve the adoption of an effort norm or expectation. Higher norms or expectations, while holding the actual level of job quality constant, will typically produce a lower subjective measure of job quality. In contrast, higher actual job

118    angela knox and sally wright quality, while holding norms and expectations constant, will typically produce a higher subjective measure of job quality. According to Brown et al. (2007: 947): ‘The influence of norms and expectations implies that the level of any given facet of job quality cannot be directly read off from its subjective measure.’ Accordingly, Brown et al. (2012) argue that we need to develop better understanding of why workers feel satisfied or dissatisfied. Moreover, they suggest that qualitative research can be used to unravel ‘the complexity and diversity of experiences’ among job-​holders (p. 1008) and explain the apparent disconnection between subjective and objective dimensions of job quality. In this regard, Watson’s (2005: 383) assertions are noteworthy: To reconcile the contentment of casual workers with their location within inferior jobs requires much more incisive research into the construction of worker subjectivity, a task that will need to move beyond simply looking at expectations. Such an agenda is more likely to bear fruit through in-​depth qualitative research, rather than through attitudinal items in quantitative surveys. The field of life history and work history analysis . . . is well suited to this agenda.

Indeed, research related to life history and work history indicates that workers assess their current jobs relative to their previous employment experiences (e.g. Walters 2005), their life stage (e.g. Belardi et al. 2021a, 2021b; Cooke et al. 2013; Pocock et al. 2012) and their alternative employment options (e.g. Knox et al. 2015) for example. In this respect, research examining temporary/​casual teachers in Australia has also illustrated the potential importance of employment status. Bamberry’s (2011) study revealed the inferior subjective experiences of casual teachers working in otherwise objectively good jobs, and therefore noted the importance of examining both objective and subjective job quality dimensions as the subjective experiences of workers performing the same (objective) jobs may not be uniform. In comparison, through in-​ depth interviews exploring employees’ personal expectations and experiences, Weaver’s (2009) qualitative research examining employees in the tourism industry revealed how employees’ preferences and personality influenced their perceptions of job quality. While research in the tourism industry generally reports poor objective job quality, involving low pay, monotonous tasks, minimal job security and low status (e.g. Baum 2006; Wood 1997), Weaver’s research illustrates very different perceptions among job-​holders themselves. Employees perceived their jobs to be ‘good’ because they aligned with their preferences and personality. Tourism jobs enabled job-​holders to consume tourism products, provide assistance to tourists and learn about the tourism industry, for example; these aspects of the job aligned with job-​holders’ preferences and personalities and were therefore valued. Research examining job quality among those employed in community aged care similarly highlights how job-​holders’ perceptions vary in accordance with their needs, life stage and even community views regarding the value of their work, leading Clarke (2015: 203) to argue that:

understanding job quality using qualitative research    119 Individual perceptions of job quality are embedded in a social context so that specific job characteristics assume relative importance depending on factors such as age, preferred lifestyle and domestic circumstances.

Other research has integrated and built on these pre-​existing conceptions of job quality by analysing the objective dimensions of the job as well as the subjective experiences and perceptions of job-​holders (Knox et al. 2015). Importantly, this research confirms that subjective experiences of jobs can vary. Different workers in the same job can perceive that job to be good or bad. Knox et al. conclude that these subjective assessments are shaped by workers’ characteristics and preferences. For example, Knox et al. illustrate how young female workers with similar qualifications can have different experiences and perceptions of the same jobs because they seek to fulfil different needs from that job. Some workers expressed a preference for the job because it aligned with their family commitments; other workers expressed an aversion to the job because it did not facilitate their pleasure-​focused social lives. Therefore, Knox et al. (2015: 1561) argue that ‘[s]‌ubjective assessments of job quality, therefore, depend on workers characteristics and preferences: who workers are as well as what they want.’ Perhaps most importantly, Knox et al.’s research demonstrates how the objective and subjective dimensions of job quality can be integrated: This conceptual categorization (see Figure 5.1) allows us to understand how objectively good or bad jobs can be assessed subjectively as ‘fulfilling’ or ‘unfulfilling’, depending on who workers are and what they want. Based on this categorization, ‘fulfilling good jobs’ are objectively good and they align with workers’ characteristics and preferences so they Subjective Job Quality +

+

Fulfilling Good Jobs

-

Fulfilling Bad Jobs

-

Unfulfilling Good Jobs

Objective Job Quality Unfulfilling Bad Jobs

Figure 5.1  Job quality categorization Source: Knox et al. 2015

120    angela knox and sally wright are assessed subjectively as ‘fulfilling’, while ‘unfulfilling good jobs’ are objectively good but they do not align with workers’ characteristics and preferences so they are assessed subjectively as ‘unfulfilling’. In comparison, ‘fulfilling bad jobs’ are objectively bad but they align with workers’ characteristics and preferences so they are assessed as ‘fulfilling’ jobs, whereas ‘unfulfilling bad jobs’ are objectively bad and they inadequately align with workers’ characteristics and preferences so they are assessed as ‘unfulfilling’ jobs. Using qualitative methods to examine job quality has created critical explanatory information regarding the consequences of job quality, changes in job quality and the concept of job quality. Such findings allow us to better understand the lived experiences of job-​holders and identify factors that produce changes in job quality and how such changes might be experienced by workers, while also demonstrating the importance of distinguishing between objective and subjective aspects of job quality. Findings of this type allow identification of the job quality levers and development of policy responses directed towards ameliorating potentially detrimental outcomes, such as turnover, at job, firm and/​or industry level. Moreover, qualitative methods highlight the importance of differentiating between objective and subjective dimensions of job quality. As per Brown et al. (2007) and Knox et al. (2015), we argue that it is then necessary to create an integrated approach, involving objective and subjective dimensions, in order to assess overall job quality. In order to achieve better understanding and conceptualization of job quality it is useful to briefly examine two intimately related terms used in psychology, concepts that may prove useful, namely: sensation and perception (see for example, Goldstein 2010). Sensation and perception perform two complementary but different roles in how we understand the world. Sensation refers to the process related to sensing our environment (through sight, sound, smell, touch and taste), based on raw—​objective—​information received by our brain. Perception refers to the process related to interpreting those sensations, thereby making—​subjective—​sense of them. Thus, the objective approach, as outlined, encapsulates the ‘sensation’-​related aspects of the job, while the subjective approach encapsulates the ‘perception’-​related aspects of the job. Illustratively, a job may entail low or minimal discretion when assessed on the basis of (objective) sensation-​related aspects of the job, whereas the job may entail adequate discretion when assessed on the basis of (subjective) perception-​related aspects of the job, that is, the job-​holder’s perception. Using this type of analytical framework, an integrated or holistic approach would encompass objective and subjective dimensions, combining the sensation-​and perception-​related aspects of the job. In this respect, such research would help in better understanding the matching process (Kalleberg 2008) with workers and their jobs.

Conclusions In this chapter we have highlighted how research on job quality is important because it helps better understand the process of creative destruction and how the quality, not just

understanding job quality using qualitative research    121 the quantity, of jobs change. A review of existing evidence has also demonstrated that industries and countries with above average job quality are more productive and innovative, meaning improvements in job quality have the potential to deliver ‘win-​win-​ win’ outcomes for individuals, firms and national economies. Yet while the potential benefits associated with improving job quality are increasingly recognized, researchers’ and policymakers’ ability to take advantage of the potential benefits that flow from ‘good jobs’ is hampered by a lack of a precise definition of job quality. Conceptual problems remain where a plethora of terms have been used to label job quality and researchers have approached job quality from a range of different disciplinary traditions. Within the existing job quality research, the majority has relied on quantitative methods, yet we have shown how qualitative research has progressed our understanding of the consequences of job quality, changes in job quality, and perhaps most importantly, the concept of job quality. As shown, qualitative research lends itself well to more exploratory and inductive purposes. Because it is conducted under ‘real world’ conditions, qualitative researchers have been able to provide important insights into how workers perceive the quality of their jobs, and how they sit within the broader context of the organization, wider labour market and society. As such, qualitative research has contributed to the development of policies that specifically address and improve the quality of jobs. Moreover, qualitative research has proved critical to the development of new conceptual understandings of job quality, a task that is essential to improving hitherto incomplete conceptual and theoretical developments related to job quality. This work needs to continue however, as effective conceptualization and operationalization of job quality has not yet been achieved, thereby impeding subsequent empirical research on job quality. What is now needed is a new research agenda that focuses on establishing the conceptual boundaries of the objective, subjective and integrated approaches to job quality. This research should initially involve qualitative methods in order to effectively establish the conceptual boundaries of these constructs, which should then be tested using quantitative methods. We argue—​based on existing qualitative research—​that it is essential to explicitly and conceptually recognize the importance and relative contributions of these three, quite distinct, constructs. Greater clarity regarding job quality will only be achieved if we conceptualize and operationalize this concept using more precise terms, and we believe that the first step involves explicit reference to and articulation of its objective, subjective and integrated components. Achieving greater clarity regarding job quality will also require researchers to bridge interdisciplinary boundaries in order to progress the research agenda.

References Appelbaum, E., Bernhardt, A., and Murnane, R. (eds.) (2003), Low-​Wage America (New York: Russell Sage Foundation).

122    angela knox and sally wright Bamberry, L. (2011), ‘As Disposable of the Next Tissue out of the Box: Casual Teaching and Job Quality in New South Wales Public School Education’, Journal of Industrial Relations, 53, 1, 49–​64. Batt, R., Hunter, L., and Wilk, S. (2003), ‘How and When Does Management Matter? Job Quality and Career Opportunities for Call Centre Workers’, in Appelbaum, E., Bernhardt, A., and Murnane, R., eds., Low-​Wage America (New York: Russell Sage Foundation), 270–​316. Baum, T. (2006), Human Resource Management for Tourism, Hospitality and Leisure: An International Perspective (London: Thomson Learning). Belardi, S., Knox, A., and Wright, C.F. (2021a), ‘Too Hot to Handle? An Analysis of Chefs’ Job Quality in Australian Restaurants’, Journal of Industrial Relations, 63, 3–​26. Belardi, S., Knox, A., and Wright, C.F. (2021b), ‘‘The Circle of Life’: The Role of Life Course in Understanding Job Quality’, Economic and Industrial Democracy, DOI: 10.1177/​ 0143831X211061185. Bohle, P, Quinlan, M., and Mayhew, C. (2001), ‘The Health and Safety Effects of Job Insecurity: An Evaluation of the Evidence’, Economic and Labour Relations Review, 12, 1, 32–​60. Brown, A., Charlwood, A., Forde, C., and Spencer, D. (2007), ‘Job Quality and the Economics of New Labour: A Critical Appraisal Using Subjective Survey Data’, Cambridge Journal of Economics, 31, 5, 941–​971. Brown, A., Charlwood, A., and Spencer, D. (2012), ‘Not All It Might Seem: Why Job Satisfaction Is Worth Studying Despite It Being a Poor Summary Measure of Job Quality’, Work, Employment and Society, 26, 6, 1007–​1018. Burawoy, M. (1979), Manufacturing Consent: Changes in the Labour Process under Monopoly Capitalism (Chicago: University of Chicago Press). Burgard S., Brand, J., and House, J. (2009), ‘Perceived Job Insecurity and Worker Health in the United States’, Social Science & Medicine, 69, 5: 777–​785. Butterworth, P., Leach, L., Strazdins, L., Oleson, S., Rodgers, B., and Broom, D. (2011), ‘The Psychosocial Quality of Work Determines Whether Employment Has Benefits for Mental Health: Results from a Longitudinal National Household Panel Survey’, Occupational & Environmental Medicine, 68, 11, 806–​812. Clark, A. (2005), ‘What Makes a Good Job? Evidence from OECD Countries’, in S. Brazen, C. Lucifora, and W. Salverda, eds., Job Quality and Employer Behaviour (New York: Palgrave Macmillan), 11–​30. Clark, A. (2011), ‘Worker Wellbeing in Booms and Busts’, in P. Gregg, and J. Wadsworth, (eds), The Labour Market in Winter: The State of Working Britain (Oxford: Oxford University Press), 128–​143. Clarke, M. (2015), ‘To What Extent a ‘Bad’ Job? Employee Perceptions of Job Quality in community aged care’. Employee Relations, 37, 2, 192–​208. Clegg, C. (1983), ‘Psychology of Employee Lateness, Absence and Turnover: A Methodological Critique and an Empirical Study’ Journal of Applied Psychology, 88, 52–​72. Cooke, G., Donaghey, J., and Zeytinoglu, I. (2013), ‘The Nuanced Nature of Work Quality: Evidence from Rural Newfoundland and Ireland’, Human Relations, 66/​4, 503–​527. Creswell, J., and Clark, V. (2007), Designing and Conducting Mixed Methods Research (New York: Sage). Emberland, J. S., and Rundmo, T. (2010), ‘Implications of Job Insecurity Perceptions and Job Insecurity Responses for Psychological Wellbeing, Turnover Intentions and Reported Risk Behaviour’, Safety Science, 48/​4, 452–​459.

understanding job quality using qualitative research    123 Erhel, C., and Guergoat-​Larivière, M. (2016), ‘Innovation and Job Quality Regimes: A Joint Typology for the EU’, QuInnE Working Paper, WP5-​2-​2016, http://​bry​der.nu/​quin​ne1/​sites/​ defa​ult/​files/​WP5_​2​_​251​1201​6_​fi​nal.pdf. Eurofound (2012), Trends in Job Quality in Europe (Luxembourg). European Commission (2002), ‘Employment in Europe in 2002’, in Synergies Between Quality and Quantity in European Labour Markets (Brussels: DG Employment). Felstead, A., Gallie, F., Green, F., and Inanc, H. (2015), ‘Fits, Misfits and Interactions: Learning at Work, Job Satisfaction and Job-​Related Well-​Being’, Human Resource Management Journal, 25, 3, 294–​310. Findlay, P., Kalleberg, A. L., and Warhurst, C. (2013), ‘The Challenge of Job Quality’, Human Relations, 66/​4, 441–​451. Freeman, R. (1978), ‘Job Satisfaction as an Economic Variable’, American Economic Review, 68/​ 1, 135–​152. Gallie, D. (2013), ‘Direct Participation and the Quality of Work’, Human Relations, 66/​4, 453–​473. Goldstein, E. (2010), Sensation and Perception (California: Wadsworth). Goldthorpe, J., Lockwood, D., Bechhofer, F., and Platt, J. (1968), The Affluent Worker: Industrial Attitudes and Behaviour (Cambridge: Cambridge University Press). Glucksmann, M. (1982), Women on the Line (London: Routledge). Green, F. (2006), Demanding Work: The Paradox of Job Quality in the Affluent Economy (Princeton: Princeton University Press). Green, F., Felstead, A., and Gallie, D. (2015), ‘The Inequality of Job Quality’, in Felstead, A., Gallie, D., and Green, F., eds. Unequal Britain at Work (Oxford: Oxford University Press), 1–​21. Green, F., Kler, P., and Leeves, G. (2010), ‘Flexible Contract Workers in Inferior Jobs: Reappraising the Evidence’, British Journal of Industrial Relations, 48/​3, 605–​629. Hackman, J., and Oldham, G. (1975), ‘Development of the Job Diagnostic Survey’, Journal of Applied Psychology, 60/​1, 159–​170. Handel, M. (2005), ‘Trends in Perceived Job Quality, 1989 to 1998’, Work and Occupations, 32/​ 1, 66–​84. Holman, D. (2013a), ‘An Explanation of Cross-​National Variation in Call Centre Job Quality Using Institutional Theory’, Work, Employment & Society, 27/​1, 21–​38. Holman, D. (2013b), ‘Job Types and Job Quality in Europe’. Human Relations, 66/​4, 475–​502. Iskander, N., and Lowe, N. (2013), ‘Building Job Quality from the Inside-​Out: Mexican Immigrants, Skills and Jobs in the Construction Industry’, International Labour Relations Review, 66/​4, 785–​807. Jencks, C, Perman, L., and Rainwater, L. (1988), ‘What is a Good Job? A New Measure of Labor-​ Market Success’, American Journal of Sociology, 93/​6, 1322–​1357. Kalleberg, A. (2008), ‘The Mismatched Worker: When People Don’t Fit Their Jobs’, Academy of Management, 22/​1, 24–​40. Kalleberg, A. (2011). Good Jobs, Bad Jobs: The Rise of Polarized and Precarious Employment Systems in the United States, 1970s to 2000s (New York: Russell Sage Foundation). Kalleberg, A. (2012), ‘Job Quality and Precarious Work: Clarifications, Controversies and Challenges’, Work and Occupations, 39/​4, 427–​448. Knox, A., Warhurst, C., Nickson, D., and Dutton, E. (2015), ‘More than a Feeling: Using Hotel Room Attendants to Improve Understanding of Job Quality’, The International Journal of Human Resource Management, 26/​12, 1547–​1567.

124    angela knox and sally wright Knox, A., Warhurst, C., and Pocock, B. (2011), ‘Job Quality Matters’, Journal of Industrial Relations, 53/​1, 5–​11. Loughlin, C., and Murray, R. (2013), ‘Employment Status Congruence and Job Quality’. Human Relations, 66/​4, 529–​553. Lupton, T. (1963), On the Shop Floor: Two Studies of Workshop Organization and Output. International Series of Monographs on Social and Behavioural Sciences, Volume 2 (New York: The Macmillan Company). MacKenzie, R., Stuart, M., Forde, C., Greenwood, I., Gardiner, J., and Perret, R. (2006). ‘All that is Solid? Class, Identity and the Maintenance of a Collective Orientation amongst Redundant Steel Workers’, Sociology, 40/​6, 833–​852. Masterman-​Smith H., and Pocock B. (2008), Living Low Paid (Crows Nest: Allen & Unwin). Morgan, J., Dill, J., and Kalleberg, A. (2013), ‘The Quality of Healthcare Jobs: Can Intrinsic Rewards Compensate for Low Extrinsic Rewards’, Work, Employment and Society, 27/​5, 802–​822. Muñoz de Bustillo, R., and Fernández-​Macías, E. (2005), ‘Job Satisfaction as an Indicator of the Quality of Work’, Journal of Socio-​Economics, 34, 656–​673. Muñoz de Bustillo, R., Fernández-​Macías, E., Esteve, F., and Antón José-​Ignacio (2011a), ‘E pluribus unum? A Critical Survey of Job Quality Indicators’, Socio-​Economic Review, 9/​3, 447–​475. Muñoz de Bustillo, R., Fernández-​Macías, E., Antón, J.-​I., and Esteve, F. (2011b), Measuring More Than Money: The Social Economics of Job Quality (Cheltenham, UK, Northampton, MA: Edward Elgar). Muñoz de Bustillo, R., Grande, R., and Fernández-​Macías, E. (2016), ‘Innovation and Job Quality: An Initial Exploration’, QuInnE Working Paper, WP5-​1-​2016, http://​www.qui​nne. eu. (Lund: University of Lund). Nolan, B. and Wheelan, C. (2014), ‘The Social Impact of Income Inequality: Poverty, Deprivation and Social Cohesion’, in W. Salverda, B. Nolan, I. Checchi, I. Marx, A. McKnight, I. Toth, and H. van de Werfhorst, eds., Changing Inequalities in Rich Countries (Oxford: Oxford University Press), 1–​22. Okay-​Sommerville, B., and Scholarios, D. (2013), ‘Shades of Grey: Understanding Job Quality in Emerging Graduate Occupations’, Human Relations, 66/​4, 555–​575. Osterman, P., and Shulman, B. (2011), Good jobs America: Making Work Better for Everyone (New York: Russell Sage Foundation). Patterson, M., West, M., Lawthorn, R. and McKell, S. (1997). The Impact of People Management Strategies on Business Performance? Issues in People Management No. 22. (London: Institute of Personnel and Development). Peetz, D., and Murray, G. (2011), ‘You Get Really Old, Really Quick: Involuntary Long Hours in the Mining Industry’, Journal of Industrial Relations, 53/​1, 13–​29. Perlow, L. (1999), ‘The Time Famine: Toward a Sociology of Work Time’, Administrative Science Quarterly, 17: 3, 57–​81. Phillimore, J., and Goodson, L. (2004), Qualitative Research in Tourism: Ontologies, Epistemologies and Methodologies (London: Routledge). Pocock, B., Williams, P., and Skinner, N. (2012), ‘Conceptualizing Work, Family and Community: A Socio-​Ecological Systems Model, Taking Account of Power, Time, Space and Life Stage’, British Journal of Industrial Relations, 50/​3, 391–​411. Pollert, A. (1981), Girls, Wives, Factory Lives (London: Macmillan).

understanding job quality using qualitative research    125 Quinlan, M., and Bohle, P. (2015), ‘Job Quality: The Impact of Work Organisation on Health’, in Knox, A. and Warhurst, C., eds., Job Quality in Australia: Perspectives, Problems and Proposals (Sydney: Federation Press), 63–​83. Rose, M. (1988), Industrial Behaviour (London: Penguin). Roy, D. (1959), ‘Banana Time: Job Satisfaction and Informal Interaction’, Human Organisation, 18/​1, 158–​168. Sengupta, S., Edwards, P., and Tsai, C. (2009), ‘The Good, the Bad and the Ordinary: Work Identities in Good and Bad Jobs in the United Kingdom’, Work and Occupations, 36/​1, 26–​55. Siebern-​Thomas, F. (2005), ‘Job Quality in European Labour Markets’, in S. Brazen, C. Lucifora, and W. Salverda, eds., Job Quality and Employer Behaviour (New York: Palgrave Macmillan), 31–​66. Simms, M. (2017), ‘Unions and Job Quality in the UK: Extending Interest Representation within Regulation Institutions’, Work and Occupations, 44/​1, 47–​67. Sledge, S., Miles, A., and Coppage, S. (2008), ‘What Role Does Culture Play? A Look at Motivation and Job Satisfaction among Hotel Workers in Brazil’, International Journal of Human Resource Management, 19/​7, 1667–​1682. Stuart, F., Pautz, H., and Wright, S. (2016), Decent Work for Scotland’s Low-​Paid Workers: A Job to Be Done (Glasgow: Oxfam Scotland). Sutherland, J. (2011), ‘Job Attribute Preferences in Scotland,’ Centre for Public Policy for Regions Working Paper 27, Centre for Public Policy for Regions. (Glasgow: University of Glasgow). Toner, P. (2011), Workforce Skills and Innovation: An Overview of Major Themes in the Literature (Paris: OECD). Vallas, S., and Prener, C. (2012), ‘Dualism, Job Polarization, and the Social Construction of Precarious Work’, Work and Occupations, 39/​4, 331–​353. Vidal, M. (2013). ‘Low Autonomy Work and Bad Jobs in Postfordist Capitalism’, Human Relations, 66/​4, 587–​612. Walters, S. (2005), ‘Making the Best of a Bad Job? Orientations and Attitudes to Work’, Gender, Work and Occupations, 12/​3, 193–​216. Warhurst, C., and Knox, A. (2015), ‘Why the Renewed Interest in Job Quality?’ in A. Knox and C. Warhurst, eds., Job Quality in Australia: Perspectives, Problems and Proposals (Sydney: Federation Press), 1–​14. Warhurst, C., Wright, S., and Lyonette, C. (2017), Understanding and Measuring Job Quality: Part 1 –​Thematic Literature Review (London: CIPD). Warhurst, C., Wright, S., and Mathieu, C. (2022), ‘Job Quality: A Family Affair?’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press). Watson, I. (2005), ‘Contented Workers in Inferior Jobs? Re-​Assessing Casual Employment in Australia’, Journal of Industrial Relations, 47/​4, 371–​392. Weaver, A. (2009), ‘Perceptions of Job Quality in the Tourism Industry’, International Journal of Contemporary Hospitality Management, 21/​5, 579–​593. Wood, R. (1997), Working in Hotels and Catering (London: Routledge). Wright, S. (2015), ‘Challenges in Researching Job Quality’, in A. Knox, and C. Warhurst, eds., Job Quality in Australia: Perspectives, Problems and Proposals (Sydney: Federation Press), 15–​36. Yin, R. (2003), Case Study Research: Design and Methods (London: Sage). Yin. R. (2011), Qualitative Research from Start to Finish (New York: The Guildford Press).

chapter 6

quantitative approaches to assess JOB QUALITY rafael muñoz de bustillo, enrique fernández-​macías and josé-​ignacio antón

Introduction For the average full-​time employee, working and related activities such as commuting absorbs well over 1/​5 of his or her working life. No wonder working conditions are a major determinant of overall well-​being on the working population (Dolan et al. 2008). This is reflected in a growing literature dealing with different aspects of job quality from diverse and complementary perspectives such as economics, psychology, sociology or health and safety studies. Although it has been always present in social sciences applied research,1 the subject of job quality has become more central in the last two decades due to concerns that the priority given to job creation, in a context of high and persistent unemployment, could lead to a deterioration of the quality of employment and the creation of a sizeable segment of low-​quality jobs (Warhurst et al. 2012). The emphasis on more, but also better, jobs of the (failed) EU Lisbon Strategy, or the importance placed by the subsequent Strategy Europe 2020 in ‘improving the quality of jobs and ensuring better working conditions’2 are good examples of such concerns at the EU policy level (Bothfeld and Leschke 2012). This growing interest of policymakers and academics in job quality is also shared by the labour movement, as shown by campaigns such as the annual 7 October Work Day for Decent Work sponsored by the International Trade Union Confederation.

1  An excellent example of the importance of this issue from the origins of industrial capitalism is the publication in 1845 of The Conditions of the Working Class in England (Engels 1891). 2  Agenda for New Skills and Jobs, flagship initiative of the Europe 2020 Strategy, EU Commission (2012).

quantitative approaches to assess JOB QUALITY    127 But the debate on job quality and its evolution is still hampered by the lack of a common understanding of what it is and how we can measure it. Lord Kelvin used to argue that ‘when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind’ (1889: 80–​81). The argument is obviously too radical (there can certainly be knowledge without measurement), but we would agree that first, having a more or less agreed measure on something facilitates considerable the debate; and second, the act of measurement itself imposes a certain discipline in terms of identifying precisely the phenomenon in question and establishing criteria for the comparability of its different instances, something which is particularly useful for the social sciences. This chapter intends to offer an up-​to-​date account of different alternative approaches to the measurement of job quality. In order to do so, in the next section we review different theoretical strategies that can be used for the measurement of job quality. Section 3 presents some of the methodological dilemmas that need to be confronted for such measurement. With that background, section 4 briefly surveys more than 20 proposals of indicators of job quality of different origins and ambitions. Finally, in section 5 we conclude with some reflections with respect to the feasibility of measuring job quality and the adequacy of the existing job quality indicators. But before engaging in a detailed discussion about how to measure job quality, it is useful to clarify the concepts of job quality and measurement. In this chapter, job quality refers to the potential impact of the characteristics of jobs on the well-​being of workers. This is a relatively open definition broadly consistent with most recent literature on this issue (Green 2006; Muñoz de Bustillo et al. 2011b), but which has some important implications which must be acknowledged: (1) it is a multidimensional concept; (2) it is anchored on objective attributes of jobs; (3) it focuses on characteristics rather than outcomes; and (4) it involves some kind of evaluation of what is good and bad in terms of the effect on the well-​being of workers. As for measurement, in this chapter it refers to the process of assigning numbers to different work situations in such a way that they reflect the previously mentioned evaluation of the potential (good or bad) impact of job characteristics on the worker’s well-​being, and that they can be compared across different types of employment, periods and countries.

Three Ways to Measure Job Quality A Spanish popular saying defends the view that ‘on tastes there is nothing written’: each person has a different taste, and none is intrinsically better than others. The same principle can be applied to job quality: different people might have different ideas about what makes a good job. Moreover, people might change their ideas (or tastes) about what is a good job as their circumstances change over time. This difficulty in the identification of

128    rafael muñoz de bustillo et al 9.0 8.0 7.0 6.8

7.0 7.0 7.1 7.1 7.1 7.1 7.2

7.2 7.3 7.4 7.4 7.4

7.4 7.5 7.5 7.5 7.5 7.5 7.5 7.5

7.8 7.8 7.8 7.9 7.9 7.5 7.6 7.7 7.7

7.9 8.1

8.2

8.4

6.0 5.0 4.0 3.0 2.0 1.0

So

ut

h

Ko r Fr ea an ce La Be tv lg ia iu m Jap /F lan an de Ru rs ss Cz T ec ai ia h wa Re n pu Sl blic ov en So Hun ia ut ga h ry Af r Sw ica Au eden str a Ca lia n Bu ada lg a Ne S ria w pa Ze in ala nd Gr ea To t B ta rit l P ain Ne ortu th ga er l la No nds Do rw m in ica Fin ay Ge n Re land rm pu an bli y- c W es Ge t rm Isr a Un ny ael ite -Ea d st St De ates nm a Cy rk pr u I r Ph ela s ili nd p Sw pi itz nes er la M nd ex ico

0.0

Figure 6.1  Average index of job satisfaction in 33 countries and territories, 2005 Source: Author´s analysis of ISSP 2005 data

what makes a good job must be somehow overcome in order to measure job quality. In this section, we review three alternative ways to address such a problem.

Job Satisfaction as an Indicator of Job Quality The first approach uses a smart shortcut intended to solve the problem of identifying what makes a good job. Instead of identifying and measuring the attributes that define the quality of a given job, the idea is to use the level of job satisfaction of the worker with her or his work as a proxy of the overall quality of the job. After all, job satisfaction can be understood as a general evaluation of how the current job affects the well-​being of the worker (by himself). This way, it is unnecessary to identify what attributes of jobs affect the worker’s well-​being, or how important each of them should be. It is enough to know how the worker evaluates the job in general terms, considering that no one knows better than the worker what is important for him or herself. Therefore, a high level of satisfaction with the job can be interpreted as an indication of good job quality (from the perspective of the specific interviewed individual worker) and averaging the level of satisfaction for particular types of jobs we can approximate and compare their quality. Figure 6.1 reproduces the average responses to the question, ‘How satisfied are you in your main job?’ (Q20) in a sample of 30 countries and three regions, obtained from the International Social Survey Program (ISSP) conducted in 2005 (ISSP Research

quantitative approaches to assess JOB QUALITY    129 Group 2013).3 Three things are worth mentioning from this chart: first, the overall high levels of job satisfaction existing in almost all countries: only in six out of the 33 countries and territories more than 10 per cent of workers are dissatisfied (completely or fairly) with their work (with a maximum of 12.5% in South Korea);4 second, the surprisingly low variation of the results, especially considering the wide existing differences between some of the countries of the sample in terms of economic structure, labour regulation and income level; and third, the seemingly odd ranking of countries in terms of job satisfaction, with Mexico at the top of the league, France in the next to last position and South Africa sharing a place with Sweden. An analysis of the distribution of results within countries shows that this is not explained by a higher dispersion in countries like South Africa or Mexico. These considerations cast doubt about the reliability of using job satisfaction as a proxy indicator of job quality (Muñoz de Bustillo and Fernández-​ Macías 2005). Such conclusion is buttressed by the analysis of the relation between job satisfaction and different job attributes, such as working time or wages, that should in principle be strongly related to job quality. There is plenty of literature, mainly in the area of social psychology, which relates job satisfaction with different aspects of the job, such as autonomy or stress (Spector 1997), usefulness of the work (Mangluione and Quinn 1975), functional flexibility (Orino and Pagani 2008), etc. Many of these studies do find an important degree of correlation between such characteristics and job quality. But these kinds of results have a significant methodological problem: if job attributes are measured through qualitative/​subjective indicators, there is a significant chance that they are endogenously linked to job satisfaction. When more objective indicators of job quality, such as wage, firm size or job stability (Clark and Oswald 1996; Brown and MacIntosh 1998; García Maynar 1999, Muñoz de Bustillo and Fernández-​Macías 2005) are used instead, the correlations are usually very low and the results not very conclusive, when not contradictory among themselves. Following Muñoz de Bustillo and Fernández-​Macías (2005), there are three different and complementary reasons that can explain the relatively high levels of job satisfaction obtained and its low variability regardless of the characteristics of the job. These mechanisms operate by aligning the expectations of the worker and the characteristics of jobs (since the mismatch between expectations and characteristics is what could lead to dissatisfaction). First, the expectations of workers are guidelines for their action: people try to find jobs that fit their expectations. Second, expectations tend to adapt to objective conditions. Since decades ago (following the seminal work of Festinger 1957), social psychologists have known of the existence of psychological mechanisms that help people to adapt their expectations and even their perception of the working environment to the actual conditions of the jobs. Finally, if there is still a significant mismatch 3  The possible responses go from completely satisfied to completely dissatisfied. The data presented in the figure is the result of transforming the categories into numbers (from 1 to 7) and rescaling the result to a 1–​10 scale. 4  The rest of the countries are Dominican Republic, Belgium/​Flanders, France, Japan and Latvia.

130    rafael muñoz de bustillo et al between expectations and actual conditions despite the two mechanisms previously mentioned, workers are likely to end up leaving their jobs.5 Together, these three factors make it very unlikely to observe a significant and persistent low level of job satisfaction, which would explain the high and concentrated values typically observed for this variable. The important role played by workers’ expectations and the mentioned processes of psychological adaptation in determining the degree of job satisfaction make it a very problematic indicator of job quality, regardless of its merit for other purposes.

Asking People What Makes a Good Job The second approach consists in presenting workers with a list of job attributes to be chosen and/​or ranked by importance: then, those answers are used to identify and weight the attributes, which are measured separately. This two-​step procedure is partially subjective, as the importance of the different facets of jobs in the measurement of job quality is obtained by asking workers themselves; but partially objective, as once we have the shopping list of the relevant characteristics of jobs according to workers (and their relative importance) we rely on objective indicators to actually measure job quality. This approach has the clear advantage of respecting workers’ own preferences, without relying on an overall satisfaction measure (which as previously discussed, may be ‘contaminated’ by expectations and psychological mechanisms of adaptation). But this approach also suffers from shortcomings. First, the pre-​selection of items to present to workers necessarily constrains their choices from the very beginning (in practice, the list of options is often relatively short in surveys using this approach). The second is related with the potential diversity of the valuation of job characteristics depending on personal, cultural and geographical characteristics. If we allow workers to select and allocate the importance of different job characteristics according to their personal or cultural preferences, we are likely to end up with many different models of what is a good job. This can be considered as the main advantage of this methodology (since it respects actual workers’ preferences), but it can make it extremely difficult to compare levels of job quality across different workers or identify the levels of quality of particular types of jobs. The results of such indicators can have very difficult interpretations, and the value for policy evaluation and design therefore significantly diminished. To illustrate this point, Figure 8.2 reproduces the average importance given by workers of different countries to different attributes of job quality according to the 2005 ISSP. The results show a very significant disparity in the importance attributed by different countries to attributes as important to employment policy as income or opportunities for advancement—​whereas interestingly, there is more consensus in the valuation of more qualitative aspects of job quality such as usefulness, which are much less relevant from

5 

In fact, job satisfaction is a good predictor of turnover intent (Freeman 1978, Lambert et al. 2001)

quantitative approaches to assess JOB QUALITY    131 90

Bulgaria

Bulgaria

80 France

70

Switzerland Bulgaria

60

Denmark

50

USA

Bulgaria

40 30 20 10 0

Finland Denmark Japan

Japan Switzerland

Job Security An interesting Useful to job society

Japan

Japan

Japan

High income Opportunities Work Help other Allows to decide the of independently people times or days advancement of work Max

Min

Average

Figure 6.2  Average and differences in importance given to job characteristics Source: Author´s analysis on ISSP 2005 data

a policy perspective. There are similar differences in the valuation of job characteristics according to social and demographic characteristics of people. Wrapping up, the use of this second approach to measure job quality has certain advantages worth acknowledging: it gives workers a say in the definition of what makes a good job; it facilitates the adaptation of the model of job quality to national specificities; and many existing surveys on quality of working life include questions about the desirability of specific job attributes. But the limitations are in our view even more important: it requires presenting workers with a predefined set of options (attributes of jobs to be ranked), necessarily leaving out elements that may be important for some workers; the adaptation of the model to different categories of workers seriously hampers the main goal of a job quality indicator, which is to be able to compare levels of job quality; and finally, it can complicate very significantly the interpretation of results and the discussion of their policy implications, since the scores will be strongly affected by subjective preferences which are outside the realm of work and the reach of policy. In our view, this approach can certainly be useful for informing the definition and measurement of job quality, but it cannot be its sole basis.

A Theory-​Based Job Quality Indicator As we mentioned in the introduction, there is a long tradition in the social sciences, starting from the contributions of the first political economists such as Adam Smith, of study of the impact of different elements of work and labour relations on workers. One

132    rafael muñoz de bustillo et al alternative way of selecting the relevant items to use in the definition and measurement of job quality is to use such literature as a guide. Table 6.1 reproduces a summary of the different areas and items of job quality considered important by different strands of social sciences, from fields as far as classical economics, medicine or health and safety studies. Due to space limitations, in this chapter we do not dwell on the different strands of job quality analysis summarized in Table 6.1, leaving the review of such developments to the chapter by Knox and Wright, ‘Understanding Job Quality Using Qualitative Research’, in this volume.

Some Methodological Dilemmas Involved in Job Quality Measurement Once the overall approach has been decided, and before embarking in the definition of a model of job quality and the identification of attributes to be measured, the researcher must confront a number of methodological and conceptual dilemmas with very significant implications in the final outcomes of the index. In this section, we briefly discuss those that we consider most important. Measuring the quality of jobs or the quality of labour markets? As we argue in the next section, the existing job quality indicators can be divided into two groups. The first one adopts a worker-​centred perspective, limiting the large or small number of indicators included in the index of job quality to those affecting the employment and working conditions of the worker. Alternatively, a second group of indicators widens their area of analysis to include elements such as the level of unemployment or the existence of unemployment benefits that reduce the cost of unemployment to the workers. From our perspective, the former option is to be preferred in order not to mix job quality with labour market quality, a wider concept that would include the level of employment and unemployment, the cost of being unemployed along with other elements such as the existence of complementary working tax credits. This last element can serve as a good example of the point we are trying to make. A low-​wage job would be a job with a caeteris paribus low quality from the perspective of a worker’s centred job quality index (at least in its wage dimension). Obviously, in a country with a generous working tax credit system, the final result in terms of well-​being for the worker after profiting from the tax credit would be a higher level of well-​being, but that higher level of well-​being would be the result of a good welfare state, not of a high level of job quality. From this perspective it is useful to distinguish between labour market quality and job quality. In fact, we could envisage the index of job quality as one of the indicators (with other such as the unemployment rate, the level and generosity of unemployment benefits or the level of employment) of a hypothetical index of labour market quality. Results or procedures? Another major choice to be made is whether the index should focus on the results in terms of the conditions of work and employment (is the job safe?

Table 6.1 Dimensions of job quality suggested by the different traditions

The orthodox economic approach: compensating differentials Labour compensation (1): wages

Behavioural The radical economic economic approach: exploitation approaches Power relations: (2) Industrial democracy as a compensating power

Source: authors’ elaboration

The traditional sociological approach: alienation and intrinsic quality of work

(3) Participation Objective strand: (4) skills (5) autonomy Subjective strand: (6) powerlessness (7) meaninglessness (8) social isolation (9) self-​estrangement

Occupational medicine and health and safety literature: risks and The institutional approach: segmentation impact of work on Work-​life balance and employment quality health studies (10) Contractual status and stability of employment (11) Opportunities for skills development and career progression

Conditions: (12) Physical risks (13) Psychosocial risks;

Working time: (16) Duration (17) Scheduling (18) Flexibility (19) Regularity (20) Clear boundaries

Outcomes: (14) Perceived impact of work Intensity: on health (21) Pace of work and (15) Absenteeism workload (22) Stress and exhaustion

134    rafael muñoz de bustillo et al Is it secure? Is the working schedule flexible?), or whether it should focus on the existence of procedures aimed at improving conditions or minimizing problems. For instance, in terms of health and safety: should the indicator measure directly the exposure to risks in the job, or whether there are adequate health and safety procedures aimed at minimizing them? In terms of the definition of job quality presented earlier in the introduction, it is preferable to measure the actual conditions rather than procedures (since the latter do not affect directly the well-​being of workers, but only indirectly through their results). But in some cases, it may be justified to include measures of procedures rather than their results in terms of conditions if there is a lack of good information on the former. And even more importantly, it can be argued that procedures themselves can be an important (and positive) attribute of a job, regardless of its impact on the job’s characteristics. For instance, the existence of interest representation and collective bargaining mechanisms can be considered as a procedure that can improve the conditions of work and employment, but also as an element of job quality on their own by providing channels of communication and participation. Measured with individual or aggregate data? There are examples of job quality indicators constructed with data at the individual level (generally, drawing from surveys on the conditions of work and employment) and indicators constructed with aggregate data (generally, drawing from national-​level information from a variety of sources). In our view, there are strong arguments for preferring indicators built from individual data. First, measures constructed from aggregate data cannot be used to study distributional or dispersion issues. Though this might not be a serious problem in other types of cross-​country comparisons, it can be very negative for job quality, which tends to vary more within than between countries (Fernández-​Macías et al. 2015). Although there are ways to account for the distribution in indicators constructed at the aggregate level, such as including dispersion measures within the indicator, this only marginally reduces the problem: once the indicator is constructed, it is fixed and there is no way to explore any distributional issue beyond the particular indicator considered. A second limitation of constructing the indicator at the aggregate level is the impossibility of studying interactions between the different dimensions forming the indicator or system of indicators: it is impossible to analyse the intersection between the different dimensions of job quality or the existence of compensation mechanisms between them. This aim is only reachable using indicators constructed at the individual level. A single index or a vector of indicators? The measurement of job quality and the presentation of results can be done either as a series of indicators (corresponding to the different facets of the phenomenon according to the underlying model) or as a single composite index. The difference between the two approaches should not be exaggerated: they can be essentially identical in all respects except for the very final step of aggregating or not the final vector of indicators into a single index. The main difference lies in the interpretation and implicit evaluation of job quality that is made in both cases. While in the case of the single index it implies a univocal and unidirectional understanding of what job quality is (no matter how many components it may be based upon), which will unambiguously position the different countries (or whatever social group we are interested

quantitative approaches to assess JOB QUALITY    135 in studying) within a one-​dimensional axis going from bad to good, in the case of a vector of indicators it will provide a more nuanced but less unequivocal comparison between different countries or social groups, since each of them can be better or worse in different dimensions. But again, we would not want to overemphasize the difference between the two approaches, which remain essentially the same except for the single index taking the aggregation a step further (a vector of indicators will necessarily aggregate information as well, at the level of each of the dimensions covered by the model). A composite index on its own, with no disaggregation of results by dimensions, would simply be useless for the analysis of job quality; whereas a system of indicators presented with no attempt of final aggregation or overall evaluation can make it very difficult to reach any normative or policy relevant conclusion. Country-​specific or universal? In principle, as in other areas of socio-​economic analysis (for example, the measurement of material deprivation), one could argue that if we aim to deliver an absolute measure of job quality, we should develop an index based on universal values that should be valid across different countries and times. But in practice, there can be very important problems when the comparison spans countries with very different levels of economic development. First, there are some basic elements of work and employment conditions that are so taken for granted in most developed economies that they are rarely even considered as elements of job quality, yet they remain extremely important in developing economies: for instance, basic rights of association in trade unions, or the existence of a formal employment contract. Second, the availability of statistical information on job quality is much more limited in the developing world, and therefore the most sophisticated proposals are in practice only applicable to the most developed nations. For such reasons, the indices proposed for worldwide use (such as some of the operationalizations based on the ILO’s decent work agenda which we discuss later) are considerably simpler than the indices intended for application to developed nations.

Indices of Job Quality: A Shopping List The growing interest in the evolution of job quality and its measurement has led to the development of many proposals, with different levels of ambition and sophistication, to operationalize the concept of job quality. In Tables 6.2 and 6.3 we have summarized 24 such proposals, produced by individuals and organizations aiming at measuring job quality in one or more countries. It is not easy to summarize the information of the 24 indicators covered here, but some relevant conclusions can be drawn from this review:6

6 

This section is an update of Chapter 4 of Muñoz de Bustillo et al. (2011a).

136    rafael muñoz de bustillo et al Table 6.2 List of acronyms, complete names, sources and databases of the indices of job quality reviewed Acronym Complete name

Scope

Source

Databases

DGBI

DGB Good Work Index

Germany

Mußmann (2009)

Ad hoc survey

DWI-​1

Decent Work Index-​1

Developed and developing countries

Ghai (2003)

ILO databases

DWI-​2

Decent Work Index-​2

Developed and developing countries

Bonnet et al. (2003)

ILO databases

DWI-​3

Decent Work Index-​3

Developed and developing countries

Anker et al. (2003)

ILO databases

DWI-​4

Decent Work Index-​4

Developed and developing countries

Bescond et al. (2003) ILO databases

EJQI

The European Job Quality Index

European Union

Leschke et al. (2008)

ELFS, EWCS, SILC. AMECO, ICTWSS

EWCI

European Working Conditions Indicators

European Union

Green and Mostafa (2012)

EWCS

EWCS

European Working Conditions Survey

European Union

Parent-​Thirion et al. (2007)

Itself a data source

GBJI

Good and Bad Jobs Index

Middle-​income countries

Ritter and Anker (2002)

IPSS

GJI

Good Jobs Index

Middle-​income and developing countries

Avirgan et al. (2005)

ILO databases

ICQE

Index of the characteristics related to the quality of employment

Chile

Sehnbruch (2004)

Ad hoc survey

IJD

Index of Job Desirability

The United States

Jencks et al. (1988)

Ad hoc survey

IJQ

Indicators of Job Quality

Canada

Jackson and Kumar (1998)

GSS, SWA

IQL

Indicators of Quality of the Labour Market

Spain

Caprile and Potrony (2006), Toharia et al. (2008)

NSI, MLI

JJI

Just Jobs Index

Developed and developing countries

Kebele et al. (2013)

KILM, WB, UNSDS

EJJI

European Just Jobs Index

EU(28)

Kebele et al. (2014)

KILM, WB, IMF, OECD EUROSAT, AIAS

JQI

Job Quality Indicator

EU(27)

Muñoz de Bustillo et al. (2011a)

EWCS

Quantitative Approaches to Assess JOB QUALITY    137 Table 6.2  Continued Acronym Complete name

Scope

Source

Databases

Laeken

Laeken indicators of job quality

European Union

European Commission (2008)

ECHP, ELFS, SILC

MJQ

Measures of Job Quality

Developed

OECD (2014)

OECD, EWCS

QEI

Quality of Employment Indicators

Canada, U.S. and Brisbois (2003) Europe

EWCS, ERNAIS

QWF

Quality of Work in Flanders

Flanders (Belgium)

Flanders Social and Economic Council (2009)

Ad hoc survey

SQWLI

Subjective Quality of Working Life Index

Czech Republic

Vinopal (2009)

Ad hoc survey

Tangian

Tangian’s proposal

European Union

Tangian (2007)

EWCS

WCI

Austrian Work Climate Index Austria

Preinfalk, Michenthaler and Wasserbacher (2006), Michenthaler (2006)

IFES omnibus survey (dedicated module)

Notes: ECHP =​European Community Household Panel; ELFS =​European Labour Force Survey; SILC =​Statistics on Income and Living Conditions; EWCS =​European Working Conditions Survey; AMECO =​Annual Macroeconomic Database of the European Commission; ERNAIS =​Ekos Rethinking North American Integration Survey; GSS =​General Social Survey; SWA =​Survey of Work Arrangements; IFES =​Institut für empirische Sozialforschung; NSI =​National Statistics Institute; MLI =​Ministry of Labor and Immigration; IPSS =​ILO People’s Security Surveys., IMF =​International Monetary Fund database, WB =​World Bank Database, AIAS =​Amsterdan Institute for Advance Studies of labour, KILM =​Key Indicators of Labour Market, ILO, OCDE =​OECD database, UNSDSD =​United Nations Statistrics Database Source: Authors’ analysis

1. A sizeable part of the indicators reviewed are not really (or not strictly) measures of job quality, since they include issues as unrelated to job quality as labour market access, the distribution of disposable income, illiteracy rates or standard macroeconomic indicators. This problem is especially evident in the case of the International Labour Organization’s Decent Work indices and the EU Laeken indicators, but also in the JJI7 or in the OECD´s MJQ that include unemployment risk, unemployment duration and effective unemployment insurance. The EJQI, the EWCS, the SQWLI, the QEI, the DGBI, the WCI and some individual academic proposals such as the JQI avoid this shortcoming. In addition, most of the indicators include some measures of social security comprehensiveness, which (on top of being unrelated to job quality in a strict sense) is problematic for

7 The

JJI includes, for example, labour force participation rate, unemployment rate, youth unemployment rate or GDP per capita in PPP.

Table 6.3 Summary of the main indicators of job quality No. of . . .

Type of indicator Index

Type of variables

Individual Multi-​ Worker-​ Periodicity purpose oriented System Aggregate Dimensions Variables Subjective Objective Results Procedures Static Dynamic data

DGBI

X

X

X

3

31

X

X

X

X

YES

Annual

DWI-​1

X

X

X

4

9

X

X

X

X

NO

Single exercise

DWI-​2

X

X

X

7

67

X

X

X

X

NO

Single exercise

DWI-​3

X

X

11

37

X

X

X

X

NO

Single exercise

DWI-​4

X

X

X

8

X

X

X

X

NO

Single exercise

X

EJQI

X

X

EWCS

X

X

EWCI

X

X

GBJI

X

X

X

GJI

X

ICQE IJD

6

17

X

X

X

X

X

NO

Annual

4

Many

X

X

X

X

X

YES

Every 5 years

4

17

X

X

YES

Every 5 years (?)

1

6

YES

Single exercise

X

X

X

X

X

X

5

16

X

X

X

X

NO

Single exercise

X

X

X

5

15

X

X

X

X

YES

Single exercise

X

X

X

11

48

X

X

X

YES

Single exercise

7

27

X

X

X

X

X

X

NO

Single exercise

X

X

X

X

IJQ

X

X

IQL

X

X

X

8

38

X

X

NO

Annual

JJI

X

X

5

14

X

X

X

NO

Single exercise, with planned updates

EJJI

X

X

5

33

X

X

X

NO

Single exercise, with planned updates

X

5

15

X

X

X

YES

Every 5 years

JQI

X

Laeken

X

10

25

X

X

MJQ

X X

X

3

11

X

X

QEI

X

X

5

11

X

X

X

QWF

X

X

X

2

10

X

SQWLI

X

X

X

6

18

X

Tangian

X

X

10

109

X

WCI

X

X

4 (16)

25

X

Source: Authors’ analysis

X

X

NO

Annual

X

X

NO

Single exercise

X

X

NO

Single exercise, with occasional updates

X

X

X

YES

Every 3 years

X

X

X

YES

Single exercise

X

X

X

X

YES

Single exercise

X

X

X

YES

Bi-​annually

140    rafael muñoz de bustillo et al













cross-​country comparisons due to the existence of very different welfare regimes, with substantially different roles for the private and public sector. 2. Some dimensions highlighted as important by the Social Sciences literature (see Table 8.2) are absent from most indicators. This particularly applies to work intensity, an omission largely conditioned by the practical absence of surveys covering this issue. Furthermore, there are some important indicators—​ the Laeken indicators, the DWI-​1, the QEI and the QWF—​that make no reference to earnings, which is definitely an important flaw. 3. One third of the indicators include no aggregate measures of job quality and only offer a system of indicators. By proceeding this way, they avoid setting—​ and justifying—​weights for the different dimensions. However, as previously mentioned, this makes the overall evaluation of job quality much more difficult as the different dimensions can yield contradictory results. 4. There is a high diversity in terms of the number of dimensions and variables used across the indicators, from six variables (the GBJI) to more than a hundred (the EWCS and Tangian’s proposal). In most of the cases, the aggregation is carried out on the basis of equal weights (although the individual variables are normally grouped within dimensions), often without any theoretically sound explanation. 5. Most indicators include both procedure and outcome variables, with only one third relying only on outcomes. Another relevant aspect has to do with the static or dynamic nature of the variables considered. Though there are many indicators including variables that refer to opportunities of advancement in the current job, only a few indices—​the Laeken indicators, the IJQ, the IQL and the GBJI—​use longitudinal measures of job or income mobility. 6. Many of the measures reviewed include as components of job quality distributional issues, such as gender or age gaps for particular variables. A more fitting approach is to design indicators that can be computed at individual level and then compare the scores (or the gaps) of job quality for the different groups of interest. Nine of the indices reviewed in the previous subsection—​the ECWS, the SQWLI, the DGBI the WCI, the QWF, Tangian’s proposal, the GBJI, the ICQE and the JQI—​have this desirable characteristic. 7. Most indicators rely on objective variables. In fact, only two of the indices reviewed—​the QWF and the SQWLI—​are based solely on such sorts of subjective variables. 8. In terms of geographical coverage, most of the indicators were designed to be applied to the EU. In fact, the underrepresentation of indices specifically developed for (or in) the United States was an unexpected result. Considering the high sophistication of social and economic statistics in the US, it is very surprising that there is not, to our knowledge, a general and periodical survey of job quality in that country (such as the European Working Conditions Survey and many other similar national surveys). In this respect, Table 6.3 only includes one genuinely American indicator, the one proposed by Jencks et al. (1988).

quantitative approaches to assess JOB QUALITY    141 Apart from this one, we can mention Tilly’s work (1996), which proposes a list of dimensions to be monitored in order to assess the evolution of job quality in the United States. However, although this proposal is useful and valid, it does not attempt any type of aggregation into an index and it only relies on secondary data, so we have not included it in our review8. On the other hand, we have to admit that the only attempt made so far by EU institutions themselves to measure job quality (the Laeken indicators of job quality) has not been particularly successful. In many ways, the Laeken indicators exemplify many of the problems highlighted in this chapter: they do not cover attributes as crucial for job quality as wages or work intensity, while including other variables obviously unrelated to job quality such as long-​term employment rate, productivity or labour market transitions. They include job satisfaction as one of the key components of job quality, which as argued earlier involves mixing inputs and outputs. They also include distributional variables (such as gender or age gaps) as components of job quality but allow no distributional analysis because they are constructed only at the country level.

Two Indicators of Job Quality as Example Although Table 6.2 demonstrates a very important and growing interest in measuring job quality, the conceptual and methodological problems previously discussed are still far from being solved. In fact, the mere existence of such a large and growing number of alternative indicators can be taken as a proof of a lack of agreement on those conceptual and methodological issues. If we compare such diversity of proposals with the more or less widespread consensus in the measurement of other crucial economic phenomena such as unemployment, GDP or PPS, we can see that the measurement of job quality is still relatively immature. In this section, to exemplify the type of results obtained when using indicators of job quality, we discuss in more detail two of the previously reviewed proposals: the JQI, an indicator of job quality constructed from individual data developed by Muñoz de Bustillo et al. (2011a), and the EJQI, produced by the ETUI from aggregate country data (Leschke and Watt 2014), in both cases applied to the EU(15).

8 

Since 2019, Alper et al, at the University of Cornell, produce on a monthly basis the so-​called U.S. Private Sector Job Quality Index, aiming at assessing job quality in the United States by measuring the evolution of desirable higher-​wage/​higher-​hour jobs vis a vis lower-​wage/​lower-​hour jobs. This indicator, although offering interesting information about wages and hours, fall short, from the perspective of this chapter, of the requirements of a job quality indicator

142    rafael muñoz de bustillo et al Table 6.4 European Job Quality Index and Job Quality Indicator European Job Quality Index (EJQI)

Job Quality Indicator (JQI)

Dimension

Indicator

Dimension

Wages

Nominal compensation in PPS; Wages gender figure adjusted by gender gap

Non-​standard employment

Temporary employment

Skills and career development

Participation in training

Indicator Nominal compensation in PPS

In work poverty

Working time and work-​life balance

Involuntary part time employment

Employment quality

Prospects for career advancement

Contractual stability Development opportunities

Work life balance Duration Scheduling Intensity

Shift-​work, Saturday, Sunday & evening work Voluntary part time work Working hours fit with family/​social commitments

Working conditions and job security

Collective interest representation

Work intensity Work autonomy Physical work factors Job insecurity

Health and safety

Physical risks

Intrinsic quality of work

Skills Autonomy Social Support

Collective bargaining coverage Trade union density Consulted about changes in work organization

Source: Leschke and Watt, (2014), Muñoz de Bustillo et al. (2011a) and Fernández-​Macías et al. (2015)

Both proposals present an aggregate indicator based on different dimensions (six in the case of the EJQI and five in the case of the JQI). Such dimensions are themselves constructed using aggregating lower-​level indicators, following a nested model. As we can see in Table 6.4, the two proposals are quite consistent in terms of the broad dimensions of job quality (although not identically structured because some dimensions in one index are split in two in the other). The only significant discrepancy is the introduction of a dimension of industrial democracy (collective interest participation) in the EJQI. Such an item is lacking in the IJQ, that focuses in indicators of outcomes as opposed to indicators of procedure.9 The major difference between the two indices is

9 In

the theoretical discussion, Muñoz de Bustillo et al. (2011a) do consider participation and industrial democracy as valid components of job quality even from an outcomes perspective (as

quantitative approaches to assess JOB QUALITY    143 1.00 Denmark 0.80

Netherlands Belgium Luxembourg Sweden Finland Austria UK EU(15) Ireland France Germany Italy Portugal Spain

EJQI

0.60

0.40

Greece 0.20

0.00 0.00

0.20

0.40

0.60

0.80

1.00

JQI

Figure 6.3  JQI and EJQI in the EU (15), 2010 Source: For the EJQI data provided by J. Leschke; for the JQI authors analysis of EWCS 2010

the type of data used in their construction: aggregate in ETUI’s EJQI and individual in the JQI. As mentioned before, the construction of an index of job quality from individual data allows us to study to what extent there is accumulation of good and bad characteristic on a given job or worker. On the downside, in Europe this approach can only be effectively applied to a single data source, the EWCS, which constrains the information that can be covered, the periodicity, etc. In contrast, the use of aggregate data allows a broader coverage of dimensions and the introduction of variables absent from the EWCS. In both cases the authors stress that the analysis should never be limited to the level of the aggregate indicator, but always presented with a decomposition of the index into its broad dimensions, and even to lower levels of sub-​components or sub-​indicators, since they can move in different directions. Figure 6.3 reproduces the results of the application of the two aggregate indices to the EU (15) in 2010. The correlation is very high (0.9), and the ranking is reasonably consistent. The most significant difference is the much lower dispersion of the JQI, where the country with best markings, Luxembourg, has a JQI ‘only’ 40 per cent higher than the country with the lowest score, Greece. In contrast, according to the EJQI the country top of the league, Denmark, has a value more than twice as high as the country with the lowest score (also Greece). But this difference is simply the result of the different normalization approach taken by the two indices: while the JQI uses a normative approach (the original values of the variables were re-​scaled to 0–​1 according to their potential

channels of expression of interest and control of workers’ own conditions). However, they do not include such indicators in the index because of a lack of suitable individual-​level variables in the 4th EWCS.

144    rafael muñoz de bustillo et al impact on the well-​being of workers, as judged by the index proponents on the basis of the existing literature; see Muñoz de Bustillo et al. 2011a: 251), the EJQI used a min–​ max normalization to the country-​level results for the six dimensions of the index (rescaling the original values so that they ranged from 0 to 1 and their relative distance was expressed in proportion with that range). The approach of the EJQI maximizes the range, with the values having only a relative interpretation (as distances between countries in their scores), while the JQI tends to produce a smaller range but the values have a more substantive interpretation (as they represent the average normative scores).

Summary and Conclusions Job quality is an important dimension of work and, as such, should be an element considered in its own right when analysing the characteristics of labour markets and the implications of labour policies. In this chapter we have discussed the alternatives and challenges faced in the measurement of job quality, reviewing a long list of existing proposals by individual researchers and organizations. The starting point of any exercise of job quality measurement is the recognition of the multidimensional nature of job quality. As argued in section 2, there are three broad strategies to select the dimensions of work to be considered in the measurement of job quality. The first one is to jump to the final outcome and use job satisfaction as a proxy of job quality. The second approach asks the subjects themselves, the workers, to select and weigh the different items to be considered. The third option draws from the growing body of social sciences literature on the effects of work attributes on workers´ well-​being to guide and model the construction of the index. From our perspective, this last option, a theory-​based index, is to be preferred to the former two, as it is inherently more transparent: since it forces to make explicit and justify the criteria and aggregating procedure that produces the index (rather than leaving it to the subjective evaluation of workers themselves), it always makes it more understandable and useful for evaluation and policy purposes. The review of a sample of existing job quality indicators undertaken in section 4 shows both the growing interest in the last decades in having such a tool and its feasibility. Although the issue of how to measure job quality is far from being resolved, it can be argued that research is moving in the right direction, with several of the proposals considered in Table 6.2 meeting most of the desirable properties for the construction of this kind on indices. That being said, it has to be acknowledged that the issue is far from resolved. From our perspective the next steps in the direction of improving the measurement of job quality should concentrate in the following areas: (1) There is a need to clearly differentiate the question of the levels of unemployment and other labour market related issues from the question of job quality. As we have argued, job quality indicators should focus on how the attributes of jobs affect the well-​being of the worker, not being the probability of finding

quantitative approaches to assess JOB QUALITY    145 employment one of them. The aggregation of indicators of quantity and quality of employment in a single index empties such index of analytical value, regardless of the potential trade-​offs or complementarities existing between both dimensions. A different question is the role that the consideration of indicators of employment and job quality might have in the development of measures of quality of the labour market. (2) We know that there are many dimensions affecting job quality, but we do not have a definitive approach as to how to weight their importance in the overall index. The two proposals used as an example in the previous section use a theoretical/​normative approach to justify the weights (each individual item is implicitly weighted in the model according to the importance given to it by the index proponents), although in both cases they opt for equal weights at the higher-​level dimensions. The implications of alternative approaches (including using workers’ preferences revealed through surveys) should be more explicitly evaluated. (3) With few exceptions, such as the EWCS, the data used in the measurement of job quality has been developed for other purposes, and seldom cover all the dimensions needed in the analysis, or with the desired periodicity. In fact, many of the decisions taken in the development of most job quality indices—​what to include and how to include it—​are driven by data restrictions. We think we can safely say that right now the major problem faced by the development of better job quality indicators is the lack of suitable and periodic data more than the shortcomings of the theoretical and methodological underpinning of the existing proposals.

References Alper, D., Ferry, D., Hockett, R. C. and Khaleghi, A. (2019), The U.S. Private Sector Job Quality Index (Ithaca: Cornell Law School). Anker, R., Chernyshev, I., Egger, P., Mehran, F., and Ritter, J. (2003), ‘Measuring Decent Work with Statistical Indicators’, International Labour Review, 142, 147–​178. Avirgan, T. Bivens, L. J., and Gammage, S. (2005), Good Jobs, Bad Jobs, No Jobs: Labor Markets and Informal Work in Egypt, El Salvador, India, Russia, and South Africa (Washington DC, Economic Policy Institute). Bescond, D., Châtaigner, A.< and Mehran, F. (2003), ‘Seven Indicators to Measure Decent Work: An International Comparison’, International Labour Review, 142, 179–​212. Bonnet, F., Figueiredo, J. B., and Standing, G. (2003), ‘A Family of Decent Work Indexes’, International Labour Review, 142, 213–​238. Bothfeld, S., and Leschke, J. (2012), ‘‘More and Better Jobs’: Is Quality of Work Still an Issue –​ and Was it Ever?’ Transfer: European Review of Labour and Research, 18/​3, 337–​353. Brisbois, R. (2003), ‘How Canada Stacks Up: The Quality of Work -​An International Perspective’, Research Paper No. 23. Canadian Policy Research Networks, Canada. Brown D., and M. McIntosh (1998), ‘If You’re Happy and You Know It: Job Satisfaction in the Low Wage Service Sector’, Centre for Economic Performance paper. London School of Economics, London. Caprile, M., and Potrony, J. (2006), ‘IQT: Objetivos y metodología’, in UGT, ed., Anuario Sociolaboral de la UGT de Catalunya 2005. Vol. II (Barcelona: UGT and CRESC), A1–​A14.

146    rafael muñoz de bustillo et al Clark A. E., and Oswald, A. J. (1996), ‘Satisfaction and Comparison Income’, Journal of Public Economics, 61, 359–​381. Dolan P., Pasgood T., and Whitte M. (2008), ‘Do We Really Know What Makes Us Happy? A Review of the Economic Literature on the Factors Associated with Subjective Well-​Being’, Journal of Economic Psychology, 29/​1, 94–​122. Engels, F. (1891), The Condition of the Working Class in England (London: Swan Sonnenschein & Co). European Commission (2008), Employment in Europe 2008 (Brussels: European Commission). European Commission (2012), New Skills and Jobs in Europe: Pathways towards Full Employment (Brussels: European Commission). Fernández-​Macías, E., Muñoz de Bustillo, R., and Antón, J. I. (2015), ‘Job Quality in Europe in the First Decade of the 21st Century’, Department of Economics Working Paper No. 1509, Johannes Kepler University Linz. Festinger L. (1957), A Theory of Cognoscitive Dissonance (Stanford University Press. Palo Alto). Flanders Social and Economic Council (2009), ‘Quality of Work in Flanders 2004-​2007’. Paper presented at the expert seminar organized by the Observatoire Social Européen, Brussels, Belgium. Freeman R. B. (1978), ‘Job Satisfaction as an Economic Variable’, American Economic Review, 68/​2, 135–​141. García Mainar, M. I. (1999), ‘La satisfacción con el trabajo en España’, Documentación laboral, 59, 113–​129. Ghai, D. (2003), ‘Decent Work: Concept and Indicators’, International Labour Review, 142, 113–​145. Green, F. (2006), Demanding Work: The Paradox of Job Quality in the Affluent Economy (Princeton, NJ: Princeton University Press). Green, F., and, Mostafa, T. (2012), Trends in Job Quality in Europe. A Report Based on the 5th European Working Conditions Survey (Dublin: Eurofound). ISSP Research Group (2013), International Social Survey Programme: Work Orientations III -​ISSP 2005. GESIS Data Archive, Cologne. ZA4350 Data file Version 2.0.0, https://​doi.org/​ 10.4232/​1.11648. Jackson, A., and Kumar, P. (1998), ‘Measuring and Monitoring the Quality of Jobs and the Work Environment in Canada’. Paper presented at the CSLS Conference on the State of Living Standards and the Quality of Life in Canada. Centre for the Study of Living Standards, Ottawa, Canada. Jencks, C., Perman, L.. and Rainwater, L. (1988), ‘What is a Good Job? A New Measure of Labor-​ Market Success’, American Journal of Sociology, 93, 1322–​1357. Kebele, T. A., Jesnes, K., Zhang H., and Pedersen J. (2014), European Just Job Index. Progressive Economy Briefing Paper Series. Kebele, T. A., Zhang H., and Pedersen J. (2013), Just Job Index 2013. Centre for American Progress. Kelvin, W. T. (1889), ‘Electrical Units of Measurement’. Lecture to the Institution of Civil Engineers, London, Popular Lectures and Addresses 1824-​ 1907, 1: 80-​ 81. Macmillan, London. Knox, A., and Wright, S. (2022), ‘Understanding Job Quality Using Qualitative Research’, in C. Warhurst, C. Mathieu, and R. Dwyer, eds, Oxford Handbook of Job Quality (Oxford: Oxford University Press).

quantitative approaches to assess JOB QUALITY    147 Lambert, E. G., Hogan, N. L. and Barton, S. M. (2001), ‘The Impact of Job Satisfaction on Turnover Intent: A Test of a Structural Measurement Model Using a National Sample of Workers’, Social Science Journal, 38/​2, 233–​250. Leschke, J., Watt, A., and Finn, M. (2008), Putting a Number on Job Quality? Constructing a European Job Quality Index, ETUI-​REHS Working Paper No. 2008/​03. European Trade Union Institute, Brussels. Leschke, J., and Watt, A. (2014), ‘Challenges in Constructing a Multi-​Dimensional European Job Quality Index’, Social Indicators Research, 118/​1, 1–​31. Manglione T. W. and Quinn, R. P. (1975), ‘Job Satisfaction, Counter-​Productive Behavior and Drug Use at Work’, Journal of Applied Psychology, 60, 114–​116. Michenthaler, G. (2006), ‘The Austrian Work Climate Index’. Paper presented at the European Working Conditions Seminar on Job Satisfaction. Helsinki. Muñoz de Bustillo R., and Fernández-​Macías E. (2005), ‘Job Satisfaction as an Indicator of the Quality of Work’, Journal of Socio-​Economics, 34/​5, 656–​673. Muñoz de Bustillo, R., Fernández-​Macías, E., Antón, J. I., and Esteve, F. (2011a), Measuring More than Money: The Social Economics of Job Quality (Cheltenham: Edward Elgar). Muñoz de Bustillo, R., Fernández-​Macías, E., Antón, J. I., and Esteve, F. (2011b), ‘E pluribus unum? A Critical Survey of Job Quality Indicators’, Socio-​Economic Review, 9/​3, 447–​475. Mußmann, F. (2009), ‘The German ‘Good-​Work’ Index (DGB-​Index Gute Arbeit)’. Paper presented at the expert seminar organized by the Observatoire Social Européen. Brussels. OECD (2014) Employment Outlook 2014 (Paris: OECD). Origo, F., and Pagani, L. (2008), ‘Workplace Flexibility and Job Satisfaction: Some Evidence from Europe’, International Journal of Manpower, 29/​6, 539–​566. Parent-​Thirion, A., Fernández-​Macías, E., Hurley, J., and Vermeylen, G. (2007), Fourth European Working Conditions Survey (Dublin, Eurofound). Preinfalk, H., Michenthaler, G., and Wasserbacher, H. (2006), ‘The Austrian Work Climate Index’. Paper presented at a Research Seminar of the European Foundation for the Improvement of Living and Working Conditions. Dublin. Ritter, J. A., and Anker, R. (2002), ’Good Jobs, Bad Jobs: Workers’ Evaluations in Five Countries’, International Labour Review, 141, 231–​258. Sehnbruch, K. (2004), ‘From the Quantity to the Quality of Employment: An Application of the Capability Approach to the Chilean Labour Market’, Center for Latin American Studies Working Paper No. 9. University of California, Berkeley. Spector P. E. (1997), Job Satisfaction. Application, Assessment, Causes and Consequences (London: Sage). Tangian, A. (2007) ‘Analysis of the Third European Survey on Working Conditions with Composite Indicators’, European Journal of Operational Research 181/​1, 468–​499. Vinopal, J. (2009) ‘The Instrument for Empirical Surveying of Subjectively Perceived Quality of Life’. Paper presented at the conference Working Conditions and Health and Safety Surveys in Europe: Stocktaking, Challenges and Perspectives. European Trade-​Union Institute, Brussels. Warhurst, C., Carré, F., Findlay, P., and Tilly C. (eds) (2012), Are Bad Jobs Inevitable? Trends, Determinants and Responses to Job Quality in the Twenty-​First Century. (London: Palgrave Macmillan).

chapter 7

job quality as the realization of democratic ideals john godard

Concern for the societal level relationships between institutions and job quality dates at least to the late eighteenth and the nineteenth centuries, when capitalism had begun to emerge as the predominant means for economic organization. For example, the putative ‘father’ of liberal economic theory, Adam Smith, evinced concern for the apparent power imbalance between workers and their masters (Winch 2004), and for the possibility that the emerging division of labour could render workers “stupid and ignorant” (19 [1776]: 267). Karl Marx, and especially the young Marx, worried not just about exploitation, but also about the quality of work under capitalism and what it meant for the intellectual capacities of workers (Marx 1965 [1844], Marx and Engels 1960 [1846]). Max Weber raised the spectre of an ‘iron cage of bureaucracy’, in which work is rendered devoid of meaning or fulfilment, with broader implications for society (Weber 1968 [1922]). Similar concerns came to be especially widespread in the middle of the twentieth century and were reflected in the work of numerous industrial sociologists and others (e.g. Mayo 1933; Seeman 1959; Blauner 1964; O’Toole et al. 1974; Braverman 1974; Erikson 1986). By the early twenty-​first century, however, the focus had come increasingly to be on the problems of workers as individuals rather than more systemic problems they encounter as members of societies. These accounts typically addressed preconceptions of job quality held by academics and policy analysts, often relying on, and conceptually constrained by, available data sets. In both respects, they tended to be theoretically weak, resulting in myriad often disparate results rather than any sort of coherent, theoretically informed conception of job quality, how it really matters, and why it varies. Efforts to address this problem have, to date, largely failed (e.g. Findlay et al. 2013). The purpose of this chapter is to establish the basis for a more systematic inquiry into job quality, one that is focused on societal level ideals or principles basic to western democracies, and addresses how institutional differences matter to the attainment of

job quality as the realization of democratic ideals    149 these ideals at work. In so doing, I draw extensively from, but expand on, Frege and Godard (2014). Although Frege-​Godard focused on the differences between the US and Germany and both whether and how they matter, a central purpose was to advance a new approach to the study of job quality, one that could help to advance the study of job quality beyond the difficulties that currently characterize the literature (e.g. see Kalleberg 2016). The present chapter seeks to further develop both the rationale for, and the potential value of, this approach. It begins by discussing the ‘quagmire’ in the job quality literature. It then develops a societal-​level approach to job quality, one that focuses on ideals on which Western democracies are founded. It then addresses the extant literature on societal-​level variation in job quality and elaborates on the institutional environments approach developed in Frege-​Godard. It concludes with the case for the development of an institutionally informed normative paradigm for the future of work.

The Job Quality Quagmire Analyses of job quality have typically fallen within one of (and, in some cases, a combination of) two categories: those that focus on the terms and conditions of employment, and those that focus on the content and experience of work. The first of these typically address pay levels and whether a job conforms to the ‘standard employment relation’ (SER) associated with the “golden age” of the 1950s and 1960s. Jobs that are low pay and/​or do not conform to the SER are considered a priori to be ‘bad’ (e.g. Osterman and Shulman 2011; Stone 2013); higher-​paying jobs that do conform to the SER are considered to be ‘good’. The second conception typically focuses on assorted job characteristics, such as task discretion (Gallie et al. 2004; Gallie 2007a), complexity and autonomy (e.g. Kalleberg 2011), and intensity (e.g. Green 2001). The higher the first three and lower the fourth, the ‘better’ a job is said to be. Both conceptions are usually based on preconceived notions of what it is that jobs should provide. The assumption seems to be that the characteristics identified either fit widespread preferences within the labour force or that they are associated with various indicators of well-​being. Yet there seems to be little evidence of either. For example, temporary employment is typically associated with a bad job, as it is at odds with the SER. Yet the associations between temporary employment and indicators of well-​being tend to be complex, varying with the type of contract involved, and with the overall or ‘average’ effect appearing to be small and often restricted to males in specific types of temporary employment (see Buddelmeyer et al. 2015: 257–​260; Wilkin 2013; Inanc 2018). Even here, the research lacks controls for job characteristics, thereby potentially conflating the effects of temporary employment status (on e.g. stress) with the effects of these characteristics (e.g. Buddelmeyer et al. 2015; Inanc 2018). The results are even weaker for part time employment, also considered to be at odds with the SER. Research has revealed little a priori association

150   john godard between hours worked and various indicators of well-​being (Warr 2007: 317–​326). In most OECD countries for which there is relevant data, fewer than 3 in 10 part-​time workers typically state a preference for full-​time (OECD Stat 2017). Although those who would prefer either permanent or full-​time employment but do not have it are likely to be less satisfied than those who do (Warr 2007: 324; Cooke et al. 2013), this may be attributable to frustrated expectations. It is quite possible that the opposite also holds—​those who would like reduced time employment but are unable to obtain it are likely also to be less satisfied. Some research does find that nonstandard work may be negatively associated with other factors often considered to comprise job quality, especially compensation levels and job content. But if so, it is at best a proxy, and to use it as such, one must establish that compensation and job content are valid indicators. This does not, however, appear to be the case. The research finds only a weak association between pay and job satisfaction (Judge et al. 2010). The implications of job content are also uncertain. In particular, autonomy and complexity are often considered to be important to job quality. Yet the association between these variables and job satisfaction is largely contingent on individual preferences and, in particular, ‘growth need strength’ (Warr 2007: 368–​376). The findings of this research are also often relatively weak and potentially inflated by common method bias. It is possible that the lack of strong findings simply reflects limitations to the research. Scholars have typically relied on secondary data sources and hence have been constrained by the available data (Lechke and Watt 2008). There has also been extensive reliance on subjective outcomes such as satisfaction and happiness as indicators of well-​ being. Not only are these unduly hedonistic and hence of questionable merit, they are also both relative (Clark and Oswald 1996) and conditioned by respondent expectations and adjustment processes. Perhaps for this reason, their range of variation also tends to be small, further limiting research findings. Finally, it is possible that the problem lies with quantitative research itself and that more interpretative analysis, in which interviewers probe more deeply into how individuals feel about, or are affected by, the terms and nature of their employment, would turn up stronger findings. It might, therefore, be argued that there is just need of ‘better’ research. Yet this is all too standard a refrain in contemporary social science research. Rather, the underlying problem may simply be that one size does not fit all and that in fact substantial variation in individual preferences (Warr 2007; Kalleberg 2008; Kalleberg 2016; Cooke et al. 2013) undermines any effort to establish universal criteria for measuring job quality. Take, for example, the young woman whose primary interest is to write novels or play in a rock band. Or the middle-​aged male with two children and a high-​earning partner in a ‘long hours’ job. Or the retired university professor who is seeking employment that provides him/​her with something to do and helps to supplement a meagre pension, but without the stresses and frustrations of academic life. In all three cases, a relatively low-​paying and routine ‘part-​time’ job may be highly desirable, provided that the individual is treated with respect, does not need to work too hard or worry about

job quality as the realization of democratic ideals    151 their performance, and is able to get some level of enjoyment through interactions with others. A temporary contract, with a two-​or three-​month period each year without work (e.g. for a parent, in the summer months, for a retiree, in the winter months) may also be highly desirable, especially if there is a reasonable certainty of re-​employment if desired. It is also possible to argue that there are labour market participants who just do not have the skill levels or capacities to qualify for a ‘good’ job, at least given the assumed definitions in the literature. Yet even these participants are not necessarily consigned to a life in Marx’s ‘Satanic mills’. Indeed, if they have a partner earning decent pay, or if they prefer an alternative, non-​materialist lifestyle, or if there are sufficiently strong labour market protections and social assistance programmes to ensure that their basic needs (social as well as economic) are met, they can live a happy and fulfilling life even if in a “bad” job. The extent to which this is the case is conditional not just on individual circumstances and preferences, but also on available social and labour market programmes and broader social and cultural norms. In this regard, even compensation is of limited value as a means of assessing individual job quality—​especially in developed economies, where income distribution rather than compensation per se has increasingly become the problem. The problem is not that various factors associated with ‘good’ jobs are in fact not ‘good’ from the perspective of a large number and perhaps even a majority of labour market participants. Rather, it is that the extent to which they are can vary considerably given individual preferences and (possibly) adjustment processes, and they may even be bad for some or many participants. Moreover, while subjective outcomes could reflect some innate bundle of human needs, they may reflect numerous additional factors, all of which can vary not only by individual but also by occupation and ultimately by nation. In short, the universalistic conception of job quality that has been implicit in much of the literature is fraught with limitations, often reflecting the prejudices and experiences of academics and (to an extent) policy analysts, but not necessarily those of their research subjects. It is likely that this disconnect will only increase in future, at least if various prognostications about the future of work and changing orientations to work hold true. If so, the above examples are likely to increasingly define how job quality is viewed. There is no perfect solution to this problem. Yet job quality is or should be an important objective of any modern democratic society and, in particular, an important component in any assessment of the performance of that society’s economy. This is relatively obvious and can almost be taken for granted. Yet it also contains the seeds of a solution. Specifically, the matter of job quality should be thought of in societal rather than individual terms. It should be defined in accordance not with individual needs or values per se, but rather with societal ideals, and it should be considered important not to the condition and well-​being of individuals per se, but rather to the condition and well-​ being of societies. In short, job quality should be studied from a societal perspective.

152   john godard

Beyond the Quagmire: A Societal Perspective Attempting to associate job quality with societal ideals might at first blush seem to be subject to the same sort of difficulties as more individualist perspectives. The term ‘ideals’ is by itself potentially contentious, as one person’s ideals may be contrary to those of another. But the intention is not to use this as a moral term. Rather it is simply to refer to what, in general, societies value or consider to be important. One approach to identifying these ideals is to do so empirically, employing a ‘public values’ approach (Bozeman 2007). Yet such an attempt ultimately assumes a consensus that may not exist either within or across nations, and as such would likely result in an empiricist muddle. Instead, the approach advocated here is to resort to more philosophic values or ideals that can be widely agreed on across Western nations and on which the development and health of these nations is or can be judged. By ‘widely agreed on’, I am not referring to some form of empirical consensus within nations, based on studies of public values, but rather on principles or ideals that can be considered as fundamental to Western civilization and against which the progress of nations as societies may be judged (e.g. see Lopes 2015).1 One source to which we might turn is the political philosophy literature (see, e.g. Kymlicka 2002). But although this literature addresses a variety of ideals (e.g. liberty, community, justice, equality, ‘recognition’), it provides little by way of a coherent set thereof. A central concern has instead been the maximization and the distribution of income, wealth, and opportunity within societies (e.g. Nozick, Rawls, Dworkin). With regard to job quality, it may be possible to argue that, from a societal perspective, it is economic outcomes that matter, and that high pay represents high job quality in general, even if it may be of limited value at the individual level (for reasons discussed earlier). A society in which everyone has access to a high-​pay job is one in which everyone is assured of financial independence, freedom and dignity. Yet to make such an argument assumes a very materialistic world, in which income is the primary motivating force and hence the primary concern of societies. Income and its distribution clearly matter, and the latter in particular may be considered as one indicator of the quality of a society. But both are at best means to ends, not ends in themselves. It is necessary to identify more fundamental ideals or principles of western democracies. These may be located in political philosophy, but they may also be

1  For example, Inglehart and Welzel (2005) draw on data from the World Values Surveys to address the progress of different nations towards the attainment of various values associated with modernization and democracy. The problem is that such analyses primarily address beliefs, attitudes, and dispositions about specific social and political issues (e.g. abortion, authority). These tend to be somewhat arbitrary and of limited value for identifying dimensions of the work experience that can be associated with job quality.

job quality as the realization of democratic ideals    153 founding principles of nations, or be basic to legal jurisprudence within nations, or figure prominently in national or international declarations of rights, or some combination of the above. However, they need to be philosophically grounded, if by philosophical we mean they are considered to be of fundamental importance to the human condition and derived through a combination of reason and widely agreed (if not universal) understandings as to the ideal nature of this condition. These principles may be deemed to be socially good in themselves, regardless of whether they have positive (or negative) associations with other outcomes (e.g. income). To the extent that they embody basic virtues considered to advance higher ideals of Western civilization, they may be considered as indicative of the democratic quality of work in a society. This is the strategy adopted in the Frege-​Godard paper referred to earlier. It identifies nine such principles: 1) liberty/​freedom, 2) fulfilment/​self-​development, 3) dignity, 4) fairness, 5) justice, 6) security, 7) fraternity/​community, 8) democracy (democratic empowerment), and 9) equality. While the first eight are ‘first order’ principles, and are measured through a battery of questions addressing individual work experiences, the ninth is a ‘second order’ principle, based on the standard deviations of an aggregate measure of the first order principles. The items used to capture these principles appear in Table 7.1. A number have been included in other research (albeit mainly in Godard 2001 and 2010), and so the question is not as much one of reinventing the wheel as it is of changing the lens through which this wheel is viewed, by interpreting the items in terms of principles rather than individual outcomes or experiences per se. This approach is not without its potential limitations. For example, it risks a possible exclusion of principles that could be considered to be of equivalent or greater importance than those selected. It also fails to address the relative importance of various principles and the likelihood that this importance can vary across nations. Finally, the way in which these principles should be defined and operationalized is not always obvious or necessarily easily agreed to, especially given that some or many (e.g. liberty/​ freedom) can be contentious in either law or philosophy. Certainly, there is ample basis for contesting the item content of various principles in the Frege-​Godard paper (although these were meant to pick up basic constructs, not to identify all criteria relevant to them). The way to address such problems is to avoid proposing a ‘totalizing narrative’ in which we claim to identify a singular, comprehensive construct or set of constructs. Instead, we can simply state that we are identifying principles that are considered to be fundamental (even if of varying importance) to a democratic society. These principles may be studied individually, without reference to the others, or they may be studied together, with comparisons of the level of attainment on each. In both cases, the question would be to address how societies compare and why they vary in the attainment of various individual principles. The principles may also be combined into a single ‘measure’ (weighted or unweighted), albeit with the caveat that doing so is a crude empirical exercise, intended to provide a summary measure for establishing how national differences appear to matter for the overall attainment of civic principles.

154   john godard Table 7.1 Frege-​Godard measures of civic principles Liberty/​Freedom (2 item* additive scale, divided by 2) ‘You are free to choose how you do your work’ ‘You are free to alter the times at which you work’. Fulfilment/​Self Development +​ (3 item additive scale, divided by 3) ‘There is a lot of variety in your job’, ‘Your job makes you keep learning new things’, ‘Your job is fulfilling.’ Dignity (3 item additive scale, divided by 3) ‘The person for whom you work treats you with respect’, ‘Management consults workers on major work-​related decisions’, ‘You feel free to openly question a manager’s decision if you disagree with it’ Fairness (4 item additive scale, divided by 4) ‘The way in which work is assigned is fair’, ‘The way in which pay levels are decided is fair.’ ‘The way in which promotions are decided is fair’, ‘Employees are treated fairly when they do something wrong.’ Justice +​(4 item additive scale, divided by 4) ‘If a co-​worker was unjustly dismissed, it is likely that something could be done to make things right’, ‘If a female co-​worker was denied a promotion due to her gender, it is likely that something could be done to make things right’ ‘If a co-​worker was bullied by a manager, it is likely that something could be done to make things right’, ‘ If a co-​ worker was denied pay or bonus money to which she was entitled, it is likely that something could be done to make things right.’ Security (3 item additive scale, divided by 3) ‘You feel that your job is secure’, ‘You can trust your manager’, ‘You get all the support and training you need’ Fraternity/​ Community (6 item additive scale, divided by 6) ‘You can trust your co-​workers’, ‘In general, you consider your co-​workers to be friends’, Everybody where you work is proud to be associated with your employer’, ‘Where you work, everyone is made to feel they belong’, ‘Where you work, everyone strives for the highest possible quality’, ‘Overall, your workplace is well run.’ Democratic Empowerment +​(3 item additive scale, divided by 3, items reverse coded) ‘You and your co-​workers are expected to do what they are told, whether they like it or not’, ‘What happens to you and your co-​workers is pretty much beyond your control’, ‘It does little good for you and your co-​workers to speak up about things.’ Equality Standard deviation of an aggregate job quality index summing each of the above. Note: The items identified with each principle are not meant to be exhaustive of all facets of that principle, but are instead meant to capture the construct empirically. Thus, in a number of cases, specific items from earlier versions were dropped, not because they were irrelevant to the construct in question, but rather because they did not matter to the reliability (Cronbach’s alpha) of the measure and there was need for parsimony in the questionnaire design. * =​all items are four-​point scales, ranging from ‘strongly disagree’ to ‘strongly agree’. +​=​adapted from Godard 2001 and 2010.

It is also possible that there will be conflicts between some principles, either in the form of direct trade-​offs, or in the conditions needed for their attainment. In other words, a particular condition may have positive effects for one principle yet negative effects for another or, at minimum, no effects; yet a different condition, which conflicts with the first, may have no effects for the first but positive effects for the second. Similar possibilities may hold with regard to potential conflicts between the attainment of these

job quality as the realization of democratic ideals    155 principles and economic performance, or between the attainment of these principles and competing principles (e.g. property rights). Yet the possibility (likelihood?) of such conflicts does not reduce the value of relying on civic principles. Rather, it may be seen as presenting interesting questions for theoretical analysis, and ultimately as yielding dilemmas in efforts to restructure institutions and ultimately work so as to enhance their democratic quality.

Studying Civic Principle Attainment The identification of civic principles provides an alternative approach to conceptualizing job quality (and ultimately, perhaps, industrial relations), but not for actually studying variation in its attainment. Such study requires not only measurement of the attainment of civil principles, but also analysis of how their attainment varies at the societal level. There are two significant (though not unrelated) barriers to doing so. The first barrier is the empirical. To begin, these principles amount to subjective experience, which is likely conditioned by expectations and norms (as well as by objective conditions), potentially rendering measures unreliable. This is not as important as for more individualistic conceptions of job quality addressed earlier, but it is likely to matter if there are systematic differences across nations (which there are). In addition, to the extent that pressures for convergence in recent decades have indeed been realized, the range of societal difference may be limited, thereby limiting the ability to establish the importance of societal differences and hence of institutional designs. Finally, attempting to establish, through quantitative analysis, just how societal level differences matter can be difficult given the level of abstraction required, especially (but not only) if these differences are not adequately theorized. The second barrier is the philosophical. Beginning with Marx (and Marx and Engels), there has been an extensive literature on how capitalism, and the nature of work within capitalism, inherently alienates ‘man’ from his self and perverts the development of human faculties (e.g. Marcuse 1964; Bourdieu 2000: 202; Burawoy 1975, 1985, 2012; also see Petersen and Willig 2004). Although authors may not explicitly single out the nature of work or explicitly refer to alienation, these analyses tend to identify similar problems (e.g. Marcuse’s technological rationality, Bourdieu’s ‘misrecognition’, Burawoy’s ‘mystification’). Under these critiques, the problem is not just that individuals experience alienation (or mystification/​misrecognition), it is that they are also unable to assess their own condition. To quote Alain Touraine (1973: 201), ‘Alienation is not the awareness of deprivation, but the deprivation of awareness.’ (as cited in Seeman 1975). If so, empirical analysis may only capture degrees of alienation at best. Indeed, the more alienated a person is, the more likely they may be to respond favourably to questions addressing various civic principles. More important, as long as societal differences rest on the deep

156   john godard structure of capitalism or the culture it generates, attempts to establish their importance are likely to be of limited value for matters of policy. Combined with the need to establish a more philosophically grounded conception of job quality, the second of these two barriers suggests a strong case for developing a ‘philosophy of job quality’ (or even of work/​industrial relations), one that might draw on empirical indicators, but which investigates in a more thorough way the question of alienation (or subjugation) and its implications for the experience of job quality in a modern democracy. The general absence of such analysis might even illustrate a ‘double deprivation of awareness’, under which subjects are not only unaware of their condition, but the academics studying them are also not conscious of their own condition. As a result, the latter engage in empiricist analysis that fails to question and may even legitimate and strengthen established systems of domination (e.g. Godard 2014: 11–​12; Bramel and Fried 1981). This may actually help to explain the problems with the job quality literature identified earlier. One possibility is to attempt to develop ‘objective’ measures. For example, it is commonly assumed that various legal rights and protections, and especially various forms of representation and voice, are associated with many if not all of the civic principles in Frege-​Godard. Yet in spite of a tendency to equate the former with the latter or at least assume a strong connection, these associations cannot be assumed. Such associations are ideally a matter for empirical inquiry, but in the conduct of research it may be neither sensible nor feasible to try to obtain data that is able to establish with a high degree of confidence that they exist. Of equal importance is the need for convincing, theoretically grounded analysis as to why particular associations can be expected or even assumed to exist. Such analysis can be as or more compelling and insightful than conventional empirical analysis, where unduly simple cause–​effect associations are often assumed and findings are often weak and unconvincing. In effect, there is need of better theory as to the ‘causes’ of job quality, ideally focusing at the societal level and, more specifically, on the role of institutional conditions in accounting for variation. This theory would ideally be subject to empirical scrutiny, but would adhere to a theoretical realist ontology (Bhaskar 1978; Godard 1994), addressing the role of both underlying and systematic factors that elude conventional quantitative analysis. In the following section, I argue for an ‘institutional environments’ approach to job quality, as developed in Frege -​Godard and elsewhere (Godard 2008; Frege and Godard 2010). This essentially entails generation of a theory of industrial relations, in which institutional conditions represent the core ‘independent’ and ‘moderating’ variables, and civic principles the core ‘dependent’ variables. However, my intention is to ground it in a deeper theoretical perspective, one that not only enables us to predict and explain these associations, but also enables us to understand them, by accounting for the differences that underpin them. In particular, it is important to not just assess whether and how institutional arrangements are associated with job quality, but also to assess the basis of these arrangements and their implications for efforts (e.g. by policy elites) to enhance job quality.

job quality as the realization of democratic ideals    157

Theorizing Societal Variation in Job Quality There are two predominant theoretical traditions that are potentially useful for exploring how job quality is affected by societal-​level differences (see Gallie 2007b): the production regimes/​varieties of capitalism (VoC) literature (Amable 2003; Coates 2000; Hall and Soskice 2001), and the power resource/​theories of employment regimes literature (Korpi 1978, 1983; Esping-​Andersen 1990; Gallie 2007a). Theories of production regimes and subsequently VoC theory have focused on the firm level, and particularly on employer strategies. ‘The central argument is that quite different employment dynamics can be found between political economies depending on the way that firms try to solve their coordination problems with respect to employment relations, vocational training, corporate governance, inter-​firm relations, and the cooperation of their employees’ (Gallie 2007a: 13). Consequently, VoC theory declares a significant difference between liberal market economies (LMEs), which coordinate their activities primarily in terms of hierarchies and competitive market arrangements, and coordinated market economies (CMEs), which depend more heavily on non-​market arrangements. Many of the factors thought to be crucial for employee well-​being are held to be substantially better in CMEs (Gallie 2007a: 13). Employers in CMEs tend to have product market strategies based on complex ‘diversified quality products’ (Streeck 1992) that ‘depend on skilled and experienced employees’ and are subject to institutional conditions that induce them to seek cooperative employment relations and pursue long-​ term committed relationships. One would therefore assume job quality to be better in workplaces of CME rather than LME countries (see also Coats and Lekhi 2008: 7; Olsen et al. 2010). In contrast, power resource/​employment regime theory focuses on the strength of national labour movements and the importance of different state interventions (employment policies). Thus, employment regimes differ in the power of their labour movements, in employment policies, in the role of the public sector and in the extent of the welfare state (Gallie 2007a: 17).2 For example, according to Gallie, the German vocational training system is not so much the outcome of certain product or firm strategies (as argued by VoC) but a reflection of the powerful German labour movement in the late nineteenth/​early twentieth century and the institutional compromise reached through state intervention. In short, employment regime theory attempts to explain national

2 Gallie actually distinguishes three employment regimes: inclusive (Scandinavian countries), dualistic (Germany) and market regimes (US, UK). Thus, coordinated market economies such as Sweden or Germany share a similar production regime but differ slightly in their employment regime (Sweden has a more centralistic inclusive approach to employment relations whereas Germany practices a dualistic regime focusing on unions at the industry level and works councils at firm level). For present purposes, I stick with the distinction between CMEs and LMEs.

158   john godard differences in employment relations primarily with regard to their employment regimes rather than firm strategies, attributing these to differences in the distribution of power resources rather than to functional requirements. Both of these approaches are of value. Not only do they provide some basis for explaining societal differences in job quality, they also help us to understand why these differences exist in the first instance. Where the power resource/​employment regime approach may help us to understand how nations come to be on particular trajectories, the production systems/​VoC approach helps us to understand how firms adapt to institutional compromises on which these trajectories are built. So, for example, if a powerful labour movement managed to produce a strong vocational training system and to ensure strong representation rights for workers, then employers have developed strategies that enable them to take advantage of both through high-​skilled, high-​value production, in which workers and their representatives are essentially partners in production. The problem is that these approaches are unduly blunt for informing analysis of variation in job quality and in the end result in largely empiricist, atheoretical analysis. Although they provide a variety of reasons to expect variation in job quality, they generally assume a singular causal effect, and provide little basis for systematically establishing the specific mechanisms through which societal differences may matter to job quality. They also tend to understate the importance of ‘cultural’ factors (e.g. norms and expectations) that are implicated in such variation (Barbier 2013: 70–​72, 105–​106). The ‘institutional environments’ approach, especially as developed in Frege-​Godard, builds upon, yet goes beyond, these approaches in order to overcome these limitations.

The Institutional Environments Approach My starting point for attempting to theorize job quality (or employment in general) is recognition that the employment relation in capitalist economies is, first and foremost, a relation of subordination, one that occurs under conditions of interest conflict between buyers and sellers of labour, and is subject to broader conflicts within the economy and society. This creates a basic problem of trust (subordination) and control (conflict). How, and the extent to which, this problem is addressed is in some measure attributable to employer practices, which in turn play a critical role in shaping job quality. However, both the nature of these practices, and the extent to which they are direct responses to the problem of trust and control, are mainly contingent on the broader institutional environment within which they are formulated (e.g. Godard 2004). This environment consists not just of dominant legal/​institutional arrangements (as in the approaches described earlier), but also of corresponding ‘logics of appropriateness’ as to the behaviour of various actors and, in particular, employers (March 1994: 57–​58). Both institutional arrangements and logics of appropriateness are, in turn, underpinned by institutional norms, or deeply held beliefs, values and principles as to the role, rationale for, and legitimacy of established institutional arrangements (Godard 2008).

job quality as the realization of democratic ideals    159 Following from the two approaches identified above, dominant legal/​institutional arrangements have major a priori implications for employer practices and ultimately job quality. These implications reflect both efficiency/​functional imperatives and the power resources that workers and their agents are able to draw upon to protect or further their goals and interests. Thus, for example, Germany can be expected to have higher levels of job quality than the US, because providing high-​quality jobs is consistent with efficiency requirements and because workers have greater power resources to ensure that these requirements are met in ways that serve worker interests as well as those of employers and, indeed, are traded off against these interests where they are in conflict. Dominant logics are important, however, because they mitigate these implications. First, they dampen variation in workplace employment regimes that might otherwise be expected from intra-​national variation in efficiency requirements and power resources. Thus, for example, Frege-​Godard (959) finds that coverage by collective bargaining in Germany makes little difference to job quality, likely because of a broader logic of appropriateness in Germany, yet it does compared to US workers without coverage, likely because a different logic applies in the US. Second, the effects of apparent changes in legal/​institutional conditions are at least partly offset by dominant logics, which may not change at the same rate or in the same way (see Godard 2021). In essence, these logics engender what organization theorists refer to as institutional isomorphism, under which employer practices do not just reflect firm-​specific or even national-​level conditions at a given point in time but also dominant ways of thinking within an ‘organizational field’ (in this case, society). Extant institutional arrangements and logics of appropriateness do not just explain variation in job quality, they also determine (through the actions of the parties) the extent and direction of institutional changes likely to impact on job quality. Institutional norms in turn matter not just because they underpin extant institutional arrangements and logics and hence variation in job quality, but also because they serve as further constraints on the possibilities for change. These norms tend to be deeply embedded, reflective of formative economic, social and political developments in a nation’s history. Although they may allow considerable scope for interpretation and hence politics (e.g. see Jackson 2003), and may even be produced through politics, they embody ‘mobilization biases’ (Godard 2009), defined for present purposes as biases that shape the relative ability of various interests to successfully mobilize resources to effect change, thereby ultimately privileging one or more institutional designs and hence configuration of interests over others.3 Thus, the question is not so much one of what institutional reforms may be conducive to job quality, but rather of whether and how such reforms are conducive to established institutional norms, and to the broader institutional structures and logics of appropriateness underpinned by these norms. In the end,

3 

This term is adapted from the term ‘mobilization of bias,’ coined by Bachrach and Baritz (1962), who give it a narrower meaning than ‘mobilization bias,’ as used in this chapter.

160   john godard changes that cannot be legitimated in accordance with these norms are likely to either receive short shrift or fail over time (see Godard 2009). This perspective is similar to the (much overlooked) employment systems approach developed by Fligstein and Byrkeflot (1996). They seek to identify ideal–​typical employment system logics, arguing that these ultimately reflect political compromises at the origins of industrialization, and that these become embedded in cultural practices and meanings that serve as conceptions of control used to maintain stable (power) relationships within organizational fields. However, the institutional environments approach focuses on underlying institutional norms and their implications for national-​ level institutional and ultimately job-​quality differences, and does not assume these norms to serve as conceptions of control.4 More important here, it posits three additional mechanisms through which institutional environments matter, beyond the a priori implications identified earlier. These are developed more fully in Frege-​Godard and so are only briefly addressed here. First, where institutional arrangements are unfavourable to workers, employer practices become central to addressing (or not addressing) problems of trust and control, and ultimately to determining the extent to which civic principles are attained. One function of a comparatively ‘strong’ institutional environment (e.g. Germany) is that it creates conditions that are more conducive, not only to the attainment of these principles, but also to ‘positive’ industrial relations. But in weaker institutional environments (e.g. the US), employers are more likely to face problems of trust and control because these are not addressed by institutional means. Employers may adopt a variety of practices in order to address these problems (e.g. Friedman 1977; Edwards 1979; Burawoy 1985), some or many of which may be largely coercive, but others of which may enhance job quality and effectively lessen the overall differences between ‘strong’ and ‘weak’ institutional environments, while giving rise to greater inequalities in job quality within the latter. Second, institutional environments can have specific implications for the functioning and characteristics of workplace-​ level representation systems and employment practices, or ‘workplace regimes’. While these differences may reflect different logics, they may also reflect specific institutional conditions. For example, the rights attaching to German works councils ensure that employment practices are not just implemented to serve employer interests, as in the US, but that they are also implemented to serve worker interests. Thus, the nature and implementation of ‘new’ work and HRM practices associated with high-​performance work organizations should be substantially more favourable in Germany than in the US, allowing for greater team autonomy and having stronger positive implications for civic principles. Third, the implications of institutional environments for the attainment of civic principles can depend on societal differences in worker subjectivities, which shape

4 

Fligstein and Byrkeflot also assume organizational fields to be purposively designed by elites. The approach advanced here does not assume purposive design.

job quality as the realization of democratic ideals    161 how workers evaluate their employment. As noted earlier, subjective measures of these principles may be especially subject to differences in expectations. Thus, for example, if workers have higher expectations, they may be more critical in their evaluations of whether various civic principles have been attained. While expectations may vary in accordance with a variety of factors, institutional environments play an important role—​not just because of dominant logics of appropriateness but also because workers enjoy various rights and protections at work that lead them to be less alienated and hence more optimistic about what they can expect. Dominant institutional norms may also play an important role in defining what workers believe they should be entitled to. Thus, for example, in a country with strong property rights norms, workers may view employer obligations to employees as lower than in a country where, these obligations are more strongly entrenched. Finally, in addition to these three basic mechanisms, institutional environments may be associated with differences in the extent to which various principles are valued relative to other principles. This may occur because of differences in logics of appropriateness (i.e. which principles should be most attended to), or in employment regimes (e.g. some principles may be more or less consistent with employer interests or with legal rights), or in expectations created by established institutional conditions. Thus, for example, fraternity/​community may be more highly achieved in the US relative to other principles, because employer logics are more consistent with a ‘unitary’ than a pluralistic workplace culture, employers are in a position to impose rules and practices to achieve such a culture, and workers are more reliant on employment to address their social needs. All of these arguments are borne out by empirical analysis reported in the Frege-​ Godard paper, using data obtained from random telephone interviews of 1,000 US and 1,000 German workers in 2009. This analysis was subject to the same limitations as those discussed earlier and so may not fully pick up ‘true’ variation in job quality or in its institutional antecedents. Nonetheless, it can be considered a starting point in the quest to establish not just how institutional environments matter, but how changes in these environments can be expected to matter, and what changes are most likely to be most conducive to the enhanced attainment of civic principles.

Conclusion This chapter has advanced two main arguments. First, there is need of a societal-​ level analysis of job quality, one that defines job quality in accordance with basic civic principles or ideals that have long been valued in, and are largely foundational to, Western democracies. Second, analysis of societal-​level variation in the attainment of these principles is best informed by an institutional environments approach, under which a ‘deeper’ theory of this variation is developed, and specific mechanisms through which institutional environments matter are identified. In developing these arguments

162   john godard and establishing empirical support for them, I drew extensively from, but went beyond, Frege-​Godard, which addressed differences in the attainment of civic principles across Germany and the US. The approach adopted in this chapter suggests the basis for a much-​needed normative paradigm for the future of work in democratic societies, one that focuses explicitly on the enhanced attainment of basic, civic principles and the institutional conditions conducive to their attainment. Any such paradigm would need to be grounded as much in philosophy as in empirical observation, especially as it would likely envisage substantial institutional and potentially cultural change. Such change could be expected to vary in accordance with institutional traditions and norms, thereby rendering universalist prognostications dominant in the ‘future of work’ literature too simplistic. But as the era of global neoliberalism comes to an end and both robotization and artificial intelligence threaten to displace even ‘good’ jobs, it is possible that we are (finally) entering a truly transformative period for work and employment. This may be especially so given the high per capita GDPs of developed economies, which, coupled with environmental crises, suggests the possibility of an alternative to consumer capitalism. It would seem critical to develop a normative paradigm for work that capitalizes on this possibility, albeit one that goes beyond the present analysis to include an action strategy for ensuring its full integration into the domains of social policy and justice (e.g. Supiot 2013).

Acknowledgements I thank Carola Frege for her contribution to many of the ideas and concepts in this chapter. I also thank both her and Christopher Mathieu for their editorial suggestions. I take sole responsibility, however, for the arguments in this chapter and for any errors or misconceptions they may contain.

References Amable, Bruno (2003), The Diversity of Modern Capitalism (Oxford: Oxford University Press). Bachrach, Peter, and Baritz, Morton (1962), ‘Two Faces of Power’, American Political Science Review, 56/​4, 947–​957. Barbier, Jean-​Claude (2013), The Road to Social Europe. Translated by Susan Gruenheck. (London: Routledge). Bhaskar, Roy (1978), A Realist Philosophy of Science (Hassocks, Eng: Harvester Books). Blauner, Robert (1964), Alienation and Freedom (Chicago: Chicago University Press). Bourdieu, Pierre (2000 [1997]), Pascalian Meditations (Stanford, CA: Stanford University Press). Bozeman, Barry (2007), Public Values and Public Interest: Counterbalancing Economic Individualism (Washington, DC: Georgetown University Press).

job quality as the realization of democratic ideals    163 Bramel, Dana, and Friend, Ronald (1981), Hawthorne, the Myth of the Docile Worker, and Class Bias in Psychology’, American Psychologist, 36/​8, 867–​878. Braverman, Harry (1974), Labor and Monopoly Capital (New York: Monthly Review Press). Buddelmeyer, Heilke, McViar, Duncan, and Wooden, Mark (2015), Nonstandard “Contingent Employment” and Job Satisfaction’, Industrial Relations, 54/​2, 256–​275. Burawoy, Michael (1975), Manufacturing Consent (Chicago: University of Chicago Press). Burawoy, Michael (1985), Politics of Production (London: Verso). Burawoy, Michael (2012), ‘The Roots of Domination: Beyond Bourdieu and Gramsci’, Sociology, 46/​2, 187–​206. Clark, A.E. and Oswald, A.J. (1996), ‘Satisfaction and Comparison Income’, Journal of Public Economics 61/​3, 359–​381. Coates, David (2000), Models of Capitalism: Growth and Stagnation in the Modern Era (Cambridge: Polity Press). Coates, David and Lekhi, R. (2008), Good Work: Job Quality in a Changing Economy (London: The Work Foundation). Cooke, Gordon B., Donaghey J., and Zeytinoglu I. U. (2013), ‘The Nuanced Nature of Work Quality: Evidence from Rural Newfoundland and Ireland’, Human Relations 66/​4, 503–​528. Edwards, Richard (1979), Contested Terrain (New York: Basic Books). Erikson, Kai (1986), ‘On Work and Alienation’, American Sociological Review, 51/​1, 1–​8. Esping-​ Anderse, Gosta (1990), The Three Worlds of Welfare Capitalism (Princeton NJ: Princeton University Press). Findlay, Patricia, Kalleberg, Arne L., and Warhurst, Chris (2013), ‘The Challenge of Job Quality.’ Human Relations, 66/​4, 441–​451. Fligstein, Neil, and Byrkeflot, Haldor (1996), ‘The Logic of Employment Systems’, in James Baron, David B. Grusky, and Donald J. Trieman, eds., Social Differentiation and Stratification (Boulder: Westview Press), 11–​34. Frege, Carola, and Godard, John (2010), ‘Institutional Norms and Cross-​national Variation in Representation Rights at Work,’ in Adrian Wilkinson, P. Gollan, M. Marchington, and D. Lewin, eds., The Oxford Handbook of Participation in Organizations (Oxford: OUP), 526–​552. Frege, Carola, and Godard, John (2014), ‘Varieties of Capitalism and Job Quality: The Attainment of Civic Principles at Work in the USA and Germany’, The American Sociological Review, 79/​5, 942–​965. Friedman, Andrew (1977), ‘Responsible Autonomy Versus Direct Control Over the Labour Process’, Capital and Class, 1/​1, 43–​57. Gallie, Duncan (2007a), ‘Task Discretion and Job Quality’, in Duncan Gallie, ed., Employment Regimes and the Quality of Work (Oxford: Oxford University Press), 105–​136. Gallie, Duncan (2007b), ‘Production Regimes, Employment Regimes, and the Quality of Work’, in Duncan Gallie, ed., Employment Regimes and the Quality of Work (Oxford: Oxford University Press), 1–​34. Gallie, Duncan, Felstead, Alana, and Green, Francis (2004), ‘Changing Patterns of Task Discretion in Britain’, Work, Employment and Society, 18/​2, 243–​266. Godard, John (1994), ‘Beyond Empiricism: Toward a Reconstruction of IR Theory and Research’, in David Lewin and Donna Sockell. Advances in Industrial and Labor Relations. Vol. 6 (Greenwich, Conn: JAI Press), 1–​36.

164   john godard Godard, John (2001), ‘High Performance and the Transformation of Work? The Implications of Alternative Work Practices for the Experience and Outcomes of Work’, Industrial and Labor Relations Review, 54/​4 (July), 776–​805. Godard, John (2004), ‘The New Institutionalism, Capitalist Diversity, and Industrial Relations’, in Bruce Kaufman, ed., Theoretical Perspectives on Work and the Employment Relationship (Urbana-​Champaign: IL: Industrial Relations Research Association), 229–​264. Godard, John (2008), ‘An Institutional Environments Approach to Industrial Relations’, in Charles Whalen, ed., Industrial Relations as an Academic Enterprise (Northampton MA: Edward Elgar), 68–​88. Godard, John (2009), ‘The Exceptional Decline of the American Labor Movement’, Industrial and Labor Relations Review, 63/​1, 81–​107. Godard, John (2010), ‘What’s Best for Workers? The Implications of Work and Human Resource Practices Revisited’, Industrial Relations, 49/​3 (July), 465–​487. Godard, John (2014), ‘The Psychologization of Employment Relations?’ Human Resource Management Journal, 24/​1, 1–​18. Godard, John (2021), ‘Work and Employment Practices in Comparative Perspective’, in Carola Frege and John Kelly, eds., Comparative Employment Relations in a Global Economy. 2nd edn. (London: Routledge). Green, Francis (2001), ‘It’s Been a Hard Day’s Night: The Concentration and Intensification of Work in Late 20th Century Britain’, British Journal of Industrial Relations, 39/​1, 53–​80. Hall, Peter, and Soskice, David (2001), ‘An Introduction to Varieties of Capitalism’, in Peter Hall and David Soskice, eds., Varieties of Capitalism: The Institutional Foundations of Comparative Advantage (Oxford: Oxford University Press) 1–68. Inanc, Hande (2018), ‘Unemployment, Temporary Work, and Subjective Well-​being: the Gendered Effect of Spousal Labor Market Insecurity’, American Sociological Review, 83/​4, 536–​566. Inglehart, Ronald, and Welzel, Christian (2005), Modernization, Culture Change, and Democracy (Cambridge: Cambridge University Press). Jackson, Gregory (2003), ‘Corporate Governance in Germany and Japan’, in Kozo Yamamura and Wolfgang Streeck, eds., The End of Diversity? (Ithaca: Cornell University Press) 261–305. Judge, Timothy A., Piccolo, Ronald F., Podsakoff, Nathan P., Shaw, John C., and Rich, Bruce L. (2010), ‘The Relationship between Pay and Job Satisfaction: A Meta-​Analysis of the Literature’, Journal of Vocational Behaviour, 77/​2, 157–​167. Kalleberg, Arne (2008), ‘The Mismatched Worker: When People Don’t Fit Their Jobs’, Academy of Management Perspectives, 22/​1, 22–​40. Kalleberg, Arne (2011), Good Jobs, Bad Jobs (New York: Russell Sage Foundation). Kalleberg, Arne (2016), ‘Good Jobs, Bad Jobs’, in Stephen Edgell, Heidi Gittfried, and Edward Granter, eds., Sage Handbook of the Sociology of Work and Employment (Thousand Oaks, CA: Sage) 111–​128. Korpi, Walter (1978), The Working Class in Welfare Capitalism (London: Routledge and Kegan Paul). Korpi, Walter (1983), The Democratic Class Struggle (London: Routledge and Kegan Paul). Kymlicka, Will (2002), Contemporary Political Philosophy. 2nd edn. (Oxford: Oxford University Press). Leschke J., and Watt A. (2008), Job Quality in Europe. ETUI-​REHS Research Department Working Paper 2008.07.

job quality as the realization of democratic ideals    165 Lopes, Helena (2015), ‘The Political and Public Dimension of Work’, Journal of Australian Political Economy, 76, 5–​26. March, James G. (1994), Primer on Decision Making: How Decisions Happen (New York: Free Press). Marcuse, Herbert (1964), One Dimensional Man: Studies in the Ideology of Advanced Industrial Society (London: Routledge Kegan and Paul). Marx, Karl (1959 [1844]), Economic and Philosophical Manuscripts (Moscow: Progress Publishers). Marx, Karl, and Engels, Frederick (1947 [1846]), The German Ideology (New York: International Publishers). Mayo, Elton (1933), The Human Problems of an Industrial Civilization (New York: The MacMillan Company). OECD (2017), OECD Employment Outlook (Paris: OECD). Olsen, Karen M., Kalleberg, Arne, and Neheim, Torstein (2010), ‘Perceived Job Quality in the United States, Great Britain, Norway, and west Germany’, European Journal of Industrial Relations, 16/​3, 221–​240. Osterman, Paul, and Shulman, Beth (2011), Good Jobs America (New York: Russell Sage Foundation). O’Toole, James et al. (1974), Work in America (Cambridge MA: MIT Press). Petersen, Anders, and Willig, Rasmus (2004), ‘Work and Recognition’, Acta Sociologica, 47/​4, 338–​350. Seeman, Melvin (1959), ‘On the Meaning of Alienation’, American Sociological Review, 24, 783–​791. Seeman, Melvin. (1975), ‘Alienation Studies’, Annual Review of Sociology, 1, 91–​123. Smith, Adam (1936 [1776]), An Inquiry into the Nature and Causes of the Wealth of Nations (New York: Modern Library). Stone, Katherine V. W. (2013), ‘The Decline in the Standard Employment Contract: The Evidence’, in Katherine Stone and Harry Arthurs, eds., Rethinking Workplace Regulation: Beyond the Standard Contract of Employment (New York: Russell Sage), 366–​404. Streeck, W. (1992), Social Institutions and Economic Performance: Studies of Industrial Relations in Advanced Capitalist Economies (London: Sage). Supiot, Alaine (2013), ‘Grandeur and Misery of the Social State’, New Left Review (July–​August), 82, 99–​113. Touraine, Alain (1973), Production de la Societe (Paris: Editions du Seuil). Warr, Peter (2007), Work, Happiness, and Unhappiness (London: Lawrence Erlbaum). Weber, Max (1968 [1922]), Economy and Society. (G. Roth and K. Wittich, ed.) (New York: Bedminster). Wilkin, Crista (2013), ‘I Can’t Get No Job Satisfaction: Meta-​Analysis Comparing Permanent and Contingent Workers’, Journal of Organizational Behaviour, 34/​1, 37–​64. Winch, D. (2004), Smith, Adam. Oxford Dictionary of National Biography (Oxford: OUP).

part I I I

K E Y I S SU E S I N J OB QUA L I T Y

chapter 8

job polarization Its history, an intuitive framework and some empirical evidence maarten goos, emilie rademakers, anna salomons and marieke vandeweyer

Introduction Between the mid-​1970s to mid-​1980s the US labour market created nearly 26 million net new jobs, more than virtually all of the other industrialized countries combined. As a result, several studies about the ‘The Great American Jobs Machine’ emerged in the literature. However, even at the time, there were questions about the quality of jobs created, with evidence suggesting that a substantial proportion of new jobs were in low-​ wage employment. This controversy over the low-​wage question intensified after the US Joint Economic Committee (JEC) released a report in December 1986 claiming that, based on Current Population Survey (CPS) data, the share of net new employment paying low wages had increased significantly after 1979. In particular, the JEC report indicated that between 1979 and 1984 the share of net new employment paying low wages (i.e. below half of the 1973 real median annual wage) was nearly 60 per cent, in contrast to the 1963–​1979 period when this share was less than 20 per cent. The JEC findings sparked a debate that has become known as the job-​quality debate of the 1980s. For example, Kosters and Ross (1987, 1988) also use CPS data but apply a different methodology to conclude that there was no rise in the share of new jobs with low wages. In a reaction to Kosters and Ross (1987), Bluestone and Harrison (1988) again use CPS data and yet another methodology to find that, in line with the JEC report, there was an increase in the share of low-​wage work. Together with an increase in the share of high-​wage work at the expense of middling jobs, Bluestone and Harrison (1988)

170    maarten goos et al concluded that the Great American Jobs Machine of the early 1980s was faltering as a result of job polarization. Starting in the mid-​1980s, the job-​quality debate faded among economists as most labour economists came to the conclusion that the problem for unskilled workers was a declining rather than an increasing number of jobs for them because of Skill-​Biased Technological Change (SBTC) (Bound and Johnson 1992; Katz and Murphy 1992). The SBTC hypothesis is based on the assumption that technological progress increases the productivity of skilled workers by more than that of unskilled workers, thereby shifting out labour demand for skilled workers by more than for unskilled workers.1 At the same time, educational attainment rates increased, thereby shifting out the supply of skilled relative to unskilled workers. An important insight from the SBTC framework is that skill-​biased technological progress and increases in educational attainment combine to skill-​upgrading in the structure of jobs. The success of this simple supply–​demand framework came from its predictive power to explain changes in the skill-​premium over time (i.e. the ratio of the average wage for skilled over unskilled workers), driven by accelerations and decelerations in the outward shifts in the supply of and demand for skilled relative to unskilled workers. For example, Goldin and Katz (2009) use US data to show that the observed short-​run decline of the skill-​premium in the 1970s and sharp increase in the 1980s is in large part explained by an acceleration and then deceleration in the supply of skilled relative to unskilled workers, with the demand for skilled workers relative to unskilled workers growing steadily during both decades due to SBTC. Because the simple SBTC framework goes a substantial distance towards explaining the evolution of skill-​premia over time in the US and elsewhere, it became known as the canonical model of the 1990s to think about the impact of technological progress on labour markets for skilled and unskilled workers.2 However, a number of studies from the 1990s continued to address the major themes of the job-​quality debate of the 1980s (see, for example, Costrell 1990; Howell and Wolff

1 Some

authors reserve the term ‘SBTC’ as an umbrella for the broad set of theories developed in the 1980s, 1990s and 2000s to think about the impact of digital technologies on labour markets. In this chapter, however, we refer to SBTC only as the set of theories developed in the 1980s and 1990s assuming that technological progress is increasing the productivity of skilled as well as unskilled workers, but more so for skilled than for unskilled workers. This assumption differs from other labour market models of technological progress. For example, there is a related hypothesis known as Capital-​ Skill Complementarity (CSC) (Griliches 1969; Berman et al. 1994). Just as SBTC, CSC also predicts an outward shift in the demand for skilled relative to unskilled workers following technological progress. In contrast to SBTC, however, CSC assumes that this shift results from a decrease in the price of capital and the relative complementarity of skilled versus unskilled workers with capital in production. More recently, task assignment models of the 2000s no longer assume that technological progress is labour augmenting, but that it directly substitutes for routine labour tasks in production. See Goos (2018) for a discussion. 2  There is considerable debate, however, whether the canonical framework can go an equal distance toward explaining differences in the skill-​premium between countries. See, for example, Blau and Kahn (1996) and Leuven et al. (2004).

job polarization   171 1991; Levy and Murnane 1992; Juhn et al. 1993; Murphy and Welch 1993; Gittleman and Howell 1995; Ilg 1996; Farber 1997; Acemoglu 1999, 2001; Juhn 1999; Ilg and Haugen 2000). A common theme that emerges from these studies is that, between 1970 and 2000, there has been an increase in the number of high-​paid but also low-​paid jobs relative to medium-​paid jobs. In February 1997, Clinton’s Council of Economic Advisors (CEA) published an influential report claiming that, when one defines a job as an industry-​occupation cell and ranks jobs by their median hourly wages, most of new employment growth in the US in the period 1994–​1996 was in jobs paying above the economy-​wide median wage (US Council of Economic Advisors 1996). In a reaction to the CEA report, Wright and Dwyer (2003) used a similar methodology to conclude that the 1990s’ economic boom indeed resulted in a strong expansion in highly-​paid jobs. However, Wright and Dwyer (2003) also show that the job expansion of the 1990s was less strong overall than in earlier periods of expansion and characterized by job polarization. Moreover, Wright and Dwyer (2003) explain this pattern of job polarization during the 1990s by a decline of middling jobs mainly in manufacturing; stronger growth of bottom-​end jobs in retail trade and personal services; and stronger growth of high-​end jobs in high-​tech sectors. Inspired by Wright and Dwyer (2003), Goos and Manning (2007) use British data to examine employment changes in much more finely grained occupation–​industry cells for the period 1979–​1999. In line with Wright and Dwyer (2003), Goos and Manning (2003, 2007) also find that there is job polarization. Moreover, the authors show that job polarization, rather than being related to the business cycle, has been a steady process at least since the early 1980s. Following Goos and Manning (2003, 2007), the secular phenomenon of job polarization was documented in the economics literature by, among others, Autor et al. (2006) and Autor and Dorn (2013) for the US; and Goos et al. (2009, 2014) for Europe. The uncovering of job polarization as a pervasive phenomenon across advanced economies resulted in the re-​emergence of the job-​quality debate, known as the new job-​quality debate of the 2000s. To explain job polarization, Goos and Manning (2007) refer to the task approach to labour markets, introduced by Autor et al. (2003) and later formalized by the task assignment models of Acemoglu and Autor (2011) and Acemoglu and Restrepo (2018). In these task assignment models, it is assumed that technological progress directly replaces workers doing routine tasks that are codifiable in computer language. These models also assume that differently skilled workers sort across different tasks in equilibrium: following the automation of routine labour tasks, unskilled workers reallocate to jobs that require less complex labour tasks. At the same time, technological progress increases the complexity of tasks done by skilled workers. To illustrate this, consider the automation of machine operators (e.g. car assembly-​ line workers being replaced by robots) or of office clerks (e.g. bank tellers being replaced by online banking). The tasks done by machine operators or office clerks aren’t necessarily the easiest for workers to do (e.g. they often require some degree of training) but are automatable because they are routine in nature such as following procedures

172    maarten goos et al according to set standards or collecting and storing information. Following the automation of these routine tasks, the labour tasks that remain for workers are either less complex or more complex for humans to do. Examples of less complex labour tasks increasingly done by unskilled workers are personal services such as cleaning or waiting tables in a restaurant. Although most humans find these jobs easy to do, their tasks are often not very routine and therefore not automatable. Finally, the increased use of new technologies also requires skilled workers to do more complex tasks that remain difficult to automate, such as performing surgery or managing a team. Also, skilled workers increasingly use new technologies to do these tasks. Because displaced workers supply their labour to jobs requiring less complex as well as more complex tasks, technological progress leads to job polarization rather than (only) to skill-​upgrading. The task assignment models of Acemoglu and Autor (2011) and Acemoglu and Restrepo (2018) are increasingly considered to be the new canonical framework of the 2010s to think about the impact of ongoing technological progress on the quality of jobs. To understand the channels that explain job polarization in the new canonical framework more intuitively, first consider the reallocation of unskilled workers to less complex jobs. Digital technologies have automated many routine tasks, directly displacing relatively unskilled machine operators and office clerks from their jobs. Consequently, these displaced unskilled workers are increasingly employed in unskilled services such as cleaning hotel rooms or waiting tables in restaurants. Although these unskilled services are easy to do for most humans, it will remain difficult for many of the tasks involved to be automated any time soon. Moreover, displaced unskilled workers move into unskilled services because consumer demand for these services is price inelastic. This is important if technological progress is leading to a cost and, therefore, a price disadvantage for unskilled services, because there is less scope to use innovation to reduce costs and therefore output prices in unskilled services relative to other more innovative sectors. However, this price disadvantage and consequent falls in relative product and therefore labour demand are quantitatively not important if consumer demand for unskilled services is relatively price inelastic. A similar intuition can explain the reallocation of skilled workers into more complex jobs that require non-​routine abstract tasks that are difficult to automate, such as teaching students or managing a company’s long-​term strategies. Moreover, new technologies can often complement skilled workers in doing these non-​routine abstract tasks. For example, computers complement the work of computer engineers. Finally, the mobility of skilled workers into more complex tasks is explained by the fact that consumer demand for products from innovative sectors are generally price elastic. What this implies is that innovative sectors can reduce their output prices relative to less innovative sectors, thereby substantially increasing their sales and therefore employment of skilled workers. In sum, the intuition that economists use to explain job polarization is based on the impact of technological progress on various channels that shape equilibria on labour and product markets. The remainder of this chapter further formalizes this intuition in

job polarization   173 the simplest possible framework and provides some empirical evidence in support of this framework. We will show that job polarization is not a natural law that must hold at all times and in all places, but that it critically depends on the nature of technological progress and how it affects equilibria on labour and product markets. This chapter is further organized as follows. Section 2 presents a stylized but intuitive framework to understand the various drivers of job polarization and the predictions we take away from it. Before testing these predictions, Section 3 presents our data. We make use of a broad sample of data spanning 28 sectors in the US and 12 European countries for the period 1980 to 2005. Section 4 presents our empirical results. Finally, Section 5 concludes.

A Simple Framework to Understand Job Polarization Set-​up: Technological Progress: Moore’s Law Following Autor and Dorn (2013), assume that efficiency units of computer capital used in Sector j at time t, K j (t ), result from:

K j (t ) = Ykj (t ) e δt (1)

where δ captures the speed of digital innovation and where Ykj (t ) is output that is not consumed but reinvested in the production of computer capital. Competitive capital and output markets imply that marginal revenue must equal marginal cost in each sectorj:

pk (t ) = p j (t )

∂Ykj (t ) ∂K j (t )

= p j (t ) e − δt (2)

where pk is the unique price per efficiency unit of computer capital and p j the price of good j produced by sector j. Defining pk (t ) ≡ pk (t ) / p j (t ), equation (2) can be written as:

pk (t ) = e − δt (3)

Equation (3) formally summarizes Moore’s Law. In 1965, Intel co-​founder Gordon Moore posited that innovation in microprocessors would lead to a doubling of the number of transistors in a dense integrated circuit every two years, implying a corresponding exponential decrease in their unit cost. This has become known as Moore’s Law and its predictions, by and large, seem to capture reasonably well the observed speed of technological progress in computing hardware as well as a wider range of digital technologies

174    maarten goos et al (Koh and Magee 2006; Nordhaus 2007; Nagy et al. 2013). For example, Nordhaus (2007) estimates that between 1980 and 2006 the real costs of performing a standard set of computations, measured by the cost expressed in constant dollars or relative to labour costs, has fallen by 60 to 75 per cent annually. Equation (3) captures this exponential decrease over time in the relative price per efficiency unit of computer capital, limiting to zero assymptotically: pk (t ) = e − δt → 0 as t → ∞.

Unskilled Services and Polanyi’s Paradox Assume that there is a sector in the economy providing Services, Ys :

Ys = Lm (4)

with Lm workers performing manual labour tasks. The manual labour tasks done in these jobs mainly are an end-​product in themselves such as assisting or caring for others, in the way receptionists, waiters in restaurants or other in-​person services do. These jobs mainly consist of manual tasks that do not require much schooling or experience. Paradoxically, as of the first decades of the twenty-​first century, many of these manual tasks remain difficult to automate. For example, most of us know how to be a good waiter in a restaurant but none of us knows how to write this down in software code. Autor (2015a) refers to this paradox as Polanyi’s paradox, arguing that tacit knowledge—​that is, knowledge that cannot be easily formalized or put into exact words—​is important because we ‘know more than we can tell’ and this poses significant hurdles to automation of in-​person services.

Innovating Sectors with Different Capital Intensities Assume that there are two additional goods producing sectors in the economy, Sector 1 and Sector 2. Sector 1 produces good Yg1 by combining abstract and routine tasks. Abstract tasks can only be supplied by (skilled) labour, whereas routine tasks are done by (unskilled) labour and capital:

1

Yg 1 = L1a−1β1 X β1 with X ≡  Lµr + K1µ  µ (5)

with La1 abstract labour tasks, X total input of routine tasks consisting of Lr routine labour tasks and K1 capital tasks. Assuming that 0 < µ < 1 implies that the elasticity of substitution between routine labour and capital tasks is larger than unity, i.e. 1 / [1 − µ ] > 1, and therefore also larger than the elasticity of substitution between abstract labour and routine tasks (which equals unity given the Cobb-​Douglas speci­ fication in equation (5)).

job polarization   175 Also assume that Sector 2 produces Yg 2 by combining abstract labour tasks and capital: Yg 2 = L1a−2β2 K 2β2 (6)



with La2 abstract labour tasks and K 2 capital. An important difference between equations (5) and (6) is the assumption that in Sector 1 routine tasks are performed by both labour and capital, whereas in Sector 2 routine tasks are performed exclusively by capital. Further assume that 1 > β2 > β1 > 0. What this implies is that Sector 2 is more routine-​ task intensive and also more capital intensive than Sector 1.3 Finally, β2 and β1 are also the constant shares of output going to routine tasks in Sectors 2 and 1 respectively. Consequently, β2 > β1 implies that the share of output going to routine tasks and, therefore, capital is larger in Sector 2 than in Sector 1. Said differently, the share of output going to workers is smaller in Sector 2 than in Sector 1.

Consumption Assume that consumers maximize utility by consuming Services, Cs , Sector 1 goods, C g1, and Sector 2 goods, C g 2, according to: σ

θ

σ −1 σ −1 θ −1 θ −1  σ −1    θ −1 U = Cs σ + C σ  with C ≡ C g 1θ + C g 2θ  (7)    



with σ the elasticity of substitution between Services and other goods and with θ the elasticity of substitution between Sector 1 and Sector 2 goods in consumption. We assume that θ > 1 and that 0 < σ < θ, or that Sector 1 and 2 goods are more substitutable in consumption than Services. This assumption implies that consumer demand for Services will be relatively price inelastic. The intuition for this is that Services are not readily dispensable if their relative prices would increase. For example, the demand for childcare by parents or for hotels and catering by business travellers is unlikely to dissipate in the face of falling prices for sending text messages. Also note that preferences are assumed to be homothetic to maintain the independence of preferences on the demand side and technological progress on the supply side, as is most standard in recent work (see, for example, Ngai and Pissarides 2007; Acemoglu and Guerrieri 2008; Autor and Dorn 2013; Goos et al. 2009, 2014). Finally, since a fraction of final goods produced is allocated towards capital production, consumption equals: C gj = Ygj − Ykj = Ygj − pk K j for j = 1, 2 and Cs = Ys . 3  That

Sector 2 is more routine-​task intensive than Sector 1 immediately follows from comparing equations (5) and (6) and the assumption that β2 > β1. In addition, it also implies that Sector 2 is more capital intensive than Sector 1. This is intuitively clear from the fact that routine tasks in Sector 2 can only be produced by capital, whereas in Sector 1 they are produced by both capital and labour. A formal proof that Sector 2 is more capital intensive than Sector 1 is given in equations (13) and (14).

176    maarten goos et al

Labour Supply Assume that there is a continuum of skilled as well as of unskilled workers, each normalized to unity. Each skilled worker can perform abstract labour tasks in Sector 1 or 2. The fractions of skilled workers employed in Sectors 1 and 2 are denoted by La1 ∈[0, 1] and La 2 = 1 − La1 ∈[0, 1] respectively. Similarly, an unskilled worker can perform manual labour tasks in Services or routine labour tasks in Sector 1. The fractions of unskilled workers employed in Services and Sector 1 are denoted by Lm ∈[0, 1] and Lr = 1 − Lm ∈[0, 1] respectively. For simplicity, skilled workers are assumed to be perfectly mobile between Sectors 1 and 2 but cannot be employed in Services. Unskilled workers are perfectly mobile between Services and Sector 1 but are not employed in Sector 2.

Model Predictions Our framework predicts job polarization under a set of critical assumptions which we discuss here. For a more detailed exposition and formal proofs we refer the reader to Appendix A.

Job Polarization Job polarization requires that unskilled workers move from Sector 1 to Services, and skilled workers move from Sector 1 to Sector 2. Asymptotically, we must have that Lr → 0 and that La1 → 0. We discuss each of these in turn. Unskilled workers reallocate to less capital-​intensive services if new technologies easily substitute for routine labour tasks and the demand for less capital-​intensive services is price inelastic.

Since the relative price of capital falls to zero in each sector, capital will accumulate and limit to infinity: K1 → ∞ and K 2 → ∞ if pk → 0. The accumulation of capital in Sector 1 drives unskilled workers to reallocate to Services under one additional condition:

Lr → 0 if

1 β1 − µ > (8) θ β1

The inequality in equation (8) requires that µ is large relative to θ. A large µ relative to θ (and thus relative to σ given that θ > σ) points to the following two opposing forces for unskilled workers: (i) a negative direct substitution effect away from Sector 1, driven by the substitution in the production of routine tasks towards cheaper capital (i.e. 1 / [1 − µ ] is large); (ii) a positive scale effect towards Sector 1 if its output price decreases relative to the price of Services. However, this scale effect is small if the demand for Services is price inelastic, that is θ and, therefore, σ is small.

job polarization   177 In sum, unskilled workers will reallocate to less capital-​intensive sectors if the substitutability between their routine labour tasks and new technologies is high relative to the price elasticity (in absolute terms) of consumer demand for unskilled services. Skilled workers reallocate to more capital-​intensive sectors if the demand for goods from innovating sectors is price elastic.

To predict job polarization, we must not only have that Lr → 0 but also that La1 → 0. In this paragraph, we show that this will always be the case in our framework given that θ > 1, that is, given that the demand for goods from innovating sectors is price elastic. Given that β2 > β1, a decline in the price of capital has two important consequences for the employment of skilled workers. First, it implies that a decline in the price of capital has a smaller positive impact on labour demand in Sector 2 than in Sector 1. To understand this, note that a large β2 implies a large constant capital share and, thus, a small constant labour share in Sector 2. Consequently, the possible increase in spending on skilled workers in Sector 2 is limited following a decline in the price of capital. This effect would tend to work against the reallocation of skilled workers from Sector 1 to Sector 2. Second, β2 > β1 leads to a larger decline in the output price of Sector 2 than of Sector 1.4 Consequently, the demand for abstract labour tasks in Sector 2 relative to Sector 1 increases because the decline in the output price of Sector 2 relative to Sector 1 leads to a stronger positive scale effect in Sector 2 relative to Sector 1. This effect is larger the larger is the price elasticity of demand for goods from innovating sectors, that is, the larger is θ. This effect drives the reallocation of skilled workers from Sector 1 to Sector 2 in our model. In sum, skilled workers reallocate to more capital-​intensive sectors if consumer demand for goods from innovating sectors is price elastic.

Changes in Relative Output Prices and Volumes From the reallocation of workers discussed earlier we can predict changes in relative output prices and output volumes. If  Lr → 0 and La1 → 0, we have that:

ps → ∞ (9) p2



p1 → ∞ (10) p2

The increase in the price of Services and of Sector 1 goods relative to Sector 2 goods is driven by our assumption that the routine task intensity of Sector 2 is larger than of Sector 1 and also larger than of Services where it equals zero (β2 > β1 > 0). Following technological progress, the greater accumulation of routine tasks in Sector 2 is driving down its output price relative to Services as well as relative to Sector 1. 4 

Also see equation (10).

178    maarten goos et al These predicted changes in relative output prices go hand in hand with relative changes in consumption and, therefore, output volumes: Cg 2



Cs Cg 2



Cg1

→ ∞ (11)

→ ∞ (12)

In sum, our framework not only predicts job polarization but also a relative decline of output prices and a relative increase of output volumes for more capital-​intensive sectors.

Empirical Implementation Define the capital intensity of Sector j = 1, 2,CAPj, as the ratio of rents paid out to capital over total output value. Assuming perfectly competitive labour, capital and output markets gives: CAP1 ≡



p k K 1 ∂ Yg 1 K 1 Kµ = β1 µ 1 µ ≤ β1 (13) = Yg 1 ∂ K 1 Yg 1 Lr + K1

CAP2 ≡



p k K 2 ∂ Yg 2 K 2 = β2 (14) = ∂ K 2 Yg 2 Yg 2

such that β2 > β1 implies that CAP2 > CAP1. Also note that the capital intensity of Services is CAPs = 0. That is, we can rank sectors by their ICT-​capital intensity and use this ranking to test predictions from our framework. Specifically, we estimate the following regression equation: Yjct = b0 + b1 rank[CAP ]jc + b2 rank[CAP ]jc 2 + γ c + ηt + ε jct (15)



where Yjct is the five-​year percentage point difference between years t and t − 5 in the outcome variable of sector j in country c; rank[CAP ]jc is the ranking of sector j in country c by ICT-​capital intensity in 2005; γ c and ηt are country and year fixed-​effects respectively; and ε jct is an error term. By including fixed-​effects we rule out that our outcomes are driven by general trends or country-​specific shocks.5 5 Note

that these fixed-​effects are redundant when the outcome variable is employment shares because these shares are calculated at the country-​year level.

job polarization   179 Note that the ranking of ICT-​capital intensity is time invariant, because the ranking of sectors by their ICT-​capital intensity is found to be stable over time.6 Therefore, the ICT-​capital intensity of a sector j in country c in 2005, CAPjc ∈[0, 1], is used to create the ranking in the empirical analysis that follows. However, using ICT-​capital intensity in 1980 or a measure averaged over time would not qualitatively change our results. Our model predicts that a U-​shaped relationship exists between a change over time in a sector’s employment share and its ICT-​capital intensity ranking. In terms of equation (15), this implies that b1 < 0 and b2 > 0. Our model also predicts monotonic changes in relative output prices and output volumes. Firstly, the conditions we have discussed showed that, over time, relative output prices decline in sectors with higher ICT-​capital intensity. Secondly, this coincides with relative output volumes increasing in sectors with higher ICT-​capital intensity. For our empirical equation (15), this implies that b2 = 0 and that b1 < 0 for output prices but b1 > 0 for output volumes.

Data Sources To test the predictions made by our framework, EUKLEMS data from the March 2011 release are used for the period 1980–​2005.7 This dataset contains information about labour and capital inputs as well as output volume and prices in value-​added for the US and 12 European countries at the 2-​digit level of ISIC(rev.3) sectors reported in Table 8.1. The primary sectors (‘Agriculture’ and ‘Mining’) and the sector ‘Private households with employed persons’ are discarded due to limited data availability, resulting in a final sample of 28 sectors in 13 countries for the empirical analyses that follow. The 13 countries are Austria, Belgium, Denmark, Finland, France, Germany, Italy, Ireland, Netherlands, Spain, Sweden, United Kingdom and United States. Column (1a) of Table 8.1 ranks sectors according to their ICT-​capital intensity in 2005. This measure of ICT-​capital intensity is given by a sector’s ICT-​capital compensation as a percentage of value added averaged across countries.8 The least ICT-​capital-​intensive sectors included in the empirical analysis that follows are ‘Real estate’, ‘Construction’, ‘Hotels & restaurants’, ‘Health & social work’ and ‘Education’. The most ICT-​capital-​intensive sectors are ‘Post & telecommunications’, ‘Financial intermediation’ and ‘Business activities’. 6  Table B.1 and Table B.2 in Appendix B show Spearman rank correlation coefficients of sectors ranked by their ICT-​capital intensity across years by country and across countries for the year 2005. 7  See Timmer et al. (2007) for further details about EUKLEMS data. Unfortunately, more recent updates of the dataset do not include the breakdown of capital compensation into ICT and non-​ICT which is crucial in linking our framework to the empirics as explained in section 2. 8  EUKLEMS defines ICT-​capital compensation as the product of the ICT-​capital stock (consisting of office, process control and computing equipment, communication equipment and software) and its user cost. Stocks are derived from investment flows in ICT-​capital using the Perpetual Inventory Method. See Table B.3 in Appendix B for more information on the content of ICT-​capital.

Table 8.1 ICT-capital intensity, employment shares, output volume and output price by sector averaged across countries, 1980–2005 (1) ICT-​capital intensity (a) % 2005

(b) %-​point change 1980-​2005

(2) Employment shares (a) % 2005

(3) Output volume

(4) Output price

(b) %-​point change % change % change 1980-​2005 1980-2​ 005 1980-​2005

Private hhs with employed persons 0.03

0.03

1.29

0.22

87.73

233.85

Agriculture

0.63

0.41

4.34

–​5.69

29.38

42.96

Real estate

0.84

0.16

1.16

0.42

116.60

262.69

Construction

1.40

0.99

8.09

–​0.02

46.41

186.92

Hotels & restaurants

1.57

–​0.21

4.66

1.23

72.44

240.29

Health & social work

1.63

0.68

10.44

3.30

115.32

246.19

Mining

1.75

1.19

0.22

–​0.45

126.56

123.28

Education

1.75

0.75

6.07

0.90

58.45

246.95

Wood

1.97

0.88

0.63

–​0.29

69.38

104.82

Textiles

2.08

1.52

0.80

–​1.95

–​30.95

102.25

Manufacturing nec; recycling

2.19

1.28

0.93

–​0.39

75.33

126.79

Non-​metallic mineral, nec

2.29

0.75

0.69

–​0.45

42.83

138.00

Rubber & plastic

2.49

1.33

0.70

–​0.08

152.74

92.76

Basic metals

2.56

1.23

2.30

–​1.04

55.28

116.52

Retail trade; household goods repair

2.80

1.23

7.53

–​0.33

95.14

154.36

Public admin & defence

2.87

1.63

6.79

–​0.34

48.50

207.74

Food

2.95

1.38

2.07

–​0.99

49.91

98.11

Motor vehicle sale, maintenance & repair

3.33

1.62

2.30

–​0.03

108.69

171.39

Electricity, gas & water

3.69

1.51

0.59

–​0.29

107.27

122.05

Machinery, nec

3.70

2.08

1.72

–​0.72

77.47

123.51

Transport eq

3.73

2.27

1.28

–​0.68

115.72

124.31

Other comm, social & personal services

3.83

–​0.08

4.80

1.63

114.38

243.39

Chemicals

4.02

2.32

0.98

–​0.39

246.59

91.71

Transport & storage

5.26

3.14

4.94

0.15

115.98

144.67

Wholesale trade

5.54

2.74

4.92

0.20

130.36

141.04

Paper

5.77

3.08

1.46

–​0.66

103.47

126.87

Electrical eq

5.96

3.12

1.79

–​0.74

736.75

37.40

Business activities

7.34

2.36

11.84

7.45

257.15

226.20

Refined petroleum

7.48

3.65

0.08

–​0.07

62.46

118.63

Financial intermediation

14.47

8.61

3.03

0.27

195.53

167.98

Post & telecommunications

18.96

7.89

1.56

–​0.18

413.60

62.26

Notes: Source: EUKLEMS. Cross-country averages of Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Italy, Netherlands, Sweden, United Kingdom and United States. ICT-capital intensity if missing in 1980 for Ireland and Sweden. Employment is defined in hours worked. Shares are defined at country-year level.

job polarization   181 Column (2a) of Table 8.1 lists sectoral employment shares in 2005 averaged across countries. Employment is measured as hours worked in a sector by all persons engaged (i.e. employees as well as self-​employed workers). The largest sectors are ‘Business activities’ (12%), ‘Health & social work’ (10%), ‘Construction’ (8%) and ‘Retail trade’ (8%), ‘Public administration & defence’ (7%) and ‘Education’ (6%). Column (2b) of Table 8.1 gives percentage-​point changes in employment shares between 1980 and 2005. In line with job polarization, the column shows an increase in some of the least capital-​ intensive sectors (e.g. ‘Hotels & restaurants’, ‘Health & social work’) and in some of the most capital-​intensive sectors (e.g. ‘Business activities’). Columns (3) and (4) in Table 8.1 use EUKLEMS data from the March 2011 release to report the percentage increase between 1980 and 2005 in sectoral output volumes and output prices averaged across the 13 countries in our sample. In EUKLEMS, the output volume series are given as an index with 1995 as the base year (1995 =​100) and are derived from growth rates in chained volume indices reported in countries’ National Accounts. EUKLEMS also obtains nominal output indices from countries’ National Accounts to construct output price indices as the ratio of nominal to volume output series. By and large, column (3) suggests that output has grown faster in more ICT-​capital-​ intensive sectors. For example, the average growth rate in output volume across the 14 sectors with lowest ICT-​capital intensity (out of the 28 sectors used in the empirical analysis that follows) is 61 per cent compared to 149 per cent for the other 14 sectors with the highest ICT-​capital intensity. Also, column (4) suggests that the opposite is true for relative output prices: 145 per cent for the 14 sectors with the lowest ICT-​capital intensity compared to 115 per cent for the 14 sectors with the highest ICT-​capital intensity.

Main Results Job Polarization Table 8.2 reports regression estimates for b1 and b2 from equation (15) using all years in column (1), and by decade in columns (2)–​(4). Starting with column (1a), estimates for b1 and b2 are negative and positive respectively and statistically significant, suggesting that there exists a U-​shaped relationship between changes in sectoral employment shares between 1980 and 2005 and sectors’ ICT-​capital intensity. The bottom of column (1a) shows that, taking the estimates to compute the turning-​point in the U-​shaped relationship at face value, on average 47 per cent of all workers are employed in sectors with ICT-​capital intensities that are lower than the estimated turning-​point. A more conservative estimate of this percentage is to set the turning-​point one standard deviation lower or higher, giving 44 per cent and 48 per cent respectively. In sum, column (1a) is consistent with the predictions from our model that a substantial fraction of unskilled workers is increasingly doing less complex tasks in less

182    maarten goos et al Table 8.2 Changes in employment shares by ICT-​capital intensity, 1980–​2005 Percentage point change in employment share

Variables:

(1)

(2)

(3)

(4)

all years

1980s

1990s

2000s

(a) rank [ICT CAP]

-​0.032*** 0.005 (0.009)

Constant

(b)

(a)

(b)

-​0.022** 0.008**

(0.003) (0.010)

(0.003)

(a)

(b)

(a)

(b)

-​0.032*** 0.006* -​0.052*** -​0.006* (0.012)

(0.004) (0.014)

0.001***

0.001***

0.001***

0.002***

(0.000)

(0.000)

(0.000)

(0.000)

(0.004)

0.118**

-0​ .066* 0.031

-​0.122*** 0.104

-0​ .086* 0.318***

0.087

(0.055)

(0.038) (0.066)

(0.044)

(0.047) (0.103)

(0.061)

(0.077)

Employment share on declining section (in %) at point estimate

47.07

41.30

47.07

57.33

at point estimate -​ st. dev.

43.86

34.27

41.30

53.70

at point estimate +​ st. dev.

48.41

47.07

48.42

61.23

Observations

1820

1820

728

728

728

728

364

364

Country fixed-​effects

-​

-​

-​

-​

-​

-​

-​

-​

Year fixed-​effects

-​

-​

-​

-​

-​

-​

-​

-​

Notes: Source: EUKLEMS. Standard errors clustered at the country-​sector level. Employment is defined as hours worked and employment shares are defined at the country-​year level. Changes are stacked 5-​year changes: 1980-​1985, 1985-​1990, 1990-​1995, 1995-​2000 and 2000-​2005. *** indicates significance at the 1%-​level, ** at the 5%-​level and * at the 10%-​level. The employment share on the declining section is the cross-​country average in 2005.

capital-​intensive sectors where there is less innovation, whereas other skilled workers are increasingly doing more complex tasks in more capital-​intensive sectors where innovation is stronger. As a further robustness check, column (1b) shows that an estimate for b1 when setting b2 = 0 in equation (15) is not statistically significant, suggesting that a linear specification is rejected as a first-​order approximation to the data. Columns (2)–​(4) of Table 8.2 repeat the exercise in column (1) by decade. The fact that all point estimates for b1 and b2 are negative and positive respectively suggests that there has been job polarization in all decades. Moreover, the significant point estimate in column (2b) indicates that relative employment growth in ICT-​capital-​intensive sectors was particularly strong in the 1980s. For the period 2000–​2005, the point estimates that

job polarization   183 are reported in column (4) show that the nature of polarization changed somewhat during this period: the point estimates in column (4a) are statistically significant but column (4b) suggests that the changes in employment shares were particularly large for sectors with the lowest ICT-​capital intensities. Figure 8.1 shows kernel plots of average 5-​year employment share changes by a ranking of sectors according to their ICT-​capital intensity in 2005 for the US and Europe separately. The left panel shows these kernel plots for the US, whereas the right panel does the same using average employment share changes and an ICT-​capital intensity ranking across our sample of 12 European countries. Figure 8.1 again shows that job polarization was present in all decades, but also that there are some differences between decades that are common in both the US and Europe: In the 1980s and 1990s the relationship clearly is U-​shaped with strong growth in ICT-​capital-​intensive sectors in particular. However, by the early 2000s a tilting has taken place, with stronger employment growth in sectors with lower ICT-​ capital intensity. This tilting over time in employment share changes corresponds to existing evidence for the US in Acemoglu and Autor (2011); Mishel et al. (2013); Autor (2015b) and Beaudry et al. (2016), although these papers use a ranking of occupations based on their mean log wage or on worker skills rather than of sectors based on ICT-​ capital intensity. Beaudry, Green and Sand (2016) call this tilting the ‘Great Reversal’. Using a task assignment framework, their explanation relies on a boom-​and-​bust cycle in the dynamic demand for cognitive skills when a new technology is introduced. Intuitively, in an initial boom phase, employment in high-​paid cognitive and low-​paid manual

Average 5y percentage point change in employment shares

United States

Europe

.6 .4 .2 0 –.2 0

10

20

30

0

10

20

30

Rank [ICT CAP] 1980s

1990s

2000s

Figure 8.1  Smoothed average 5-​year employment share changes by ICT-​capital intensity (in percentage points, 1980–​2005) Source: EUKLEMS. Percentage point changes in employment share by average capital intensity ranking in 2005 using a locally weighted smoothing regression (bandwidth 0.8 with 27 observations). Employment expressed in hours worked. European employment shares defined as the average of country-​level shares in 12 European countries

184    maarten goos et al occupations increases, whereas employment in middle-​paid routine jobs and unemployment decreases. However, the marginal productivity and therefore investments in new technologies decrease as capital accumulates, leading to a decrease in the demand for high-​paid cognitive and middle-​paid routine occupations, and an increase in low-​paid manual jobs and unemployment. Beaudry, Green and Sand (2016) provide some indirect evidence for this hypothesis and, if true, predict a continued pattern of skill-​downgrading unless there is a new technological breakthrough that would start a new boom-​and-​bust cycle in the demand for skilled workers.9 Interestingly, the right panel of Figure 8.1 shows that a similar tilting exists across the 12 European countries in our sample.10 Another challenge is to better understand whether there are differences between countries and, if so, why? For example, if anything, Figure 8.1 suggests that job polarization has been somewhat more pronounced in the US than in Europe. There is little doubt that labour market institutions could be important in explaining differences in employment changes between countries. For example, there could be a trade-​off between employment and wage-​setting institutions for low-​skilled workers. However, Oesch (2013) finds little evidence in support of such a trade-​off and the author provides several explanations as to why this could be the case (e.g. the supply of unskilled workers has decreased at least as fast as its demand since countries educated their workforce; technologies could be endogenous such that higher wages for low-​ wage workers also shifts their demand curve out). Fernandez-​Macias (2012) and Fernandez-​Macias et al. (2012) examine cross-​country differences in employment changes. But convincing evidence for institutional drivers that can explain these differences remains notoriously elusive, although it would be interesting to know more about it in light of the ongoing job quality debate also among policymakers. In sum, consistent with the simple framework we have presented, we find that in both the US and Europe there has been job polarization when sectors are ranked by their ICT-​capital intensity. That is, workers are increasingly employed in the least and most ICT-​capital-​intensive sectors in the economy. Moreover, we find that this process of job polarization has characterized our labour markets at least since the 1980s, although some differences seem to exist between decades and countries. Most notably: whereas the 1980s and 1990s were characterized by relatively strong employment growth in the most ICT-​capital-​intensive sectors (especially in the US), the opposite holds after 2000 with stronger employment growth in the least ICT-​capital-​intensive sectors.

9  In line with this, Deming (2017) finds that the share of STEM jobs decreased in the US between 2000 and 2012, after growing in the previous two decades. However, he also finds an increase in STEM jobs that require social skills, attributing it to the specialization of skilled workers in coordinating team production following technological progress. 10  The right-​panel of Figure 8.1 shows a kernel plot of the percentage point changes in employment shares by sectors ranked by their cross-​country average ICT-​capital intensity in 2005. Plotting kernel estimates using the employment share changes after pooling employment across European countries instead gives a very similar picture.

job polarization   185

Baumol’s Cost Disease As discussed in section 2, our framework also predicts a decrease in the relative output price for sectors with higher ICT-​capital intensities. Said differently, the least ICT-​ capital-​intensive sectors will see their relative price increase over time. This prediction is in line with the well-​known phenomenon of Baumol’s cost disease. Baumol’s (1967) original thesis stated that, if productivity growth is unbalanced across sectors, sectors with lower productivity growth will see their relative output price as well as their share in total employment increase.11 The author refers to teaching as an example where labour itself is the output with very little potential for productivity growth: ‘Regardless of new inventions and gains from specialization, there still seems to be fairly firm limits to class size.’ An alternative view, and the one that we take here, is that Baumol’s cost disease does not result from unbalanced productivity growth across sectors but from differences in ICT-​capital intensities between sectors (see also Acemoglu and Guerreri 2008). In particular, the framework we have presented assumed that sectors are more ICT-​capital-​intensive if they are more routine-​task intensive. Consequently, consistent with Baumol’s cost disease, more ICT-​capital-​intensive sectors will innovate more, resulting in lower relative marginal costs and, therefore, lower relative output prices. Also consistent with Baumol’s cost disease is that the least ICT-​capital-​intensive sectors experience both increasing relative prices and relative employment. Table 8.3 shows regression estimates for equation (15) using as the dependent variable the percentage change in the output price for sector j in country c between years t and t −1. In line with predictions from our model, columns (1a) and (1b) of Table 8.3 show that there is a significant negative relationship between the ICT-​capital intensity of sectors and growth in output prices. Columns (2)–​(4) show relatively small differences between decades. Figure 8.2 further illustrates the pervasiveness of these relative price changes in both the US as well as Europe.12 Finally, our framework also predicts that output growth is faster the more ICT-​capital-​ intensive a sector is. Table 8.4 therefore shows estimates for equation (15) using as the dependent variable the percentage change in output volume and shows this is the case indeed. Figure 8.3 further illustrates that these patterns hold in the US as well as in Europe. In sum, there are important technology-​induced changes in relative output prices on product markets in line with Baumol’s cost disease. These changes in relative output 11  Moreover, Baumol (1967) argued that unbalanced productivity growth would lead to an increasing share in GDP for less innovative sectors, and thus to a slowdown in sector-​weighted aggregate growth. Because this is inconsistent with Kaldor facts, more recent work has shown how unbalanced productivity growth can lead to structural changes between sectors while still predicting constant aggregate growth. See, for example, Ngai and Pissarides (2007). 12  The sudden changes in slope at the top end of the ranking may be caused by the less accurate performance of kernel plots at the boundary of the sample range because of the use of a symmetric estimation window. For the most recent decade, the sector ‘Refined petroleum’ experienced high positive growth in the US during the 2000s, likely caused by mechanisms outside our framework, while the most capital-​intensive sector ‘Post & telecommunications’ experienced an absolute decline in prices.

186    maarten goos et al Table 8.3 Changes in output price by ICT-​capital intensity, 1980–​2005 Percentage change in output price

Variables:

rank [ICT CAP]

Constant

(1)

(2)

(3)

(4)

all years

1980s

1990s

2000s

(a)

(b)

(a)

–​0.417

–​0.375*** –​0.585

–​0.353*** –​0.212

–​0.316*** –​0.490

–​0.540***

(0.404)

(0.073)

(0.105)

(0.094)

(0.114)

(0.420)

(b)

(a)

(0.822)

(b)

(a)

(b)

(0.425)

0.001

0.008

–​0.004

–​0.002

(0.013)

(0.014)

(0.027)

(0.015)

42.482*** 42.273*** 37.741*** 36.577*** 16.589*** 17.107*** 12.721*** 12.970*** (3.586)

(0.038)

(3.585)

(3.038)

(4.243)

(2.272)

(2.926)

(2.393)

Observations

1820

1820

728

728

728

728

364

364

Country fixed-​effects

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Year fixed-​effects

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Notes: Source: EUKLEMS. Standard errors clustered at the country-​sector level. Output prices are in value added and indexed, 1995 =​100. Changes are stacked 5-​year changes: 1980-​1985, 1985-​1990, 1990-​1995, 1995-​2000 and 2000-​2005. *** indicates significance at the 1%-​level, ** at the 5%-​level and * at the 10%-​level. The employment share on the declining section is the cross-​country average in 2005.

United States

Europe

Average 5y percentage change in output price

40 20 0

–.20 0

10

20

30

0

10

20

30

Rank [ICT CAP] 1980s

1990s

2000s

Figure 8.2  Smoother average 5-​year output price changes by ICT-​capital intensity (in percentage points, 1980–​2005) Source: EUKLEMS. Average 5-​year percentage changes in output prices by average capital intensity ranking in 2005 using a locally weighted smoothing regression (bandwidth 0.8 with 27 observations). Prices are in value added and indexed, 1995=​100. European output prices defined as the average of country level output prices in the 12 European countries

job polarization   187 Table 8.4 Changes in output volume by ICT-​capital intensity, 1980–​2005 Percentage change in output volume

Variables:

(1)

(2)

(3)

(4)

all years

1980s

1990s

2000s

(a) rank [ICT CAP]

(b)

(a)

(b) 0.559***

(a)

(b)

(a)

-​0.713

1.114***

(b)

-​0.597

0.812***

-​0.369

-​0.819

0.712***

(0.424)

(0.184)

(0.427) (0.126)

(0.708) (0.350)

(0.531)

(0.145)

0.049***

0.032**

0.063*

0.053**

(0.019)

(0.016)

(0.033)

(0.021)

5.483*

-​1.559

4.428

-​0.215

8.125

(2.985)

(4.112)

(3.118)

(3.152)

(5.483) (5.644)

(3.494) (3.483)

Observations

1820

1820

728

728

728

728

364

364

Country fixed-​effects

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Year fixed-​effects

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Constant

-​1.010

5.146

-​2.510

Notes: Source: EUKLEMS. Standard errors clustered at the country-​sector level. Output volumes are in value added and indexed, 1995 =​100. Changes are stacked 5-​year changes: 1980-​1985, 1985-​1990, 1990-​1995, 1995-​2000 and 2000-​2005. *** indicates significance at the 1%-​level, ** at the 5%-​level and * at the 10%-​ level. The employment share on the declining section is the cross-​country average in 2005.

Average 5y percentage change in output volume

United States

Europe

40 30 20 10 0 0

10

20

30

0

10

20

30

Rank [ICT CAP] 1980s

1990s

2000s

Figure 8.3 Smoothed average 5-​year output volume changes by ICT-​capital intensity for manufacturing and all sectors (in percentage points, 1980–​2005) Source: EUKLEMS. Average 5-​year percentage changes in output volume by average capital intensity ranking in 2005 using a locally weighted smoothing regression (bandwidth 0.8 with 27 observations). Output volume is in value added and indexed, 1995=​100. European output volume defined as the average of country level output prices in the 12 European countries

188    maarten goos et al prices entail changes in consumer demand and, therefore, in output volumes. In turn, these changes in relative output volumes on product markets lead to scale-​effects in labour demand. In particular, these scale-​effects mute the direct substitution-​effect of technological progress leading to a reallocation of unskilled workers to sectors that are less capital intensive, and they amplify the reallocation of skilled workers to sectors that are more capital intensive. What our analyses therefore show is that Baumol’s cost disease and job polarization are phenomena that are intrinsically related by the interplay between the nature of technological progress, the direct substitutability between differently skilled workers and new technologies in production, and the price elasticities of product demand for different goods and services in consumption.

Robustness IV Regression Results One concern in our empirical analysis is that the ranking of sectors by their ICT-​capital intensity in 2005 is endogenous. To address this concern, we can use 2SLS estimators to instrument the sector ranking at the beginning of a five-​year period (indicated by the subscript and t − 5) with the ranking at the beginning of a previous five-​year period relative to the period of analysis (indicated by the subscript t −10). The results are shown in Table 8.5. The first-​stage estimates presented in the bottom panel of Table 8.5 confirm the strong correlation of sector rankings over time. In the upper part of the table we report second-​stage estimates for our three outcome variables.13 In sum, our results remain qualitatively unaffected.

Excluding Public Sectors Outcomes in public sectors are not exclusively determined by market forces, as is assumed by our framework. In addition, information about ICT-​capital intensity, output prices and volumes for public goods and services is difficult to measure. Therefore, there might be substantial measurement error in them, such that including them in the analyses gives inconsistent point estimates and/​or incorrect standard errors for these point estimates. Therefore, we re-​estimate our empirical specification leaving out the following sectors: ‘Education’, ‘Health & social work’ and ‘Public administration & defence’. Table 8.6 shows estimates for our three main outcome variables that are qualitatively similar to the estimates shown in previous tables.

Employment De-​Industrialization and Globalization Rodrik (2016) documents a process of employment de-​industrialization in advanced economies: a structural employment shift away from manufacturing industries due to technological progress in those (but not other) sectors and/​or increasing imports of 13 

Note that we lose the first year of data, which we can only use as an instrument.

job polarization   189 Table 8.5 Changes in employment shares, output price and volume by ICT-capital intensity, 1980–2005 2SLS estimates

Variables: rank [ICT CAP] in t-​5

(1) Percentage point change in employment shares

(2) Percentage change in output price

(3) Percentage change in output volume

(a)

(a)

(a)

(b)

(b)

(b)

-​0.039***

0.004

-​0.760

-​0.355***

-​0.878*

0.832***

(0.013)

(0.004)

(0.523)

(0.109)

(0.462)

(0.252)

0.001***

0.014

0.059***

(0.000)

(0.018)

(0.020)

0.162

-​0.059

25.334***

23.276*** 16.292*** 7.615*

(0.113)

(0.083)

(4.043)

(2.496)

(3.534)

(4.387)

Observations

1456

1456

1456

1456

1456

1456

Country fixed-​effects

-​

-​

Yes

Yes

Yes

Yes

Year fixed-​effects

-​

-​

Yes

Yes

Yes

Yes

Constant

First-​stage estimates (1) Quadratic specification (a) rank [ICT CAP] in t-​5 rank [ICT CAP] in t-​10

Constant

(2) Linear specification

(b) rank [ICT CAP] in t-​5

0.837***

1.613

0.823***

(0.070)

(1.909)

(0.020)

-​0.000

0.773***

(0.002)

(0.070)

2.501***

39.032**

2.573***

(0.615)

(16.383)

(0.518)

Observations

1456

1456

1456

Country fixed-​effects

Yes

Yes

Yes

Year fixed-​effects

Yes

Yes

Yes

F-​statistic

1308.31

1423.39

1653.5

F-​statistic p-​value

0.000

0.000

0.000

Notes: Source: EUKLEMS. Standard errors clustered at the country-sector level. Employment is defined as hours worked and employment shares are defined at the country-year level. Output prices and output volumes are in value added and indexed, 1995 = 100. Changes are stacked 5-year changes: 1980-1985, 1985-1990, 1990-1995, 1995-2000 and 2000-2005. *** indicates significance at the 1%-level, ** at the 5%-level and * at the 10%-level.

190    maarten goos et al Table 8.6 Changes in employment shares, output price and volume by ICT-​capital intensity, 1980–​2005 Excluding public sectors

Variables:

rank [ICT CAP]

(1)

(2)

(3)

Percentage point change in employment shares

Percentage change in output price

Percentage change in output volume

(a)

(b)

(a)

(b)

(a)

(b)

-​0.026***

0.009***

-​0.544

-​0.335***

-​0.414

0.847***

(0.010)

(0.003)

(0.475)

(0.081)

(0.478)

(0.193)

0.001***

0.007

0.043**

(0.000)

(0.016)

(0.020)

-​0.024

-​0.208***

42.771***

41.664***

4.020

-​2.681

(0.092)

(0.080)

(4.258)

(2.902)

(3.402)

(4.556)

Observations

1625

1625

1625

1625

1625

1625

Country fixed-​effects

-​

-​

Yes

Yes

Yes

Yes

Year fixed-​effects

-​

-​

Yes

Yes

Yes

Yes

Constant

Notes: Source: EUKLEMS. Standard errors clustered at the country-​sector level. Employment is defined as hours worked and employment shares are defined at the country-​year level. Output prices and output volumes are in value added and indexed, 1995 =​100. Changes are stacked 5-​year changes: 1980-​1985, 1985-​ 1990, 1990-​1995, 1995-​2000 and 2000-​2005. *** indicates significance at the 1%-​level, ** at the 5%-​level and * at the 10%-​level. The employment share on the declining section is the cross-​country average in 2005.

manufactured goods through international trade. If manufacturing sectors would be exclusively situated in the middle of our ICT-​capital intensity ranking, this process of employment de-​industrialization would be observationally equivalent to our framework in that it would also predict job polarization. However, our data suggest that manufacturing sectors are located across a wider range in our ICT-​capital intensity ranking. To see this, Figure 8.4 overlays the predicted change in the employment shares for all sectors in the economy (lines) with the predictions for manufacturing sectors separately (markers). Although predominately situated in the middle, manufacturing sectors are in both the declining as well as inclining halves of the overall U-​shape. What this suggests is that de-​industrialization is less suited to explain economy-​wide job polarization. But Figure 8.4 also clearly shows that the process of employment de-​ industrialization does exist: Predicted employment share changes for manufacturing sectors are almost always negative, especially after 2000. An explanation for this is found in Autor et al. (2013), who study the impact on US employment of increased import competition from China, especially after its accession to the World Trade Organization

job polarization   191

Average 5y percentage point change in employment shares

United States

Europe

.6 .4 .2

+

0

+ ++

–.2 0

+ + + 10

+ +++ + + + 20

+ + ++ 30

0

+++ +

10

+ ++ +

20

30

Rank [ICT CAP]

1980s, all

1980s, all +

2000s, all

1980s, manufacturing

1980s, manufacturing

2000s, manufacturing

Figure 8.4  Smoothed average 5-​year employment share changes by ICT-​capital intensity for manufacturing and all sectors (in percentage points, 1980–​2055) Source: EUKLEMS. Percentage-​point changes in employment share by average capital intensity ranking in 2005 using a locally weighted smoothing regression (bandwidth 0.8 with 27 observations). Employment expressed in hours worked. European employment shares defined as the average of country-​level shares in 12 European countries

in December 2001. They show that increased import competition from China decreases employment in manufacturing relative to services when the US increased its current account deficit with China after 2001.14

ICT-​Capital Intensity of Sectors and Routineness of Occupations A difference between our approach in this chapter and most of the previous literature on job polarization is that we rank jobs by their sector’s ICT-​capital intensity instead of by an occupation’s task content. In particular, Autor et al. (2003) introduced an occupation-​ specific Routine Task Intensity index (RTI-​index) that has been used extensively in the literature on occupational job polarization.

14 

An alternative explanation is given by Grossman and Rossi-​Hansberg (2008) who introduce a task assignment model of offshoring. In particular, they assume that advanced economies are increasingly offshoring unskilled jobs abroad, resulting in a decrease in the relative demand for unskilled workers domestically. However, Grossman and Rossi-​Hansberg (2008) also show that the relative demand for domestic unskilled workers could increase if offshoring enables sectors using unskilled workers intensively to reduce their output prices disproportionately, thereby increasing output and the relative demand for domestic unskilled workers.

192    maarten goos et al The RTI-​index of Autor et al. (2003) corresponds to the three tasks done by workers in our framework: Manual tasks (M), Routine tasks (R), and Abstract tasks (A). For each occupation, these three measures are derived from the US Dictionary of Occupational Titles (DOT). The Manual task measure (M) corresponds to the DOT variable which captures the importance of ‘Eye-​ hand-​ foot coordination’ in an occupation. The Routine task measure (R) is a simple average of two DOT task measures for an occupation: ‘Set limits, tolerances and standards’ and ‘Finger dexterity’, capturing the cognitive and manual aspect of routine tasks respectively. The Abstract task measure (A) is also an average of two DOT measures: ‘Direction control and planning’ and ‘GED math’, capturing the interactive and more analytical aspect of Abstract tasks. The RTI-​index is then computed for each occupation by taking the log of the ratio of the Routine task measure (R) over the sum of Manual (M) and Abstract (A) task measures:  R  RTI = ln   (16)  M + A



with positive values if R > ( M + A) and negative values otherwise. To see whether our framework and empirical results are consistent with studies defining job polarization in terms of occupations rather than sectors, we need to convert the occupation-​specific RTI-​index into a sector-​specific measure that can be compared to its ICT-​capital intensity. To do this, we use occupation-​industry employment shares from the European Union Labour Force Survey (EULFS) for the 12 European countries in our EUKLEMS sample. In particular, we use the EULFS to compute employment shares by 2-​digit ISCO08-​ occupations in 1-​digit NACE(rev.1)-​industries.15 With these data, we can then compute the share of workers employed in routine occupations for each sector, that is, the share of employment in occupations with RTI > 0 within each sector at the 1-​digit level. Occupational job polarization and job polarization in our framework would then be capturing the same phenomenon if the share of workers in routine occupations is lower in the least ICT-​capital and most ICT-​capital-​intensive sectors. Figure 8.5 shows that this is the case indeed: The sector with the largest share of occupational employment in routine tasks is ‘Manufacturing’ with 63 per cent, and second is ‘Wholesale & retail’ with 58 per cent. These sectors have average levels of ICT-​capital intensity. ICT-​capital-​intensive sectors such as ‘Business services’, ‘Transport, storage & communications’ and ‘Financial intermediation’ have shares of employment in routine occupation between 28 per cent and 37 per cent. The smallest shares of employment in routine labour tasks can be found in the in-​person sectors, around 20 per cent or below. 15  ISIC(rev.3)

1-​digit level.

used in EUKLEMS and NACE(rev.1) used in the EULFS are harmonized at the

job polarization   193

.4

n ed i

sto

Fin

an

cia

lI

co d an ge

ra

at io

n tio

m m

es s sin bu d

an g tin

nt er m

un

ac

ls na pe rso

d an

Tra

ns

po

rt,

en ,r te ta

es

ica

tiv i

er vic

go ld ial

tie s

es

s od

y pl eh o oc ity ,s un

al

rc he Ot

Re

m om

ta re d an W

ho

les

ale

us

air ep ;r de

tra il

Pu

bl

Ele

ic

ct

ad

ric

m

in

ity ,g

as

of

an

ho

d

M

up

tu rin

wa te rs

uf ac

an

an

d

uc n ist ra tio

te ls Ho

g

e de fe nc

at io

ts

Ed

an

an He alt h

ur an

re sta

cia d

Co

so

ns

tru c

tio

lw or k

n

0

n

.2

d

Share of routine employment

.6

Figure 8.5  Share of employment in routine occupations over 1-​digit sectors, ranked by their ICT-​capital intensity Source: EUKLEMS (2011) and Goos, Manning and Salomons (2014); EULFS. The United States is left out because it is not included in Goos, Manning and Salomons (2014) and EULFS. Sectors on the x-​axis are ranked by their employment weighted average ICT-​capital intensity. Occupation level RTI is taken from Table 1 of Goos, Manning and Salomons (2014). These are aggregated to 1 digit NACE Rev. 1 sector level by taking the share of employment in 2005 with RTI>0 and averaged across European countries

Conclusion This chapter gave a brief history of the literature on job polarization: the idea that job markets are polarizing into lovely and lousy jobs at the expense of middling jobs is not new and dates back to at least the 1980s. Between the mid-​1970s to mid-​1980s labour markets in advanced economies created many good jobs. However, even at the time, evidence also suggested that a substantial proportion of new jobs were in low-​wage employment, resulting in a job-​quality debate in the 1980s. But starting in the mid-​ 1980s, the job-​quality debate faded among economists as most labour economists came to the conclusion that the problem for unskilled workers was a declining rather than an increasing number of jobs for them because of Skill-​Biased Technological Change (SBTC). The simple SBTC framework has been extremely successful in explaining the evolution of skill-​premia over time in advanced economies. However, most recently, the task approach to labour markets has emerged, leading to a renewed interest in the

194    maarten goos et al phenomenon of job polarization and a new job-​quality debate that is unlikely to disappear in a hurry as it did in the mid-​1980s. To summarize the intuition that economists use to explain job polarization, this chapter then provided a simple intuitive framework. The contribution of this framework is that it illustrates how the interplay of various channels determining outcomes on labour and product markets can predict job polarization. Importantly, this chapter made clear that job polarization is not a natural law that must hold at all times and in all places. This is important because it illustrates that general equilibrium effects in labour and product markets and, possibly, institutions shaping those markets are fundamental in understanding job creation, job destruction and job reallocation, as well as growth and inequality. This chapter then ended with some evidence in support of our framework, further stressing the academic and societal importance of the ongoing job-​quality debate.

Acknowledgements We are grateful to Rachel Dwyer, Thomas Lemieux and seminar participants at the University of Leuven, Utrecht University, University of Groningen, ASSA and ZEW for comments. Salomons gratefully acknowledges funding from the Netherlands Organization for Scientific Research (NWO) under the Veni scheme. Rademakers gratefully acknowledges funding from the Flemish Research Foundation (FWO) under the Aspirant Mandate.

References Acemoglu, D. (1999), ‘Changes in Unemployment and Wage Inequality: An Alternative Theory and Some Evidence’, American Economic Review, 89/​5, 1259–​1278. Acemoglu, D. (2001), ‘Good Jobs versus Bad Jobs’, Journal of Labour Economics’, 19/​1, 1–​21. Acemoglu, D., and Autor, D. (2011), ‘Skills, Tasks and Technologies: Implications for Employment and Earnings’, in D. Card, and O. Ashenfelter, eds., Handbook of Labour Economics. Volume 4 (North-​Holland: Elsevier), 1043–​1171. Acemoglu, D., and Guerrieri, V. (2008), ‘Capital Deepening and Nonbalanced Economic Growth’, Journal of Political Economy, 116/​3, 467–​498. Acemoglu, D., and Restrepo, P. (2018), ‘The Race Between Man and Machine: Implications of Technology for Growth, Factor Shares, and Employment’, American Economic Review, 108/​ 6, 1488–​1542. Autor, D. (2015a). ‘Polanyi’s Paradox and the Shape of Employment Growth’. Jackson Hole: Federal Reserve Bank of Kansas City: Economic Policy Proceedings, Reevaluating Labour Market Dynamics, 129–​177. Autor, D. H. (2015b), ‘Why Are There Still So Many Jobs? The History and Future of Workplace Automation’, Journal of Economic Perspectives, 29/​3, 3–​30.

job polarization   195 Autor, D. H., and Dorn, D. (2013), ‘The Growth of Low-​Skill Service Jobs and the Polarization of the US Labour Market’, American Economic Review, 103/​5,1553–​1597. Autor, D. H., Dorn, D., and Hanson, G. H. (2013), ‘The China Syndrome: Local Labour Market Effects of Import Competition in the United States’, American Economic Review, 103/​6, 2121–​2168. Autor, D. H., Katz, L. F., and Kearney, M. S. (2006), ‘The Polarization of the US Labour Market’, American Economic Review, 96/​2, 189–​194. Autor, D. H., Levy, F., and Murnane, R. J. (2003), The Skill Content of Recent Technological Change: An Empirical Exploration’, Quarterly Journal of Economics, 118/​4, 1279–​1333. Baumol, W. J. (1967), ‘Macroeconomics of Unbalanced Growth: The Anatomy of Urban Crisis’, American Economic Review, 57/​3, 415–​426. Beaudry, P., Green, D. A., and Sand, B. M. (2016), ‘The Great Reversal in the Demand for Skill and Cognitive Tasks’, Journal of Labour Economics, 34/​S1, S199–​S247. Berman, E., Bound, J., and Griliches, Z. (1994), ‘Changes in the Demand for Skilled Labour within U.S. Manufacturing: Evidence from the Annual Survey of Manufacturers’, Quarterly Journal of Economics, 109/​2, 367–​397. Blau, F. D., and Kahn, L. M. (1996), ‘Wage Structure and Gender Earnings Differentials: An International Comparison’, Economica, 63/​250, S29–​S62. Bluestone, B., and Harrison, B. (1988), ‘The Growth of Low-​Wage Employment: 1963-​86’, American Economic Review, 78/​2, 124–​128. Bound, J., and Johnson, G. (1992), ‘Changes in the Structure of Wages in the 1980s: An Evaluation of Alternative Explanations’, American Economic Review, 82/​3, 371–​392. Costrell, R. M. (1990), ‘Methodology in the “Job Quality” Debate’, Industrial Relations: A Journal of Economy and Society, 29/​1, 94–​110. Deming, D. J. (2017), ‘The Growing Importance of Social Skills in the Labour Market’, Quarterly Journal of Economics, 132/​4, 1593–​1640. Farber, H. S. (1997), ‘Job Creation in the United States: Good Jobs or Bad?’ Working Paper 385, Industrial Relations Section, Princeton University. Fernández-​Macías, E. (2012), ‘Job Polarization in Europe? Changes in the Employment Structure and Job Quality, 1995-​2007’, Work and Occupations, 39/​2, 157–​182. Fernández-​ Macias, E., Hurley, J., and Storrie D. (eds.) (2012), Transformation of the Employment Structure in the EU and USA, 1995-​2007. (London: Palgrave Macmillan). Gittleman, M. B., and Howell, D. R. (1995), ‘Changes in the Structure and Quality of Jobs in the United States: Effects by Race and Gender, 1973-​1990’, ILR Review, 48/​3, 420–​440. Goldin, C. D., and Katz, L. F. (2009), The Race Between Education and Technology (Harvard University Press, Cambridge, MA). Goos, M. (2018), ‘The Impact of Technological Progress on Labour Markets: Policy Challenges’, Oxford Review of Economic Policy, 34/​3, 362–​375. Goos, M., and Manning, A. (2007), ‘Lousy and Lovely Jobs: The Rising Polarization of Work in Britain’, Review of Economics and Statistics, 89/​1, 118–​133. Goos, M., Manning, A., and Salomons, A. (2009), ‘Job Polarization in Europe’, American Economic Review, 99/​2, 58–​63. Goos, M., Manning, A., and Salomons, A. (2014), ‘Explaining Job Polarization: Routine-​Biased Technological Change and Offshoring’, American Economic Review, 104/​8, 2509–​26. Griliches, Z. (1969), ‘Capital-​Skill Complementarity’, Review of Economics and Statistics, 51/​4, 465–​468.

196    maarten goos et al Grossman, G. M., and Rossi-​Hansberg, E. (2008), ‘Trading Tasks: A Simple Theory of Offshoring’, American Economic Review, 98/​5, 1978–​1997. Howell, D. R., and Wolff, E. N. (1991), ‘Trends in the Growth and Distribution of Skills in the US Workplace, 1960-​1985, ILR Review, 44/​3, 486–​502. Ilg, R. E. (1996), ‘The Nature of Employment Growth, 1989-​95’, Monthly Lab. Rev., 119, 29. Ilg, R. E., and Haugen, S. E. (2000), ‘Earnings and Employment Trends in the 1990s’, Monthly Lab. Rev., 123, 21. Juhn, C. (1999), ‘Wage Inequality and Demand for Skill: Evidence from Five Decades’, ILR Review, 52/​3, 424–​443. Juhn, C., Murphy, K. M., and Pierce, B. (1993), ‘Wage Inequality and the Rise in Returns to Skill’, Journal of Political Economy, 101/​3, 410–​442. Katz, L. F., and Murphy, K. M. (1992), ‘Changes in Relative Wages, 1963-​1987: Supply and Demand Factors’, Quarterly Journal of Economics, 107/​1, 35–​78. Koh, H., and Magee, C. L. (2006), A Functional Approach for Studying Technological Progress: Application to Information Technology’, Technological Forecasting and Social Change, 73/​9, 1061–​1083. Kosters, M. H., and Ross, M. N. (1987), ‘The Distribution of Earnings and Employment Opportunities: A Re-​ Examination of the Evidence’, American Enterprise Institute Occasional Paper No. 87. Washington, DC: American Enterprise Institute, September 1987. Kosters, M. H., and Ross, M. N. (1988), ‘A Shrinking Middle Class?’ Public Interest, 3–​27. Leuven, E., Oosterbeek, H., and Van Ophem, H. (2004), ‘Explaining International Differences in Male Skill Wage Differentials by Differences in Demand and Supply of Skill’, Economic Journal, 114/​495, 466–​486. Levy, F., and Murnane, R. (1992), ‘US Earnings Levels and Earnings Inequality: A Review of Recent Trends and Proposed Explanations’, Journal of Economic Literature, 30/​3, 1333–​81. Mishel, L., Shierholz, H., and Schmitt, J. (2013), ‘Don’t Blame the Robots: Assessing the Job Polarization Explanation and Growing Wage Inequality’, Technical report, Center for Economic Policy and Research. Murphy, K. M., and Welch, F. (1993), ‘Occupational Change and the Demand for Skill, 1940-​ 1990’, American Economic Review, 83/​2, 122–​126. Nagy, B., Farmer, J. D., Bui, Q. M., and Trancik, J. E. (2013), ‘Statistical Basis for Predicting Technological Progress’, PLOS ONE, 8/​2, 1–​7. Ngai, L. R., and Pissarides, C. A. (2007), ‘Structural Change in a Multisector Model of Growth’, American Economic Review, 97/​1, 429–​443. Nordhaus, W. D. (2007), ‘Two Centuries of Productivity Growth in Computing’, Journal of Economic History, 67/​01, 128–​159. Oesch, D. (2013), Occupational Change in Europe: How Technology and Education Transform the Job Structure (New York: Oxford University Press, New York). Rodrik, D. (2016), ‘Premature Deindustrialization’, Journal of Economic Growth, 21, 1–​33. Timmer, M. P., O’Mahony, M., and Ark, B. v. (2007), ‘EU KLEMS Growth and Productivity Accounts: An Overview’, International Productivity Monitor, 14, 71–​85. US Council of Economic Advisors (1996), ‘Job Creation and Employment Opportunities: The United States Labour Market, 1993-​1996’. Report. Washington, D.C: US Council of Economic Advisors,. Wright, E. O., and Dwyer, R. E. (2003), ‘The Patterns of Job Expansions in the USA: A Comparison of the 1960s and 1990s’, Socio-​Economic Review, 1/​3, 289–​325.

job polarization   197

Appendix A Model A.1 Equilibrium Since all markets are assumed to be perfectly competitive, the second welfare theorem implies that the equilibrium is characterized by solving the social planner’s problem of maximizing utility of the representative household: σ

θ

σ −1 σ −1 θ −1 θ −1   σ −1   θ −1 max U = Cs σ + C σ  with C ≡ C g 1θ + C g 2θ  (A1) K1 , K 2 , Lr , La 1    

where:



Cs = Ys = Lm = 1 − Lr (A2)



C gj = Ygj − Ykj with Ykj = pk K j for j = 1, 2 (A3)



Yg 1 = L1a−1β1 X β1 with X ≡  Lµr + K1µ  µ (A4)



1

1 − β2

Yg 2 = L1a−2β2 K β2 = 1 − La1 

K β2 (A5)

It is further assumed that the relative price per efficiency unit of computer capital limits to zero asymptotically: pk → 0 as t → ∞ ; 0 < µ < 1 implying that the elasticity of substitution between routine labour and capital tasks is larger than unity, i.e. 1 / [1 − µ ] > 1 ; 1 > β2 > β1 > 0 implying that Sector 2 is more routine-​task intensive (and more capital intensive) than Sector 1; 0 < σ < θ and θ > 1 implying that consumer demand for Services will be relatively price inelastic and that consumer demand for Sector 1 and 2 goods will be price elastic; La1 ∈[0,1] are abstract labour tasks in Sector 1, La 2 = 1 − La1 ∈[0,1] are abstract labour tasks in Sector 2, Lm ∈[0,1] are manual labour tasks in Services, and Lr = 1 − Lm ∈[0,1] are routine labour tasks in Sector 1; skilled workers supply their labour inelastically to abstract labour tasks in Sector 1 or Sector 2 and unskilled workers supply their labour inelastically to manual labour tasks in Services or to routine labour tasks in Sector 1.

Efficient allocations of K1 , K 2 , Lr and La1 are given by the first-​order conditions:

pk = β1L1a−1β1 X β1 − µ K1µ −1 (A6) pk = β2 1 − La1  / K 2  L1r− µ [1 − Lr ]

−

1 σ

=C

σ −θ θσ

−

1

1 − β2

(A7)

C g 1θ β1L1a−1β1 X β1 − µ (A8)

198    maarten goos et al

1

L−a1β1 1 − La1  2 = C g 1 / C g 2  θ [1 − β2 ]/[1 − β1  X −β1 K 2β2 (A9) β

Note that equations (A6) –​(A9) also give the equilibrium allocations over time for pk → ∞ because it is implicitly assumed that income can be saved but not for future consumption and that the capital stock can increase every period—​that is, there is capital accumulation—​but capital fully depreciates between periods. The remainder of this section shows the conditions under which these efficient allocations are consistent with between-​sector job polarization.

A.2 Job Polarization Job polarization requires that unskilled workers move to less capital-​intensive sectors, that is, from Sector 1 to Services, and skilled workers move to more capital-​intensive sectors, that is, from Sector 1 to Sector 2. Asymptotically, we must have that Lr → 0 and that La1 → 0 which is now shown to be the case under some additional parameter restrictions. Unskilled Workers Reallocate from Sector 1 to Services Since the relative price of capital falls to zero in each sector, capital in each sector will accumulate and limit to infinity: K1 → ∞ and K 2 → ∞ if pk → 0. Because K1 → ∞ and Lr is bounded, X → ∞ . Consequently, X + K 2 → ∞ or the number of routine tasks done in the economy will limit to infinity. Due to technological progress embodied in (computer) capital, there is routinization through capital accumulation in the economy.16 Given that capital accumulates in Sectors 1 and 2, equation (A8) determines the conditions needed for Lr → 0 . To see this, first consider the consumption terms C g1 and C g 2 (which are included in C ) that occur in the expression. Making use of equations (A6) and (A7) gives:



C g 1 = Yg 1 − pk K1 = 1 − β1[K1 / X ]µ  L1a−1β1 X β1 (A10) C g 2 = Yg 2 − p k K 2 = [1 − β2 ]1 − La1 

1 − β2

K 2β2 (A11)

which shows that the limits for terms C g1 and C g 2 depend on the limit for La1 . If La1 → 0 , we get that the limit for C g1 is undetermined but that C g 2 → ∞ . If La1 → 1 , we get that the limit for C g1 → ∞ and that the limit for C g 2 remain unspecified. However, note that in any case C → ∞ because either for C g1 → ∞ or C g 2 → ∞ . Substituting equation (A10) into equation (A8) gives:



L1r− µ [1 − Lr ]

−

1 σ

=C

σ −θ θσ

−

1

[θ −1][1− β1 ]

µ θ 1 − β1[K1 / X ]  β1La1

θ

X

β1 − µ −

β1 θ

(A12)

Given that µ < 1 , we must have that Lr → 0 on the left-​hand side of equation (A12) if the right-​ hand side of equation (A12) is zero asymptotically: (i) C → ∞ implies that C

σ −θ θσ

→ 0 if  σ < θ ;

16  Moreover, given that the number of manual and abstract tasks are bounded by L and L m a respectively, X + K 2 → ∞ implies that the share of routine tasks done in the economy converges to unity.

job polarization   199 1

1

(ii) K1 / X → 1 implies that 1 − β1[K1 / X ]µ  θ converges to converges to [1 − β1 ] θ ; (iii) θ −1 1 − β / θ La1 ∈[0,1] implies that L[a1 ][ 1 ] converges to unity given that θ > 1 and β1 < 1 ; (iv) Given that X → ∞ , a sufficient condition for the right-​hand side of equation (A12) to converge to zero and therefore Lr → 0 then is: −

−

Lr → 0 if [1 / θ ] > [β1 − µ ] / β1 (A13)



Equation (A13) has a straightforward interpretation. For given β1 , Lr → 0 if µ is sufficiently large relative to θ and therefore also relative to σ (given that σ < θ ). Said differently, the elasticity of substitution between routine labour tasks and capital in production, 1[1 − µ ] , must be sufficiently large compared to the elasticities of substitution in consumption, θ and σ . On the one hand, a large elasticity of substitution between routine labour tasks and capital in production implies a large decrease in the demand for routine relative to manual labour tasks—​this is a substitution effect. On the other hand, the decrease in the capital price could also lead to a fall in the output price of Sector 1 goods relative to Services, thereby increasing the demand for routine relative to manual labour tasks—​this is a scale effect. But if θ and σ are small relative to1 / [1 − µ ] , this scale effect will not dominate the substitution effect and Lr → 0 . The role played by β1 in equation (A13) also has a straightforward interpretation. All else equal, a small β1 implies that the inequality in equation (A13) is more likely to hold or that it is more likely that unskilled workers will reallocate from Sector 1 to Services. The reason for this is that a small β1 implies a small constant expenditure share on routine tasks in Sector 1. Consequently, following a decline in the price of capital, a possible increase in spending on unskilled workers in Sector 1 is small, which would tend to amplify the reallocation of unskilled workers from Sector 1 to Services.

Skilled Workers Reallocate from Sector 1 to Sector 2 Because Sector 2 is more routine task intensive than Sector 1 given that β2 > β1 , a uniform decline in the price of capital across both sectors leads to a greater adoption of routine tasks in Sector 2 than in Sector 1. So, we not only have that routine tasks accumulate in both sectors, that is, X → ∞ and K 2 → ∞ , but also that routine tasks accumulate faster in Sector 2 than in Sector 1, that is, X / K 2 → 0 . Given that X / K 2 → 0 , equation (A9) determines the asymptotic allocation of skilled labour. Substituting equations (A10) and (A11) into equation (A9) gives:



β1 +

La1

1− β1 θ

β2 − 1 − β2 θ

1 − La1 

= [1 − β1 ][β2 − 1]

θ −1 θ

1 − β1[K1 / X ]µ 

−

1 θ

 X β1 / K 2β2 

θ −1 θ

(A14)

showing that we must have that La1 → 0 on the left-​hand side of equation (A14) because the right-​hand side of equation (A14) is zero asymptotically: (i) K1 / X → 1 such that 1 − β1[K1 / X ]µ  β1 β2  X / K 2 

θ −1 θ

−

1 θ

1

converges to [1 − β1 ] θ ; (ii) X / K 2 → 0 because β2 > β1 implies that

→ 0 if θ > 1 .

−

200    maarten goos et al That skilled workers reallocate from Sector 1 to Sector 2 if θ > 1 also has a straightforward interpretation. For given relative output prices, β2 > β1 implies that a decline in the price of capital has a smaller positive impact on labour demand in Sector 2 than in Sector 1. To understand this, note that a large β2 implies a large constant capital share and, thus, a small constant labour share in Sector 2. Consequently, the possible increase in spending on skilled workers in Sector 2 is limited following a decline in the price of capital. This effect would tend to work against the reallocation of skilled workers from Sector 1 to Sector 2. However, because a fall in the price of capital also leads to a decrease in the output price of Sector 2 relative to Sector 1 goods (also see equation (A16)), the demand for abstract labour tasks increases in Sector 2 relative to Sector 1. Because the substitutability in consumption between Sector 1 and 2 goods is larger than unity, that is, θ > 1 , this increase in the demand for abstract labour tasks in Sector 2 relative to Sector 1 dominates such that La1 → 0 . More generally, skilled workers doing abstract labour tasks reallocate towards the more capital-​intensive sectors in the economy. This is because these sectors will see their relative output price decrease, thereby increasing their demand for skilled abstract labour tasks the more substitutable their goods are in consumption.

A.3 Changes in Relative Output Prices and Volumes To derive an expression for the price of Services, ps , relative to Sector 2 goods, p2 , note that utility maximization implies:  θ−σ 1  1 ps / p2 = C θσ C gθ2  / Csσ → ∞ (A15)  



Consider the different terms on the right-​hand side of equation (A15): (i) Given that (i)  C → ∞ implies that C 1 θ g2

θ−σ θσ

→ ∞ given that σ < θ ; (ii) equation (A11) shows that C g 2 → ∞ if La1 → 0

such that C → ∞ ; (iii) Given that Cs → 1 as Lr → 0 , equation (A15) limits to infinity. Similarly, we can derive the following condition for the price of Sector 1 relative to Sector 2 goods:



1

p1 / p2 = C g 2 / C g 1  θ → ∞ (A16)

where the limit follows because C g 2 / C g 1 → ∞ . To see this, substitute equations (A10) and (A11) into C g 2 / C g 1 : C g 2 / C g 1 = [1 − β2 ] / 1 − β1[K1 / X ]µ  1 − La1 

1 − β2

/ L1a−1β1 K 2β2 / X β1 → ∞ (A17) which limits to

infinity if La1 → 0 given that K1 / X → 1 , K 2 / X → ∞ and 1 > β2 > β1 . Finally note that equation (A17) predicts an increase in output volume that is larger for Sector 2 than for Sector 1. Moreover, C g 2 → ∞ and Cs → 1 implies that C g 2 / Cs → ∞ .

job polarization   201

Appendix B Data 8B.1 ICT-​Capital Intensity Rank Stability Over Time

Table 8B.1 Rank stability of ICT-​capital intensity across sectors over time Average Spearman rank correlation coefficient Austria

0.91

Belgium

0.89

Denmark

0.75

Spain

0.94

Finland

0.89

France

0.93

Germany

0.95

Ireland

0.76

Italy

0.87

Netherlands

0.90

Sweden

0.95

United Kingdom

0.79

United States

0.91

Notes: Each number is the average of the pairwise year-​by-​ year Spearman rank correlation coefficients of the ICT-​capital intensity across sectors in a country.

8B.2 ICT- Capital Intensity Rank Stability across Countries

Table 8B.2 Rank stability of ICT-​capital intensity across sectors and countries in 2005 Spearman rank correlation coefficient Austria

0.83

Belgium

0.79

Denmark

0.87

Spain

0.83

Finland

0.85

France

0.79

Germany

0.85

Ireland

0.73

Italy

0.71

Netherlands

0.78

Sweden

0.81

United Kingdom

0.65

United States

0.79

Notes: Each number is the Spearman rank correlation coefficient between a country’s ICT-​ capital intensity and the average ICT-​capital intensity of all other countries in 2005.

Table 8B.3 shows the decomposition of capital assets in the EUKLEMS database. The components indicated in bold are part of the definition of ICT-​capital.

Table 8B.3 Decomposition of capital assets Total tangible assets (A)

Total construction Residential structure Total non-​residential investments

(B)

Machinery and equipment Transport equipment Machinery and other equipment computing equipment Communications equipment other machinery and equipment Other tangible assets Total intangible assets

Software Other intangibles Notes: Source: Timmer et al. (2007). Assets that fall under the definition of ICT-​capital are denoted in bold.

chapter 9

geographies of job quality sally weller, tom barnes and nicholas kimberley

Introduction In recent times, the OECD has renewed its focus on local development and the importance of place. The OECD’s approach to job quality, as a balance between job demands and job resources, focuses on job dimensions that are, in practice, place-​specific (Cazes et al. 2015). National data on job quality, such as the data reported by the OECD, aggregates the outcomes of these place-​based processes. This makes understanding the geography of jobs crucial to an adequate conceptualization of job quality. For geographers, job quality is one aspect of the wider social relations of labour exchange. With the inherently spatial nature of all social relations at the forefront of geographical analyses, job quality is approached relationally and understood as a product of contested labour market processes that materialize unevenly across different types of economic and socio-​political settlements. These processes generate patterns of job quality that vary between industries, occupations and social groups (Herod et al. 2007). For geographers, therefore, job quality is a product created at the intersection of labour supply, labour demand and frameworks of regulation in specific placed contexts. As a consequence of this holistic orientation, it is rare for geographical work to tackle job quality as a stand-​alone category. Rather, geographers understand job quality in the context of the multiple factors shaping the lives of working people (Castree 2007; Ward 2007; Weller and Campbell 2015). Geographical inquiry examines how contemporary reorganizations of work, employment, and labour markets shift market risks onto workers and weaken the association between work and worker livelihoods. There is a strong interest in the changing landscapes of employment relations and labour regulation. Geographical research also spans beyond workplace issues to examine the ways individual workers reconcile paid work with family obligations, the relationship between labour markets and housing

204    sally weller, tom barnes and nicholas kimberley markets and between labour markets and migration, the differentiated impact of industry reorganizations on particular groups of workers, and the effects of plant closures and deindustrialization on individuals, households and communities. Consequently, although geographers share with other disciplines a normative or ethical commitment to improving job quality, there is no sub-​disciplinary geography dedicated to the topic. In fact, the only mention of ‘job quality’ in the geographically oriented Handbook of Employment and Society: Working Space (Zook and Samers 2010 in McGrath-​Champ et al. 2010) retains scare quotes around the term to mark the authors’ discomfort with separating the qualities of jobs from the characteristics of their incumbents. This chapter examines job quality as a topic situated within geographical studies of labour’s position in the shifting landscapes of capitalism. It commences by fleshing out geography’s approach to labour studies where process, place, relationality and scale are central interests. It then discusses five areas of scholarship in which job quality is a prominent dimension: regional polarization, global production networks, global care networks, platform and gig economies, and labour agency. The conclusion underlines the distinctiveness of geography’s approach.

Geographical Understandings of Labour Markets The central tenet—​and point of difference—​in geographical inquiry is that space is not just the arena in which social processes play out but, rather, is constitutive of all social praxis. Uneven spatial patterns are a part of the explanation of social relations as well as being their outcome (Massey 1984). Building on Harvey’s (1978, 1982) insight that capitalism’s dynamism stems from and relies on spatially uneven distributions of resources, geographies of work and employment are attuned to the ways that differently spatialized and scaled landscapes of socio-​economic interactions conjure local employment profiles and divisions of labour. Understanding what happens in work, employment and labour relations therefore requires consideration of the inter-​relationships among discourses, institutions, structures and different forms of agency. It also requires sensitivity to political contestation. Like all relationships and processes, labour market struggles are political struggles in which actors and groups of actors seek to advance their projects. Politics involves spatial practices that reshape arenas of action, alter territorial boundaries, reframe the way people make sense of the world, and rescale perceptions of both existing conditions and the realms of future possibility. Geographical studies contextualize labour, labour market processes, and labour agency in the dynamic geographies of economic, political, social and environmental change. As a consequence, labour issues are inseparable from ‘the wider question of how people live and seek to live’ (Castree 2007: 859). From this socialized understanding, geographers begin from the assumption that labour can never be bought and sold in the

geographies of job quality    205 same way as everyday commodities such as tomatoes or mobile phones. Working people must earn sufficient wages to sustain life, provide for their families, and reproduce the labour force. This means that labour is at best a pseudo-​commodity and that labour markets can never be true markets; at least, not in the neoclassical economic sense of an unfettered and aspatial interaction of labour supply and labour demand equilibrated by a market price for ‘human capital’ and skill. Rather, for geographers, labour markets are the product of the multiple, inherently spatialized structures and processes that work to allocate labour to work tasks (Castree et al. 2004). Since labour markets cannot function without external regulation to guarantee labour’s reproduction, governments will always have a role in securing job quality. Geography’s insistence that labour market processes are both inherently spatialized and structured geographically conditions conceptualizations of labour markets (Peck 1996, 2003; Martin 2000). In the twentieth century, prior to the revolutions of globalization and internet connectivity, labour markets were viewed as essentially local phenomena created by the interaction of labour supply with labour demand in the context of a regulatory framework (Peck 1996). The idea of regulation here extends beyond the laws governing labour relations to include all the rules and norms of what Jonas (1996: 325) termed ‘local labour control regimes’. These are ‘historically contingent and territorially embedded sets of mechanisms that coordinate the reciprocities between production, work, consumption and labour reproduction within a local labour market’ and combine all the regulations pertaining to labour market operation, including tacit norms, local social arrangements, and associated ‘ways of doing’ interaction. The everyday expectations of localized work cultures are created by the history of those rules and the experiences they produce, which means that processes of labour market formation—​of which issues of job quality are a part—​are inextricably embedded in the challenges of daily living in places (Walker 1985). Local markets are influenced inter alia by the contested strategies and tactics of competing firms, the interventions of trade unions and industry lobby groups, the policies of governments, and the interventions of a range of intermediaries (Storper and Walker 1983). With globalization, technological change and increasing mobility, however, the geographical scope of labour markets has expanded to create multiple occupationally specific labour markets with varying territorial reach (Weller 2008). At the same time, labour markets are embedded in and shaped by other types of institutions and wider economic processes. Local areas operate within regional and national jurisdictional and regulatory boundaries, and local firms often follow the strategies of distant corporate head offices. These links mean that place-​based processes represent local variations of regional and national arrangements (Jessop 1990). As suggested by the ‘varieties of capitalism’ literature (Hall and Soskice 2001), the extent of national and regional differences in formal regulatory environments are highly significant for any discussion of differences in job quality. National and regional differences in labour regulation reflect different ideological positions on the question of how best to secure prosperity (and for whom). All this means that the localized functioning of labour markets is crosscut by processes that operate across territories and scales.

206    sally weller, tom barnes and nicholas kimberley In specific places, therefore, the quality of jobs is governed by direct labour regulation (such as minimum wages, work hours and rights to collective bargaining); by rules governing the conditions of entry into and exit from the labour market (school leaving age, gender regimes, visa rules for immigrant workers, education and training policies, dismissal, redundancy and retirement policies); and indirectly by related laws (such as competition laws, taxation arrangements and social security eligibilities). This institutional perspective produces explanations attuned to the dynamic and unpredictable nature of competitive processes as they re-​adjust constantly to this plethora of influences. This orientation produces a preference for extended and comparative case study research that spans and connects places and scales and is sensitive to causal mechanisms that might be invisible or incidental to empirical analysis.

Placing and Scaling Job Quality Geography’s wider contextualization of labour issues produces three principal contributions to discussions of job quality. First, geographers view all social processes, including the social relations that produce different sorts of jobs, as inherently spatialized. At the most obvious level, an appreciation of spatiality highlights that the characteristics of workers, jobs and labour markets vary from place to place. However, the notion of inherent spatiality goes further by implying that all the social processes contributing to job design and job quality are positioned somewhere, in a relational spatial arrangement. Thus, for Castree (2007: 855, after Herod 2001), ‘[G]‌eography is more than simply “background scenery” . . . [i]nstead, it has a constitutive role to play in the drama of what happens to workers and what workers can do to alter the terms and conditions of their employment’. Thus, for example, the dearth of job creation in regional towns is juxtaposed to the concentration of jobs in adjacent major cities, job opportunities to the positioning of workers and their households relative to transport corridors, and job creation to the locational preferences of globalized networks of firms. Changes in the regulatory context and vagaries of the business cycle have uneven spatial effects as threats to and opportunities for firm profitability translate into firm re-​ locations and workplace re-​organizations. A spatialized perspective highlights and explains the increasing geographical concentrations of both ‘good’ and ‘bad’ quality jobs. The social world of work is also ‘scaled’ in the sense that the spatial arrangements of labour markets are ordered by local, regional, national and global rules, norms and imperatives. The staccato rhythms of labour market regulation and re-​regulation at different scales create tensions that on the one hand provide opportunities for workers’ struggles (Herod 2001), but on the other hand combine to strengthen the coercive function of regulatory states and to open up new spaces for the expansion of poor quality jobs. Second, as well as being ‘placed’, geographers view work as also unavoidably ‘embodied’, performed by real people with real lives who have responsibility for and

geographies of job quality    207 obligations to families and communities. It follows that job quality must be understood as a dynamic social phenomenon embedded in, shaped by, and governed through, multiple social processes that interact dynamically, albeit in often contradictory ways. The evolving character of jobs is produced as employers compete to attract skilled labour and create new jobs, or restructure existing jobs, in ways that take advantage of the characteristics of available and under-​utilized segments of the (potential) workforce. High quality jobs—​in terms of remuneration, working conditions and career paths (but perhaps not hours and work–​life balance)—​are created in high-​skill, cutting-​edge occupations where skills are both highly valued and scarce. The poorest quality jobs are created specifically for vulnerable or desperate population segments, such as undocumented agricultural workers in the US and refugees in Europe (Mitchell 2006; Lewis et al. 2015). Other jobs are designed to target specific groups, such as low-​paid suburban retail and office jobs that fit to the schedules of women with school-​age children (Hanson and Pratt 1995). This process has created millions of jobs that do not provide sufficient wages to support a household. The fragmentation of work assumes households contain multiple wage earners or are supported by other sources of income (such as welfare transfers, rents, dividends and private pension payments). This understanding suggests that the quality of jobs cannot be separated from the personal characteristics and social positioning of their incumbents. It follows that typologies based on job characteristics alone, without regard to these relational and contextual dimensions, cannot produce an adequate definition of job quality. Third is geography’s insistence that job quality, as with all matters pertaining to labour markets, is a political issue. Geographical analyses highlight that job characteristics are the product of political struggles (Herod 1997; Castree 2000). The extent of labour regulation by governments reflects the different spatial and temporal resolutions of ongoing contestation. It produces uneven job quality in different industries, places or social groups. In places where the interests of employers prevail, politically motivated re-​regulation actively compromises job quality and expands labour market inequalities. Whenever firms seek to operate with fewer workers and with more flexible labour arrangements, job quality comes under threat (Gough 2003). Geographical work rejects outright economists’ expectation that market forces will produce inter-​place convergence in the long run. On the contrary, market processes increase differentiation and inequality because they concentrate poor-​ quality jobs among vulnerable groups of workers in peripheral places and good-​quality jobs among privileged groups in world cities.

Sites of Geographical Job Quality Research Five areas in which job quality has been integral to geographical analysis are the polarization of regional employment, the dynamics of labour in global production networks

208    sally weller, tom barnes and nicholas kimberley and in global care chains, platform and gig economies, and labour’s capacity to reshape job quality. At each of these sites, technological and social change translates into employer strategies that, while reproducing diversity in job quality, also increasingly produce precarious lives and polarize the fortunes of places.

Regional Polarization of Job Quality The price and productivity of labour has always varied from place to place, and capital has always exploited these differences. However, firms now have greater capacity than ever to compare locations and to act on those comparisons (Moretti 2012). Existing employers reconfigure the available work incessantly to optimize their operations and extract the maximum rate of profit: for example, by reducing staff numbers, reducing working hours, using labour hire workers or turning to zero-​hours contract arrangements. Changes in the geography of jobs create opportunities to reconfigure work arrangements: when existing firms elect to relocate their operations, as established but uncompetitive firms fail, and as new firms set up in optimal locations. In places where there are more qualified workers than there are jobs, competitive pressures threaten job quality. Over-​supplied labour markets create ‘queues’ for access to quality work. These are ordered by employers’ discriminatory preferences to create starkly contrasting opportunities for different subgroups of workers. McDowell (2011), for example, shows how financial services firms order queues by class, gender and ethnicity. Contemporary technologies of connectivity—​including the internet and mobile phones—​are another factor promoting polarization by encouraging firms to locate in well-​connected places. Connectivity depends on, and is made possible by, physical networks of satellites and cables. This infrastructure has material effects for the distribution of jobs of different types as firms avoid places with poor internet connections and cluster at nodes of fibre-​optic connectivity (Graham and Marvin 2001). Relatedly, firms’ increasing use of Geographical Information Systems (GIS) to track consumers and products is reshaping the geography of business location, and therefore the distribution of employment globally (Graham and Wood 2003). Geographers are interested in how spatial arrangements—​such as the distance between places, the topology of the landscape or the physical design of a workplace—​impact employment arrangements and job quality. Within workplaces, for example, the arrangement of workstations conditions work intensity and interpersonal interactions. State strategies are another important contributor to diverging regional fortunes. Governments have responded to increasing firm mobility by adopting competitive, entrepreneurial policies and market-​friendly regulatory approaches. Regardless of whether these policy settings are motivated by ideology or practicality, they reshape labour markets and labour market participation patterns in ways that increase spatially based divergences. In the competition to attract business, governments in less advantaged regions are prone to adopt the low-​road tactic of disciplining labour by, for example reducing real wages and employment conditions relative to competing

geographies of job quality    209 locations. In contrast, governments in wealthy urban growth centres are able to invest in enhancing amenities to attract a ‘creative class’ of knowledge workers (Florida 2002). Regulatory innovations intended to attract capital often function in practice to undermine job quality. In ‘workfare’ states, draconian welfare conditions combine with migration rules to produce vulnerable workers who have no choice but to accept poorquality jobs (Peck 2001; Peck and Theodore 2001; Bauder 2006; Smith et al. 2008). Geographical research often traces how social groups such as migrant workers, students, sole parents or the unemployed are channelled toward a narrow range of insecure low-​wage jobs. In places where governments allow subsistence welfare payments to be augmented by some earned income, policy actively creates a workforce available for intermittent, part-​ time and low-​wage jobs. Recursively, this triggers employer responses that create the types of low-​quality jobs tailored to this labour supply (see Peck and Theodore 2001; see also Rubery and Wilkinson 1994). These polarizing forces compromise job quality in different ways depending on the place, the occupation and the characteristics of the workforce. Places have always developed reputations based on the characteristics of their labour forces but global connectivities increase the importance of this external gaze. As discursive constructions, place reputations have brand-​like qualities that do not necessarily accord with the factual evidence. They might be positive, as in a place known for the skill, ingenuity or reliability of its workforce; or negative, as in a place renowned for conflict, industrial unrest or poor product quality. Place reputation influences the likelihood that capital will invest in new job-​creating initiatives (Kavaratzis and Ashworth 2005; Hall and Hubbard 1996). Consequently, in the absence of redistributive interventions by governments, poorer rates of job creation in places with adverse reputations compound local disadvantage. In addition, governments that promote an upbeat and business-​ friendly entrepreneurial image might simultaneously close down public debates about job quality. These forces combine to distribute good and bad jobs unevenly between nations and among regions within nations. Polarization produces the derogatory ‘rustbelt’ descriptor of deindustrializing cities and the optimistic promise of ‘sunbelt’ locations. A vibrant stream of empirical research in geography examines regional ecologies of employment (Martin 1997; Rey and Janikas 2005; Fingleton et al. 2012). Aggregated data reveal the intergenerational persistence of spatial divisions such as the North–​South divide in the UK and sunbelt–​rustbelt divide in the US. Areas with high unemployment, poorer labour market participation rates and abundant poorer quality jobs are often juxtaposed to areas of low unemployment, high labour market participation and plentiful high-​quality jobs. Only in major metropoles do large numbers of workers with well-​paid, high-​quality jobs generate work for and coexist with vulnerable and underpaid personal services industry workers (Sassen 1991). Overall, these polarizing forces compromise job quality in different ways depending on the place, the occupation and the characteristics of the workforce. The processes that increase inter-​regional differences simultaneously differentiate job quality. With nation states in market economies less likely to support redistributive transfers into declining

210    sally weller, tom barnes and nicholas kimberley regions, the places bypassed by capital have become sites of entrenched disadvantage. In places where the political promises of capitalist development are no longer credible, disadvantage is often translating into disenchantment and political disengagement.

Job Quality and Globalized Production The structural imperatives of globalizing capitalism further differentiate job quality. A vast literature documents the rise of distanciated, networked and interlinked forms of organization—​variously described as Global Commodity Chains (GCC), Global Value Chains (GVC) or Global Production Networks (GPN)—​in which inter-​firm relationships have a profound impact on labour, labour relations and job quality. Geographers have tended to favour the GPN approach, rather than the GVC framework, because it recognizes both tiered hierarchies and lateral network interactions and is more able than GVC and GCC to incorporate local specificities and inter-​locational differences (Coe and Yeung 2015). GPNs operate across the globe, in densely tiered networks of firms. Different aspects of the production sequence, from raw materials to finished products, are optimally located relative to local factor prices, amenable regulatory frameworks, proximity to inputs and outputs, and accessibility to transport corridors. The design, management and integration of these components are organized in head offices, which are typically located in core global cities adjacent to centres of financial and political power. Labour is central to the network form: for Smith et al. (2002), production networks are ultimately networks of embodied labour that embroil people and communities in global geographies of work and employment but which at the same time are shaped by the spatial contexts within which people live. Global production networks threaten job quality in some places and enhance it in others. First, as these networks developed and matured in the 1980s, firms shifted many routine and less-​skilled production tasks to places where labour was less expensive, where worker collectives such as unions were less well developed, and where authoritarian governments induced workers to be regarded as more compliant. This changed the global distribution of ‘good’ and ‘bad’ jobs. For manufacturing, secure, well-​paid permanent positions in advanced economies transformed into often insecure, poorly paid and precarious positions in emerging economies whose governments’ export-​oriented industrialization policies hastened the shift. Second, once networks were established, firms were more able to compare production costs in different locations. There were two consequences: firms detached from a sense of obligation to local workforces, and governments were thrown into competition with each other as they struggled to attract investment. Both opened the way for reduced local job quality. Third, the hierarchical axis of the structure of global production networks is designed to extract surplus value and deliver profits ‘up’ to leading firms, via competition within the chain. As pressure to compete on prices increases further ‘down’ the pyramid, job quality deteriorates (Rainnie et al. 2013; Starosta 2010). The poorest quality jobs are found at network peripheries, where firms scavenge for short-​term overflow

geographies of job quality    211 and supplementary contracts. Fourth, global networks harness and accelerate the dissemination of technologies that standardize and harmonize business routines, software packages and accounting practices. This development adds to the polarization of the global workforce by dissolving routine administrative jobs, making more tasks relocatable, and concentrating highly skilled strategic, system design and regulatory tasks at core sites. Fifth, in high-​wage locations, firms use the threat of sending work offshore as a disciplining device, a weapon wielded to disable collective bargaining and wage demands. Globally, the corporate strategies of major core firms are designed to favour short-​term returns on shareholders’ investments and shift business risks to workers, with the effect of producing more volatile labour markets across all networks (Christopherson 2002). Sixth, to complete the circle, inexpensive consumer goods created in low-​wage locations enable workers in advanced economies to maintain living standards for less cost, which helps them to manage household provisioning despite reduced job opportunity and diminished job quality.

Global Care Chains In 2009, during the Global Financial Crisis, the World Bank’s report Economic Geography recognized and legitimated the global relocation of workers from the global South to advanced economies through permanent and temporary migration (World Bank 2009). By then, migration was already firms’ preferred means for coordinating geographically uneven patterns of labour supply and labour demand (World Bank 2006; see also Bauder 2006). The new strategy would be to bring people to jobs rather than jobs to people. Permanent and temporary migration policies are designed to channel workers from the global South into occupations experiencing skill shortages in advanced economies—​ supplying skilled professionals such as doctors, nurses, accountants and IT specialists—​but in practice they also create workforce segments available for poorquality service jobs. In particular, these networks deliver workers, mostly women, to work in less-​skilled caring occupations, such as aged-​care attendants and hospitality-​industry workers (Perrons 2004). Global care chains are created by entrepreneurial actors negotiating the spaces created by migration rules, welfare eligibilities and labour deregulation. These chains function to allocate care and domestic work in advanced economies to women who hail, predominantly, from Asia and Latin America. Care chains highlight the embodied nature of work and the global socio-​spatial reach of labour market processes. Migrant women care workers are among the most exploited labour in advanced economies. McDowell and colleagues (McDowell 2008; McDowell et al. 2007; Dyer et al. 2011; Wills et al. 2009) show how the intersection of migration, citizenship and gender creates a space that enables the institutionalization of the poor-​ quality jobs these women occupy at particular sites. The embodied character of care work, the construction of worker identities, workplace practices, and interpersonal

212    sally weller, tom barnes and nicholas kimberley interactions work to produce and reproduce the labour market vulnerabilities of these workers. Here, the extraction of value is more concentrated and more personalized than in global production networks. Care workers often work in less unionized sectors, where workers use a variety of ‘flexible tactics’ (Datta et al. 2007) to cope with the unstable and poor-​quality nature of their employment (see also Smith and Stenning 2006). Through remittances and electronic connections to home, these workers create trans-​nationalized lives that transcend the local scales of daily existence (Kelly and Lusis 2006). The rise of global care chains, and their functional dependence on informalized workspaces, also serves to highlight the geographies of the ‘alternative economic spaces’ that exist in the gaps not filled by the formal economy (Leyshon et al. 2003). In a world containing far more potential workers than job opportunities, increasing numbers of people are engaged in paid work without having ‘a job’ in the formal contractual sense. The quality of this work is not only an issue for those who work at the margins of modern slavery (Strauss 2012; Kara 2009). In advanced economies, large numbers of people contribute as volunteers and interns, or in cooperatives, the illicit economy, family enterprises, and arrangements where work is undertaken to meet social or community obligations. This work, despite being unpaid or paid in-​kind or conducted on a barter basis, has become so professionalized that it is barely distinguishable from paid work. The quality of these non-​jobs is an issue when moral obligations make it hard for workers to walk away from unsatisfactory working arrangements.

Technological Change and the Emergence of Platform and Gig Economies Technological change has a major impact on employment profiles, divisions of labour, reskilling and deskilling processes and, therefore, on job quality. As technology alters the profile of the demand for labour with particular skills and qualities, the bundling of tasks into occupations changes, creating new specializations and revising views about the types of worker best suited to particular tasks (Sayer and Walker 1992). These redefinitions evolve over time, for example, as privileged occupational groups claim interesting and challenging new tasks and as routine tasks are automated. Thus, although the transformation of occupational skill trajectories is an indeterminate process, power differentials imply the reproduction and intensification of skill-​based labour market polarizations. As technologies enable tasks to be relocated both geographically and institutionally, restructuring processes facilitate a reshaping of the relationships between labour, management and government actors. As with globalization, in dynamic labour markets technological change creates and destroys both good and bad jobs. Geographers tend to emphasize the complexity, contingency and ‘temporally provisional and spatially uneven’ nature of technology-​induced changes (Zook and Samers 2010: 128).

geographies of job quality    213 Contemporary technologies are reinscribing and deepening the division between what used to be called the primary and secondary labour markets (see Doeringer and Poire 1985). For many full-​time workers in the secure primary (or formal) sector, technologies of connectivity have intensified work and compromised work–​life balance. They create the expectation that workers will be available for work continually, allow work to intrude into non-​work time, blur the distinction between at-​work and at-​home identities, and socially proscribe unprofessional behaviour in private time. Workplace surveillance technologies now enable employers to monitor work performance more closely and to apply sophisticated metrics to compare workers’ outputs. For workers, these intrusions undermine the quality of life as well as the quality of work. Software technologies that standardize and routinize office work and harmonize it with international business routines deskill workers and increase the monotony of office work (Thrift 2005). Standardization enables firms to outsource accounting and related services to firms that compete by offering less attractive employment conditions to their workforces. For workers in the insecure secondary sector, on the other hand, work is often precarious and insecure. This trend has been fuelled in part by the emergence of labour market intermediaries, such as labour-​hire firms, that enable firms to access labour on a temporary or casual basis (Peck and Theodore 2001), and in part by computer programs that enable firms to tailor employment profiles to the spatial and temporal rhythms of demand for different skills. Recently, the technology-​enabled emergence of the ‘platform’ or ‘gig’ economy has increased firms’ and other consumers of services’ access to a flexible supply of labour. Platforms and labour-​hire arrangements both change the spatial organization of labour practices. Platforms use internet applications (apps) to link providers of tasks and users of tasks. Platforms take many forms, some exchanging goods and others exchanging services. The platform-​based coordination of economic transactions relies on algorithms that in practice function to localize physical tasks (e.g. lawn-​mowing) and internationalize both the exchange of goods and the performance of knowledge-​based tasks (e.g. copy-​editing). Platforms facilitate individualized, arms-​length, short-​term interactions between employers as consumers and workers as sellers of services. Platform-​based businesses interact with individual workers as contractors rather than as employees, a strategy that avoids on-​costs (like sick pay allowances) and shifts the costs of training and equipment to workers. Competition for work then pushes down the price of labour. The platform firms charge a percentage fee for each sale, so their incentive is to fragment work into multiple short interactions. As a result, workers are catapulted into a world of uncertainty. Platforms have enabled labour subcontracting agencies to flourish in multiple sectors. In this sense, firms in the platform economy commonly act as a specific form of labour-​market intermediary, by linking jobseekers to income-​earning opportunities, as well as acting as arms-​length employers. These arrangements accelerate the proliferation of short-​term, task-​specific contract-​based engagements and serve to further undermine job quality (Warhurst et al. 2017). When third-​party firms contract to use

214    sally weller, tom barnes and nicholas kimberley agency labour, the putative employer is external to the workplace. This tends to compromise worker voice in negotiations over work tasks and to stifle collective forms of worker organization. Platforms are also amenable to use in the allocation of work to creative workers of various types. Creative work throws out challenges to understandings of job quality because such workers see their work as a vocation and work long hours, often in the capacity of independent contractors, for remuneration that does not reflect the time or effort expended honing their craft (Gill and Pratt 2008). Yet creative workers have autonomy and artistic freedom, which makes their jobs ‘good’ despite their less good characteristics. Many artistic workers live in cosmopolitan cities but work from home and struggle at subsistence levels. For them, however, working from home is a choice made in the light of economic realities but from a position of relative power. Space allocations often distinguish artists (able to dedicate a space at home to their work) from workers pushed into home-​based work as a means of diminishing their wages, working rights and working conditions. The diverse modalities of platform and gig economies highlight the importance of applying a relational understanding of ‘work’, ‘workers’ and ‘regulation’ to assessments of job quality.

Worker Agency The right to representation is a foundational dimension of job quality (Castree 2000; Herod 2001). Geographies of labour focus on worker agency, mostly in collective organizational forms such as in unions, and with interest directed to its trans-​locational and multiscalar dimensions. Agency can be defined as the various ways that workers act, both collectively and individually, within and beyond traditional union movements, to improve the quality of their jobs and the quality of their working lives. This work seeks to see employment arrangements ‘through the eyes of labour’ to understand how ‘workers . . . make space in particular ways’ (Herod 1997: 3). These geographies recognize the irresolvable tension between worker strategies to defend working conditions in particular places—​tactics that often lead workers to cooperate with local capital or pit workers in one place against workers in other places (Herod 1997; Ince et al. 2015)—​and strategies that build collaborative links among workers in different places. Of particular interest to geographers is the rescaling of struggles and the creation of new scales of organizing through the involvement of national and international trade unions and labour rights groups with local unions in places (Cumbers et al. 2008; Lier 2007; Herod 1997; Savage 2006; Wills 1998). Applying these ideas to the analysis of GPNs reveals how the ‘just-​in-​time’ structure of globalized production increases the bargaining power of workers at the choke points of global production chains, such as dockworkers (Castree 2000; Sadler and Fagan 2004; Selwyn 2012). Platform technologies have changed the geographies of worker resistance by enabling unions and other advocacy groups to harness community-​based strategies and build extended international linkages. They enable individuals to connect to unions in

geographies of job quality    215 new ways, provide the platforms for social movement politics, and make spatially extensive organizing models possible (Castree 2000). At the same time, there has been considerable debate over whether new forms of collective agency have shifted the loci of actions to defend and improve job quality from the traditional workplace to worker–​community collaborations (Silver 2003). These debates are increasingly important given that traditional forms of workplace-​based organizing are prohibited in many jurisdictions and compromised by subcontracting, agency and casual work. Geographers are interested in global activist networks and the role of NGOs, such as anti-​sweatshop campaigners, especially the ways they link consumer activists in advanced economies with the struggles of production workers in distant places (Johns and Vural 2000; Silvey 2004). Much of this work examines the scales of regulation—​including civic, state and global institutions—​required to generate and enforce labour standards. Worker agency can also be found at the workplace scale in acts of resisting, circumventing, sabotaging or reworking particular employer strategies (Datta et al. 2007) or discerned in local ‘ways of doing’ work practices that limit the range of strategic options available to both employers and employees (Hudson 2001; Warren 2014). The decline in the quality of jobs in advanced economies is a direct consequence of the decline of both unionism and of the political power of working people.

Conclusion: Precarious Working Lives The five research areas we have summarized share a commitment to documenting and analysing the different ways that contemporary capitalism is undermining job quality and working lives. For geography, the qualities of jobs and characteristics of job holders are two sides of the same coin. This understanding is typically expressed as an indissoluble coupling of precarious work and precarious workers. It follows that creative and ‘gig’ economy work tailored to the preferences of people who have secure livelihoods is not precarious. Geography’s interweaving of precarious work and precarious lives emphasizes the specificities of spatio-​temporal contexts to produce a conditional and contingent understanding of job quality centred on lived experience (Ettlinger 2009; after Butler 2006). From this viewpoint, assessments of any job’s quality depend on how people think and feel and interact, not only on the material characteristics of their working conditions. Precarious circumstances span work and other domains because precarious lives are on the receiving end of uneven power relations. Since the precise modalities of power vary with the context, research focuses on identifying the processes that engender precarity and understanding their spatial and temporal dimensions in different social, economic, political and cultural contexts. Precarity is characterized not only by work and life insecurity but now also by political insecurity as abandoned workers realize that traditional labour-​oriented political parties have failed to protect their interests. This produces ‘untidy’ geographies of job quality that reject the artificial

216    sally weller, tom barnes and nicholas kimberley separation of objective conditions of work from accompanying perceptions and emotional responses (Ettlinger 2004; see also Waite 2009). To sum up: the distinctiveness of geographical approaches to work and employment is the insistence that all labour market processes are inherently spatial and unavoidably placed. Geographical research in labour studies aims to ‘connect(s) work and the reproductive sphere, class and non-​class identities, local affairs and global forces’ (Castree 2010: 468). This approach insists that working people have the capacity to actively produce their working lives and that they are not simply the passive victims of firm and state restructuring strategies (Smith et al. 2002). Job quality is the product of social processes that construct the relationship between jobs and their incumbents in systematic ways in different locations. These processes generate diverging and uneven trajectories between nations and between regions within nations.

References Bauder, H. (2006), Labor Movement: How Migration Regulates Labor Markets (Oxford: Oxford University Press). Butler, J. (2006), Precarious Life: The Powers of Mourning and Violence (London: Verso). Castree, N. (2000), ‘Geographic Scale and Grass-​Roots Internationalism: The Liverpool Dock Dispute, 1995–​1998’, Economic Geography, 76/​3, 272–​292. Castree, N. (2007), Labour Geography: A Work in Progress’, International Journal of Urban and Regional Research, 31/​4, 853–​862. Castree, N. (2010), ‘Workers, Economies, Geographies’, in S. McGrath-​Champ, A. Rainnie and A. Herod, eds., Handbook of Employment and Society: Working Space (London: Edward Elgar), 457–​476. Castree, N., Coe, N., Ward, K., and Samers, M. (2004), Spaces of Work: Global Capitalism and Geographies of Labour (London: Sage). Cazes, S., Hijzen, A., and Saint-​Martin, A. (2015), ‘Measuring and Assessing Job Quality: The OECD Job Quality Framework’, OECD Social Employment and Migration Working Papers No. 174. OECD Publishing: Paris, http://​dx.doi.org/​10.1787/​5jrp0​2kjw​1mr-​en (last accessed 19 October 2017). Christopherson, S. (2002), ‘Why Do National Labor Market Practices Continue to Diverge in the Global Economy? The ‘Missing Link’ of Investment Rules’, Economic Geography, 78/​ 1, 1–​20. Coe, N. M. and H. W. Yeung (2015), Global Production Networks: Theorising Economic Development in an Inter-​connected World (Oxford: Oxford University Press). Cumbers, A., Nativel, C., and Routledge, P. (2008), ‘Labour Agency and Union Positionalities in Global Production Networks’, Journal of Economic Geography, 8/​3, 369–​387. Datta, K., McIlwaine, C., Evans, Y., Herbert, J., May, J., and Wills, J. (2007), ‘From Coping Strategies to Tactics: London’s Low-​Pay Economy and Migrant Labour’, British Journal of Industrial Relations, 45/​2, 404–​432. Doeringer, P., and Poire, M. (1985), Internal Labor Markets and Manpower Analysis (New York: M E Sharp). Dyer, S., McDowell, L., and Batnitzky, A. (2011), ‘Migrant Work, Precarious Work-​Life Balance: What the Experiences of Migrant Workers in the Service Sector in Greater London

geographies of job quality    217 Tell us about the Adult Worker Model’, Gender, Place and Culture: A Journal of Feminist Geography, 18/​5, 685–​700. Ettlinger, N. (2004), ‘Toward a Critical Theory of Untidy Geographies: The Spatiality of Emotions in Consumption and Production’, Feminist Economics, 10/​3, 21–​54. Ettlinger, N. (2009), Precarity Unbound, Alternatives, 32/​3, 319–​340. Fingleton, B., Garretsen, H., and Martin, R. (2012), ‘Recessionary Shocks and Regional Employment: Evidence on the Resilience of UK Regions’, Journal of Regional Science, 52/​1, 109–​133. Florida, R. (2002), ‘The Economic Geography of Talent’, Annals of the Association of American Geographers, 92/​4, 743–​755. Gill, R., and Pratt, A. (2008), ‘In the Social Factory? Immaterial Labour, Precariousness and Cultural Work’, Theory, Culture & Society, 25/​7–​8, 1–​30. Gough, J. (2003), Work, Locality and the Rhythms of Capital (London: Continuum). Graham S., and Marvin, S. (2001), Splintering Urbanism: Networked Infrastructures, Technological Mobilities and the Urban Condition (London: Routledge). Graham, S., and Wood, D. (2003), ‘Digitizing Surveillance: Categorization, Space, Inequality’, Critical Social Policy, 23/​2, 227–​248. Hall, P., and Soskice, D. (2001), Varieties of Capitalism: The Institutional Foundations of Comparative Advantage (Oxford: Oxford University Press). Hall, T., and Hubbard, P. (1996), ‘The Entrepreneurial City: New Urban Politics, New Urban Geographies?’ Progress in Human Geography, 20/​2, 153–​174. Hanson, S., and Pratt, G. (1995), Gender, Work and Space (London: Routledge). Harvey, D. (1978), ‘The Urban Process under Capitalism: A Framework for Analysis’, International Journal of Urban and Regional Research, 2/​1–​4, 101–​131. Harvey, D. (1982), The Limits to Capital (London: Blackwell). Herod, A. (1997), ‘From a Geography of Labor to a Labor Geography: Labor’s Spatial Fix and the Geography of Capitalism’, Antipode, 29/​1, 1–​31. Herod, A. (2001), Labor Geographies: Workers and the Landscapes of Capitalism (New York: The Guilford Press). Herod, A., Rainnie, A., and McGrath-​Champ, S. (2007), ‘Working Space: Why Incorporating the Geographical is Central to Theorizing Work and Employment Practices’, Work Employment and Society, 21, 247. Hudson, R. (2001), Producing Places (New York: The Guilford Press). Ince, A., Featherstone, D., Cumbers, A., MacKinnon, D., and Strauss, K. (2015), ‘British Jobs for British Workers? Negotiating Work, Nation, and Globalisation through the Lindsey Oil Refinery Disputes’, Antipode, 47/​1, 139–​157. Jessop, B. (1990), State Theory: Putting the Capitalist State in its Place (University Park PA: Penn State Press). Johns, R., and Vural, L. (2000), ‘Class, Geography, and the Consumerist Turn: UNITE and the Stop Sweatshops Campaign’, Environment and Planning A, 32/​7, 1193–​1213. Jonas, A. (1996), Local Labour Control Regimes: Uneven Development and the Social Regulation of Production’, Regional Studies, 30/​4, 323–​338. Kara, S. (2009) Sex Trafficking: Inside the Business of Modern Slavery (New York: Columbia University Press). Kavaratzis, M., and Ashworth, G. (2005), ‘City Branding: An Effective Assertion of Identity or a Transitory Marketing Trick? Tijdschrift voor economische en sociale geografie, 96/​5, 506–​514. Kelly, P., and Lusis, T. (2006), ‘Migration and the Transnational Habitus: Evidence from Canada and the Philippines’, Environment and Planning A, 38/​5, 831–​847.

218    sally weller, tom barnes and nicholas kimberley Lewis, H., Dwyer, P., Hodkinson, S., and Waite, L. (2015), ‘Hyper-​Precarious Lives: Migrants, Work and Forced Labour in the Global North’, Progress in Human Geography, 39/​5, 580–​600. Leyshon, A., Lee, R., and Williams, C. (eds.) (2003) Alternative Economic Spaces (London: Sage). Lier, D. (2007), ‘Places of Work, Scales of Organising: A Review of Labour Geography’, Geography Compass, 1/​4, 814–​833. Martin, R. (1997), ‘Regional Unemployment Disparities and Their Dynamics’, Regional Studies, 31/​3, 237–​252. Martin, R. (2000), ‘Local Labour Markets: Their Nature, Performance, and Regulation’, in G. Clark, M. Feldman, and M. Gertler, eds., The Oxford Handbook of Economic Geography (Oxford: Oxford University Press), 445–​476. Massey, D. (1984), Spatial Divisions of Labour (London: Macmillan). McDowell, L. (2008), ‘Thinking through Work: Complex Inequalities, Constructions of Difference and Transnational Migrants’, Progress in Human Geography, 32/​4, 491–​507. McDowell, L. (2011), Capital Culture: Gender at Work in the City (London: John Wiley & Sons). McDowell, L., Batnitzky, A., and Dyer, S. (2007), ‘Division, Segmentation and Interpellation: The Embodied Labours of Migrant Workers in a Greater London Hotel’, Economic Geography, 83/​1, 1–​25. Mitchell, D. (1996), The Lie of the Land: Migrant Workers and the California Landscape (Minnesota: University of Minnesota Press). Moretti, E. (2012), The New Geography of Jobs (New York: Houghton Mifflin Harcourt). Peck, J. (1996), Work-​Place: The Social Regulation of Labor Markets (New York: Guilford Press). Peck, J. (2001), Workfare States (New York: Guilford Press). Peck, J. (2003), ‘Places of Work’, in E. Sheppard and T. Barnes, eds., A Companion to Economic Geography (London: Blackwell), 133–​148. Peck, J., and Theodore, N. (2001), ‘Contingent Chicago: Restructuring the Spaces of Temporary Labor’, International Journal of Urban and Regional Research 25/​3, 471–​496. Perrons, D. (2004), Globalization and Social Change: People and Places in a Divided World (London: Routledge). Rainnie, A., Herod, A., and McGrath-​Champ, S. (2013), ‘Global Production Networks, Labour and Small Firms’, Capital & Class, 37/​2, 177–​195. Rey, S. and Janikas, M. (2005), ‘Regional Convergence, Inequality, and Space’, Journal of Economic Geography, 5/​2, 155–​176. Rubery, J., and Wilkinson, F. (1994), ‘Introduction’, in J. Rubery and F. Wilkinson, eds., Employer Strategy and the Labour Market (Oxford: Oxford University Press), 1–​17. Sadler, D., and Fagan, R. (2004), ‘Australian Trade Unions and the Politics of Scale: Reconstructing the Spatiality of Industrial Relations’, Economic Geography, 80/​ 1, 23–​43. Sassen, S. (1991), The Global City (Princeton, NJ.: Princeton University Press). Savage, L. (2006), ‘Justice for Janitors: Scales of Organizing and Representing Workers’, Antipode, 38/​3, 645–​666. Sayer, A., and Walker, R. (1992), The New Social Economy: Reworking the Division of Labour (Cambridge MA: Blackwell). Selwyn B. (2012), ‘Beyond Firm-​Centrism: Re-​Integrating Labour and Capitalism into Global Commodity Chain Analysis’, Journal of Economic Geography, 12/​1, 205–​226. Silver, B. (2003), Forces of Labor: Workers Movements and Globalisation since 1870 (Cambridge: Cambridge University Press). Silvey, R. (2004), ‘A Wrench in the Global Works: Anti‐Sweatshop Activism on Campus’, Antipode, 36/​2, 191–​197.

geographies of job quality    219 Smith, A., Rainnie, A., Dunford, M., Hardy, J., Hudson, R., and Sadler, D. (2002), ‘Networks of Value, Commodities and Regions: Reworking Divisions of Labour in Macro-​Regional Economies’, Progress in Human Geography, 26/​1, 41–​63. Smith, A., and Stenning, A. (2006), ‘Beyond Household Economies: Articulations and Spaces of Economic Practice in Post-​Socialism’, Progress in Human Geography, 30/​2, 190–​213. Smith, A., Stenning, A., Rochovská, A., and Świa̧tek, D. (2008), ‘The Emergence of a Working Poor: Labour Markets, Neoliberalisation and Diverse Economies in Post-​Socialist Cities’, Antipode, 40/​2, 283–​311. Starosta, G. (2010), ‘The Outsourcing of Manufacturing and the Rise of Giant Global Contractors: A Marxian Approach to Some Recent Transformations of Global Value Chains’, New Political Economy, 15/​4, 543–​463. Storper, M., and Walker, R. (1983), ‘The Theory of Labour and the Theory of Location’, International Journal of Urban and Regional Research, 7/​1, 1–​43. Strauss, K. (2012), ‘Coerced, Forced and Unfree Labour: Geographies of Exploitation in Contemporary Labour Markets’, Geography Compass, 6/​3, 137–​148. Thrift, N. (2005), Knowing Capitalism (Thousand Oaks: Sage). Waite, L. (2009), ‘A Place and Space for a Critical Geography of Precarity?’ Geography Compass, 3/​1, 412–​433. Walker, R. (1985), ‘Class, Division of Labour and Employment in Space’, in D. Gregory and J. Urry, eds., Social Relations and Spatial Structures (London: Macmillan), 164–​89. Ward, K. (2007), ‘Thinking Geographically about Work, Employment and Society’, Work, Employment and Society, 21/​2, 265–​276. Warhurst, C. Mathieu, C., and Wright, S. (2017), ‘Workplace Innovation and the Quality of Working Life in an Age of Uberisation’, in P. Oeij, D. Rus, and F. Pot, eds., Workplace Innovation: Theory, Research and Practice (Cham: Springer), 245–​259. Warren, A. (2014), ‘Working Culture: The Agency and Employment Experiences of Nonunionized Workers in the Surfboard Industry’, Environment and Planning A, 46, 2300–​2316. Weller, S. (2008), ‘Are Labour Markets Necessarily ‘Local’? Spatiality, Segmentation and Scale’, Urban Studies, 45/​11, 2203–​2223. Weller, S., and Campbell, I. (2015), ‘The Contextualised Geographies of Job Quality’, in A. Knox and C. Warhurst, eds., Job Quality in Australia (Sydney: Federation Press), 84–​102. Wills, J. (1998), ‘Taking on the CosmoCorps? Experiments in Transnational Labor Organization’, Economic Geography, 74/​2, 111–​130. Wills, J., Datta, K., Evans, Y., Herbert, J., May, J., and McIlwaine, C. (2009), Global Cities at Work: New Migrant Divisions of Labour (London: Pluto Press). World Bank (2006), World Development Report 2006: Equity and Development (New York: World Bank). World Bank (2009), World Development Report 2009: Economic Geography (New York: World Bank). Zook, M., and Samers, M. (2010), ‘Telemediated Servants and Self-​Servants of the Global Economy: Labour in the Era of ITC-​ Enabled e-​ Commerce’, in S. McGrath-​ Champ, A. Rainnie and A. Herod, eds., Handbook of Employment and Society: Working Space (London: Edward Elgar), 124–​143.

chapter 10

the cornerstone of job quality Occupational safety and health maria albin, chris mathieu, esa-​pekka takala and töres theorell

Introduction Most domains of job quality pertain to things beneficial to employees—​wages, training and education, discretion and voice, job security, control over scheduling, promotion and progression opportunities, etc. However, one domain, Occupational Safety and Health (OSH), literally is about life and death—​the quality of one’s physical and psychic life, and abrupt or premature death. Envisioning job quality as a continuum spanning from benefit to harm, OSH expressly seeks to prevent harm and injury occurring from work activities and the work environment. OSH is a cornerstone of job quality in two senses. First, illness, injury, and death are the most serious negative consequences of work from an employee perspective in absolute terms, as health and well-​being are paramount. Second, if instead of conceptualizing job quality as a horizontal continuum from harm to benefit one conceptualizes it vertically, as a type of Maslow’s pyramid, OSH toils at the bottom seeking the prevention of somatic and psychosocial harm, making continued, undiminished work possible at higher levels. Physical and psychic health are prerequisites for being able to hold, carry out, and develop in a job. Thus, the prerequisites for labouring and employment are largely secured by OSH. Employees’ inherent interest in their well-​being, and employers’ and states’ interest in labour capacity (and employee/​citizen well-​being) leads to a general confluence of interests in protecting workers from harm. However, parties are not always in agreement about what practical steps to take to secure safety and health, as there can be overt and

the cornerstone of job quality     221 productivity costs to OSH activities, as well as savings in financial and health terms. But because OSH is fundamental to both employee well-​being and productivity, it is arguably the domain in the employment field where explicit regulation is both most comprehensive and legitimate to all social partners. Regulation takes both hard and soft approaches, producing obligatory standards and monitoring compliance regarding the former, and producing guidelines, advice, and educational materials regarding the latter. Both are premised on monitoring developments by gathering statistical data on OSH-​related phenomena (accidents, deaths, illness, sick-​days, etc.) and prognosticating future developments and the need for new or altered approaches to OSH work. These activities take place at national and supra-​national levels (EU-​OSHA, ISO, ILO). Setting recommendations, standards and especially exposure limits is however contentious, as costs in terms of lives and health are weighed against economic costs. Even with current toxicological and epidemiological data, agencies still set exposure limit values exceeding what some researchers find acceptable risk, thereby contradicting basic human rights to a safe work environment (Johanson and Tinnerberg 2019). OSH can be seen from two distinct, but not mutually exclusive perspectives—​ economics and human suffering. The human suffering perspective extends beyond the individual employee directly affected as injury, impairment, and death, also impact family, relatives, colleagues, and friends via affective empathy, and have economic consequences for a wider circle of persons around the afflicted individual. Occupational illness can also affect biological reproductive functions, generating direct intergenerational implications (see section entitled ‘Adverse Reproductive Effects’). Therefore, the extent of impact of occupational injury must be understood as extending far beyond individuals and the individual numbers found in reported statistics and estimates. The extent of occupational injury is difficult to precisely ascertain due to under-​ reporting, even in developed economies. A recent ILO-​commissioned study estimating occupational injury at the global level found that 2.78 million deaths in 2014, or 5 per cent of all deaths globally, have work-​related causes. Work-​related illnesses and diseases accounted for 86.3 per cent of this mortality, and accidents accounted for 13.7 per cent (Hämäläinen et al. 2017: 4). In addition to direct mortality, another 374 million work-​ related injuries or illnesses (leading to at least four days absence from work) were estimated for 2014 (Hämäläinen et al. 2017: 12). Beyond human suffering, occupational injury’s economic costs are tremendous, including insurance and liability costs and reduced productivity. In purely economic terms at the global level, fatal and non-​fatal work-​related accidents and illness are estimated to cost 3.9 per cent of GDP, while the figure for Europe is 3.3 per cent (EU-​ OSHA 2017: 2). Several categorical distinctions are made in OSH work. First, injury or harm can arise in two basic ways, through accidents, which is primarily associated with safety, or through recurrent exposure to hazards in occupational activities, which is primarily

222    maria albin, chris mathieu, esa-pekka takala and töres theorell associated with occupational health. A second distinction is between fatal and non-​ fatal injury; both accidents and exposure to workplace hazards can lead to death (immediate, often in the case of accidents, or premature due to occupational illness) and injury or illness. Regarding the latter, one can distinguish between temporary and long-​ term impairment. Finally, illnesses are often categorized in terms of the general type of hazards causing injury. Distinctions are made between biological and chemical causes, psychosocial causes, repetitive physical strain, etc. Other important causes not further described in this chapter are violence and hostility at work from managers, colleagues, clients, etc. (Burke and Cooper 2018), noise (hearing impairment), vibrating tools (neurological and circulatory impairment in the hand), and thermal factors. The remainder of this chapter focuses on three primary hazard domains: chemicals and biological illness; ergonomics; and social-​psychological factors and illnesses.

Chemical and Biological Occupational Risks Chemicals and biological agents can cause a wide range of diseases affecting virtually all bodily organs. Insights about respiratory disorders caused by inhaling certain dusts came early while, for example, cardiovascular effects are more recent. Although some chemicals may cause instant death at high doses (e.g. through blocked oxygen uptake or other vital bodily functions), they often cause disease long after exposure started or even after cessation. Moreover, the disease is often also common in the general population. This masks or limits awareness of the lived experience of such risks at the workplace as they take time to develop or are attributed to other sources. Thus, systematic approaches to assess and manage such risks are fundamentally important. Exposure to air pollutants (dust, vapours and fumes) and chemicals is common, both reported by around one of six respondents in the 2015 European Working Conditions Survey (EWCS 2015–​EU28). Cancer, hand eczema, cardiovascular, and reproductive effects are briefly described in the following sections in relation to chemicals’ exposure, followed by respiratory risks which may occur from both chemical and biological agents, and risks associated with handling biological material. Finally, different strategies to manage risks are described.

Cancer and Workplace Exposures The burden of occupational cancer on society is high (estimated at 102,500 deaths per year in EU 28, which is twenty times as many as deaths caused by occupational accidents). One of five workers in the EU are exposed to carcinogens at work. Exposure mainly occurs through inhalation, but also by skin contact. Lung cancer accounts for 54–​75 per cent of occupational cancer, conversely among men 17–​29 per cent of all lung

the cornerstone of job quality     223 cancer deaths may be attributable to occupational exposures. Overall, occupational exposures are estimated to cause 5–​8 per cent of all cancer deaths, with a higher proportion for men than women (Takala 2015). Authoritative evaluation of carcinogenic hazards for humans are made by the International Agency for Research on Cancer (IARC), which currently lists 47 occupational exposures shown to cause cancer (Loomis et al. 2018) and many others as probably carcinogenic to humans. Estimates for the UK indicate that the major causes are exposure to asbestos, shift work, mineral oils, solar radiation, crystalline silica, and diesel engine exhaust gases. For men the main sector of concern is construction, for women the main exposure is shift work (classified as probably carcinogenic to humans based on an increased risk for breast cancer) (Rushton 2017). It’s estimated that with modest intervention over 26 per cent of occupational cancers, or with stricter interventions more than 82 per cent could be avoided by 2060 (Hutchings et al. 2012). Many carcinogens act through mechanisms that also increase the risk for other diseases (e.g. respiratory, cardiovascular). This should be considered when assessing the total risk for an exposed population (see the section entitled ‘Management’).

Chemicals and Hand Eczema Skin contact with chemicals may cause eczema, usually on the hands, by inducing an allergic reaction (allergic contact dermatitis) or by irritation (non-​allergic or irritant contact dermatitis). Skin irritation is caused by broad groups of substances like cleaning agents, organic solvents, and cutting fluids, but also by natural fluids from, for example, plants and shellfish, prolonged contact with pure water, and by friction. Workers who had atopic eczema in childhood are at especially high risk from these exposures. Allergic contact dermatitis is caused by a sensitizing agent penetrating the skin and an immune reaction. Subsequently, eczema can be triggered by skin contact all over the body, not only where sensitizing contact occurred. If further exposure is avoided the eczema will usually heal or improve, but reappears after renewed contact. Over 3,000 agents causing allergic contact dermatitis have been identified, often small reactive chemicals binding to proteins, but 25 agents account for around half of the cases, including metals (nickel, chromium, cobalt), resins (including colophonium), rubber chemicals, preservatives, formaldehyde, dyes, acrylates, antiseptics, and fragrances. The main preventive principles are elimination of strong contact allergens, reduced skin contact and protective gloves. However, gloves must be carefully chosen since components in them may be sensitizing and permeation of chemicals through gloves is highly variable. Gloves used in contact with chemicals which permeate the glove, such as hair dyes, must not be used again since the chemical is released and uptake into the skin is enhanced inside the glove.

224    maria albin, chris mathieu, esa-pekka takala and töres theorell

Chemicals and Cardiovascular Disease Chemicals may cause cardiovascular disease through direct cardiotoxic effects on, for example, energy production, contractile mechanisms or rhythm regulations. This applies to several toxic metals (e.g. cobalt, arsenic, mercury). However, cardiovascular disease related to exposure to chemicals, dust and fumes is more generally believed to occur through oxidative stress and maybe a low-​grade inflammation. The exposure route is predominantly inhalation, and the dust/​chemical causes a local inflammation in the lung spilling over to systemic circulation, adding to the arteriosclerotic process as well as thrombosis formation (Fang, Cassidy and Christiani 2010). Exposures known to increase the risk for myocardial infarction include crystalline silica, diesel engine exhaust, welding fumes, polycyclic aromatic hydrocarbons, environmental tobacco smoke, and exposure to lead, arsenic and cadmium (Sjögren et al. 2020). Risk increase seems to occur at moderate (current) exposure levels although sufficient data to describe excess risk at specific air concentrations is rare. Belief that relatively low levels may suffice to increase risk is supported by findings from exposures in the general population. Thus, exposures to air pollution from road traffic in the order of micrograms/​m3 increase the risk for myocardial infarction, non-​smoking spouses of smokers are at increased risk, and lead exposure has been shown to increase blood pressure (a risk factor for heart disease) in the general population. The increase in risk observed in the general population is generally low for the individual. However, epidemiological studies of, for example, construction workers, suggest that the risk increase is substantial and contemplated when considering the need to reduce exposure levels. Therapeutic improvements permit many to return to work after myocardial infarction. Very few studies have investigated if the long-​term prognosis is affected by work exposures, although job-​strain as well as fear of unemployment were found to increase risk for reinfarction. A general and reasonable assumption is that factors known to induce the first-​time event are likely to increase the risk for a repeat, suggesting that possibilities to reduce the exposure to, for example, dust should be considered in dialogue with the worker returning to work after a cardiac event.

Adverse Reproductive Effects Chemical exposures may harm both male and female fertility, and foetuses. Lead, methylmercury and cadmium exposure, and common plasticizers like certain phthalates, have been shown to reduce sperm quality. Males exposed in utero to maternal, possibly also paternal, smoking tend to have lower sperm counts. These effects are discernible on a population level, but not thought to be of individual clinical importance unless counts are low for other reasons. Male sperm counts have declined over time in high-​income countries, being likely multifactorial.

the cornerstone of job quality     225 Chemicals’ effects on the ovaries are less studied, although exposure to polycyclic aromatic hydrocarbons (PAHs), lead, cadmium or mercury seem to induce toxic effects. Effects on foetuses are usually monitored as miscarriage, malformations, or reduced intrauterine growth. Many chemicals are known or highly suspected of inducing such adverse effects: lead exposure, methylmercury, ethylene oxide, and poorly protected work with cytostatic agents. Exposures shown to have reproductive toxicity, and those being mutagenic, carcinogenic or endocrine disrupting (disturbing hormonal balance) should be avoided (Lindbohm and Sallmén, n.d.). Risk assessment must include inhalation and potential systemic uptake through skin contact. Risk assessment in the absence of specific knowledge about safe levels, usually includes a safety factor in relation to known effect levels in adults. Furthermore, biological agents, most importantly rubella and toxoplasma, adversely affect foetuses unless the mother is immunized. Protection and rights of pregnant and breastfeeding women are regulated in EU Directive 92/​85. Recent findings from exposure to road traffic air pollution indicates adverse effects on intrauterine growth, and increased risk for maternal hypertension and pre-​eclampsia, also of concern for similar occupational exposures.

Respiratory Disorders from Chemicals, Dust, Fumes, and Biological Agents Inhalation of chemicals, dust, and fumes are shown in meta-​analyses to cause around 115 per cent of all cases of Chronic Obstructive Pulmonary Disease (COPD, in non-​ smokers 30–​50 per cent of cases). Occupational groups at increased risk include construction workers, miners, foundry workers, welders, farmers, cooks, and cleaners (Blanc et al. 2019). Similarly, around 15 per cent of all adult asthma cases have in several reviews been estimated to have occupational origin. Exposure to irritative agents (e.g. ozone, cleaning agents), small reactive molecules (e.g. isocyanates, persulphates, cyanoacrylate), and high-​molecular weight agents (allergens; e.g. flour and grain dust, animal proteins, latex) are the main exposures (Torén and Blanc 2009, Blanc et al. 2019). Specific dusts can also induce lung fibrosis (pneumoconiosis), with asbestos (asbestosis) and crystalline silica (silicosis) the main agents (Donaldson and Seaton 2012). Disease risk is related to cumulative exposure, often occurring late in working life and may progress after exposure has stopped. Although silicosis is the oldest known occupational respiratory disease, epidemics of severe, often lethal, silicosis have recently occurred in work environments with extremely high levels of crystalline silica, for example cutting and grinding artificial stone for benchtops, and dry sandblasting (denim jeans, reconditioning oil field equipment). Clusters of disabling lung disease (bronchiolitis obliterans) were originally described for workers exposed to diacetyl, a butter-​ flavouring agent in popcorn production, but later also with other flavouring agents used in the food industry and e-​cigarettes (Holden and Hines 2016).

226    maria albin, chris mathieu, esa-pekka takala and töres theorell Specific national regulations usually exist for exposure to fibrogenic dust and some small reactive molecules (e.g. isocyanates) requiring monitoring of exposure levels, pre-​ employment medical examination (fitness to work), and periodic examinations to detect early-​stage disease. Under these regulations diseased workers mustn’t continue in exposed work. Organic dust (mouldy material, endotoxins, compost, poultry and swine farming) may cause inhalation fever (Quirce et al. 2016). Fever episodes are usually short, but their occurrence indicates exposure levels sufficient to cause allergic alveolitis, a grave disease unless exposure ceases.

Hazards from Handling Biological Material Workers handling biological material may be exposed to viruses, bacteria, and other microorganisms which can cause acute and chronic, sometimes life-​threatening disease (EU-​OSHA 2009). Such exposures are especially common in the healthcare sector but can also occur in animal-​handling (zoonoses). In healthcare, high-​risk tasks include handling body fluids (especially blood), handling sharp instruments and needles with a risk for stick injuries and cuts in, for example, surgery, dentistry or taking blood samples, but also handling contaminated waste. HIV and Hepatitis B are serious risks from contact with blood. Moreover, bacteria which are multi-​resistant to antibiotics are an increasing problem as exemplified by multi-​resistant tuberculosis and methicillin resistant staphylococcus aureus (MRSA). During epidemics and pandemics, many occupational groups especially in the health care sector, may be at high risk, as illustrated by the covid-​19 pandemic (Carlsten et al. 2021). Bacteria forming spores may survive a long time outside the body; the bacillus causing anthrax can survive for decades in soil, and the bacterium causing tuberculosis can survive for months and is relatively resistant to disinfectants. During chemotherapy, patients’ urine, sweat, and vomit often contain active metabolites of the chemotherapeutic agent. Concentrations may be high, entailing a need to protect care-​giving staff and cleaners against skin contact. Measures preventing transmission of disease from direct contact with patients are well known and established in healthcare, but strict routines need to be followed to control risks. Hepatitis B vaccination of medical, nursing and paramedical staff is mandatory in some countries, and otherwise recommended.

Managing Chemical and Biological Risks National and International Level Managing chemical and biological risks relies on a systematic approach to risk assessment, control and monitoring exposure levels. We cannot use our senses to assess

the cornerstone of job quality     227 if an exposure to some agents is safe. A chemical may have a strong unpleasant smell at extremely low air concentrations far below the lowest observed adverse effect level (LOAEL), it may also be totally without odour up to lethal concentrations (carbon monoxide), and the small particles which harm the lung in exposure to, for example, crystalline silica are not visible to the naked eye. International agencies, like the International Labour Organization (ILO), and the European Agency for Safety and Health at Work (EU-​OSHA), establish norms regarding handling dangerous substances and dusts. The European Commission initiatives on occupational exposure limits (OELs) may be indicative (health-​based) or, for a more limited number of substances, binding (taking socioeconomic and technical feasibility factors into account). Most limits refer to air concentrations, but biological limit values may also be set and sometimes the preferred option (e.g. monitoring lead exposure). National OELs may be based on, for example, EU standards, or own assessments. Comparisons of national OELs show significant differences even when revised at a similar point in time (Schenk 2010). An overview of national OELs is given by the Institut für Arbetsschutz der Deutschen Gesetzlichen Unvallsversicherung (IFA; https://​www.dguv.de/​ifa/​ges​tis/​ges​tis-​int​erna​tion​ale-​gre​nzwe​rte-​fuer-​chemis​che-​sub​ stan​zen-​limit-​val​ues-​for-​chemi​cal-​age​nts/​index-​2.jsp). The national OELs generally include consideration of economic consequences and feasibility. This is important. They cannot generally be seen as a ‘safe’ limit. For example, the OEL for respirable silica is 0.1 mg/​m3 in most countries, while the consensus among experts in the field is that serious adverse effects are seen down to 0.025 mg/​m3. The American Conference of Industrial Hygienists (ACGIH) provides a yearly update with thorough documentation of health-​based recommended threshold limit values for more than 700 substances and physical agents, and is the most cited source for such values. However, they have in some instances been seen as high, especially for endocrine disruptors.

Workplace Level Systematic approaches to safe use and handling of chemicals is vital, as is necessary expertise. The Occupational Health Service (OHS) is described in the ILO convention as a service for advising employers, workers, and their representatives in establishing and maintaining a safe and healthy working environment, including identification and assessment of risks from health hazards in the workplace (ILO 1985). In some instances, several exposures at the workplace may have the same target organ (e.g. be neurotoxic or harmful to the airways) which needs to be considered. Assessment and management of risks requires both high professional skills and good knowledge about processes and conditions at the workplace. A challenge is that the OHS may often have a short contract with the employer, and not include risk identification and assessment. Further, access to OHS is in some instances poorly matched to the needs for external competence, that is, much lower in small-​and medium-​sized enterprises. Factors contributing to a good collaboration between the OHS and employers were recently reviewed (Halonen et al. 2017).

228    maria albin, chris mathieu, esa-pekka takala and töres theorell

Individual Awareness Individuals should be informed by their employer about potential health risks and necessary precautions. Safety measures are often poorly implemented, both due to circumstances (time pressure, availability of adequate personal protective equipment with good fit) and insufficient knowledge (mask removed while air is still contaminated). Educational activities are important (Schulte et al. 2005), such as teaching of adequate protective measures at vocational schools and improving the use of gloves among hairdressing apprentices. Such strategies are important to make young workers aware of health risks in, for example, construction, hairdressing or healthcare in a situation where traditional approaches (labour inspection, union safety officers) are weakened in many countries. Digital technologies enhance possibilities to exchange and get adequate safety information, and new technologies to assess exposure which may be self-​administrated (passive air samplers, analyses of dried blood sampled as in self-​control of diabetes) may contribute to workers’ future possibilities to obtain independent risk assessments.

Ergonomics The word ‘ergonomics’ is a combination of two Greek words, ‘ergon’ (‘work’) and ‘nomos’ (‘natural law’). It means scientific approaches to designing and arranging things people use so that the people and things interact most efficiently and safely. Synonyms are ‘human engineering’ and ‘human factors’ (International Ergonomics Association http://​iea.cc). Ergonomics’ aim is to adjust things in work and leisure according to people’s needs, abilities, and limitations, and simultaneously to improve productivity. The domain of ergonomics has commonly been described in terms of three approaches: Physical ergonomics concerns the human body and functions related to physical actions. Relevant topics are, for example, variation in body size (anthropometrics) and physical capacity and how they influence the performance of work. Common targets are heavy manual work, repetitive work, and working postures. Cognitive ergonomics concerns mental processes, such as perception, memory, reasoning, and motor response. Mental workload, decision-​making, human reliability, and skills training are among the targets of actions. Organizational ergonomics aims to optimize sociotechnical systems, including organizational structures, policies, and processes. Related issues are, for example, communication and co-​working, distribution of human and other resources, working schedules, and quality management. This section presents two interaction models of the human and work: 1) In the Human–​Work interaction model the scope is in the structural and functional abilities and limitations of the human being. 2) The Human–​Work System model’s viewpoint is work as a system. After these theoretical models are presented, some practical issues on ergonomics are described.

the cornerstone of job quality     229

Human–​Work Interaction Figure 10.1 describes the interaction between the human individual and work. Based on the information signals received from the thing (work), the human makes decisions (conscious or unconscious), and acts towards the work. Within this loop there are several points where ergonomics can help the human act more efficiently. 1. The human needs information about the object. Our environment is full of all kinds of information detected through our senses: visual, auditive, tactile, kinetic, odour, temperature, pressure, and other signals. To help the human detect the important signals from the continuous mass of information (noise), work should be designed so that relevant information and signals are strong enough to be detected easily. Sometimes technical amplifiers are needed. 2. Even though the work gives clear and relevant signals that the human can detect in good conditions, the environment can blur the signals so that our senses cannot perceive them. Lighting and reflections may hinder visual detection. Physical barriers can prevent seeing, for example walls or beams in the cab of a vehicle or control room may form blind spots. In an office loud speech can confuse our auditive performance. Personal (safety) equipment can sometimes reduce the signals, for example protective gloves, clothing, or hearing protectors. Ergonomics means designing and reshaping the environment so that it doesn’t prevent the human from perceiving relevant information from work. Sometimes personal aids like eyeglasses help the person’s impairments.

F a t i g u e 5

Human

2

Information

1

Perception Work

3 Processing of information

4 Muscular action

Figure 10.1  Interaction between the human and work

230    maria albin, chris mathieu, esa-pekka takala and töres theorell 3. Processing of information. Mental models guide information-​processing and decision-​making. They are dependent on previous experiences, training, and the person’s mental skills. Mental models also help the perception, detection, and filtering of relevant from irrelevant information. For example, a radiologist can see with a short glance signs of a disease in an x-​ray picture whereas a layman cannot even know the organ it presents. Likewise, we cannot follow instructions given in a language we don’t know. Similarly, the use of symbols that have different meanings in different cultures may result in misunderstanding. Ergonomics’ role is to design and reshape work so that the workers can interpret information and make relevant decisions with the mental models they have. Our vision is superior to other senses for processing information. Therefore, multiple information is easier to conceptualize from graphic presentations than from facts described with words. Visual maps, flowcharts, and statistical graphs are examples of how to describe complicated information in an easy to handle way. If words are needed, plain language is preferred. In an ideal ergonomic design, the result is such that we can intuitively understand the meaning of information by the mental models that are common for most people. Still, specific training is often needed to clarify things. An ergonomic training programme takes account of the background skills and mental models of the trainees in order to help effective learning. 4. Muscular action. Neuromuscular activity results in actions to manipulate the work. Gross movements or holding of postures are examples. In physically light work without forceful actions small muscles are active. Writing or drawing with a pen or a computer includes visible small movements. Eye muscles are active to keep the gaze to detect visual signals. Speech or other auditive action requires activity of numerous muscles from those for breathing, the vocal chords, tongue and mouth. Muscular activity must resist biomechanical forces. Biomechanics applies the mechanics of physics to the human body. In a simple biomechanical model, joints are hinges or lever-​points, bones and the insertion points of the muscles are lever-​arms, and muscular forces keep the balance against the external forces or produce movements. Gravity is the most important external force acting on the weights handled and on body parts. Other forces we must resist are those used in several man-​made things (e.g. springs, magnetic locks, machines), and friction. The lever-​arms of muscles are short with respect to the length of the bones that serve as lever-​arms of external forces. Therefore, quite small external forces can result within the body in such high forces that they are close to or even higher than what muscles and other tissues can resist. For biomechanical reasons, more muscular effort is needed to resist external forces in constrained body postures and extremities than in neutral postures. Therefore, muscles must work more in an ‘unnatural’ posture.

the cornerstone of job quality     231 The human body is an adaptive system. Muscles, bones, tendons, skin, and the cardiorespiratory system react to external loads by strengthening; provided the forces are not too great and cause tissue damage. On the other hand, low external loads weaken the body. For strengthening and keeping fitness, sufficient load and recovery periods are necessary. Ergonomics aims to optimize biomechanical loads so persons can use their muscles most effectively, risk of injuries is kept low, and loading sufficient to keep basic fitness. This can be reached with many physical ergonomics classical solutions targeting excessive loads or postures. 5. Fatigue. Our body is not a machine and cannot maintain the same performance for long periods. Signs of fatigue will appear within tens of seconds in high-​intensive muscular effort and within a few tens of minutes in continuous medium intensive effort. Muscle fatigue can be seen within tens of minutes in continuous static holding of the same posture with low external load, for example in office work. Also, the senses and mental processes fatigue with high levels of information and mental load. Decrease in performance can be seen within few tens of minutes in intensive vigilance tasks (Basner et al. 2011). In addition to reducing performance, fatigue increases the risk of errors, which can lead to accidents. Errors also reduce productivity and quality. Therefore, the savings gained for production via extra working hours in a hurried or intensive situation may be lost due to the extra time needed for corrections of failures in the following days. Sufficient recovery is needed to maintain human performance over workhours. The higher the load, the longer the time needed for recovery. Scientific research, unfortunately, cannot give clear time limits for loading and recovery. A good piece of advice is to rest as often as possible. In data-​entry work five extra minutes rest every hour has reduced discomfort and increased productivity despite the shortening of the total active working time of the eight-​hour-​day (Galinsky et al. 2007). In the recommendations of manual material handling, the maximal weights are related to the duration and frequency of the tasks (the NIOSH equation, NIOSH 1994). Fatigue can be avoided by technical and organizational means to reduce external loads, and via organizational means by adding sufficient work variation and recovery periods.

Human–​Work System Model Interaction between the human and work can also be modelled by four basic elements of the work system around the individual worker (Fig. 10.2; Smith and Sainfort 1989). 1) The job includes tasks placing several physical and mental requirements on the worker. Due to individual variation, identical sets of tasks can be perfect for

232    maria albin, chris mathieu, esa-pekka takala and töres theorell Technology, tools

Organisation

Worker

Design of job/tasks

Environment

Figure 10.2  The work system Modified from Smith and Sainfort 1989

one person but straining for another. 2) The tasks are performed with suitable technologies and tools. Similar tasks can be performed with various styles and technologies that make task performance easy or difficult. 3) The work environment consists of the physical and social surroundings influencing the performance of work and the worker’s loading. Accomplishing the same task in poor outdoor conditions among hostile people is totally different from an office-​type environment and supportive social atmosphere. 4) Organization of work has a high impact of variation of working conditions. Work can be done alone or shared with colleagues; there may be sufficient or insufficient resources; the work community can consist of workers with similar or different backgrounds and culture; etc. All these elements influence work performance and worker loading. Dynamic interactions mean that changes in one element usually result in changes in some other elements. Worth noticing is that this simple system close around an individual worker may be influenced by several external conditions (e.g. legislation, macroeconomics, industrial standards etc. Carayon et al. 2015). Additionally, production systems often include several working units and the product output of one unit is an input to the next unit in the process. The described system around an individual worker is a part of a larger system of units and processes, and therefore changes in any parts of the whole system may have effects on other parts elsewhere. An ergonomic solution is always an optimizing task between these elements. In designing or remodelling work elements, the potential effects of solutions on other parts of the system should be evaluated in order to avoid unexpected problems elsewhere in the whole system.

Ergonomics in Practice At its best, ergonomics is involved in the workplace’s continuous development and improvement processes. The organization’s top management has to make decisions on development policy in order to give sufficient resources and to define the limits and constraints.

the cornerstone of job quality     233 Table 10.1 Ergonomics actions as a process 1

Policy

2

Locate the job within the whole production process

3

2.1

Split the job to tasks and subtasks

2.2

Evaluate each task for problems

2.3

Evaluate the importance of problems

2.4

Try to find the roots of the problem in the process

Generate solutions and select those to be used 3.1

For each potential solution evaluate the potential benefits and harms in the whole production system

3.2

Rank the solutions by benefits, harms, and needs for investment

4

Plan the process to fulfil the changes

5

Perform the changes

6

Check the effects

Actions start with a thorough analysis of the work system to identify the location of potential for development. A usual way is to start with different kinds of documentation available in the organization, for example safety documents (safety inspections, accident reports, near accidents), production documents (variation and errors in productivity), occupational health documents (sick-​leave in different working units), or data from human resource departments (high worker turnover, surveys of worker well-​being). Individual workers with work-​related health symptoms also indicate what should be checked regarding working conditions. After identification of the problems or inconveniences in some units or jobs, the next step is to go further by splitting the job and working process into tasks and subtasks. These can sometimes be seen in the technical documentation of the quality management system. Often the real work deviates from the official process description or there is no such data. Therefore, it is important to interview the workers and supervisors about task performance. As the tasks and work processes have been identified, each should be evaluated for harms and needs for improvement. Checklists for ergonomics and safety can be used to identify the problems in the human-​work interaction (see Fig. 10.1) or in the parts of the human-​work system (Fig. 10.2). For each task or point detected in need of improvement, the importance is evaluated, for example with the common risk-​evaluation system matrix of likelihood and severity of harm. The most important points are selected for further analysis. In the next step the potential causes of the selected problems are identified and analysed. For this it is important to evaluate the whole process of the production system in order to find the roots of problems. Actions on the roots usually have highest potential

234    maria albin, chris mathieu, esa-pekka takala and töres theorell effect. For example, if an individual worker has problems due to manual handling of heavy objects, the cause may be in the task, environment, tools, or organization of this particular job and task. Manual handling of heavy objects may also be due to some other phase in the whole production process. In that case it’s important to focus analysis on these phases and causes. The final solution may be elimination of the problem by changes in other parts of the whole production process. The next step is to generate potential solutions. Each solution is further analysed for the benefits and harms in the job/​task where the problem was identified, but also in the other parts of the whole work system and processes.

Participation and Co-​Work Regardless of the technical documentation, tasks are often performed in different ways, which workers know best. Workers also may have ideas for improvements and are in the frontline to implement many solutions. Foremen and upper-​level staff know the whole production system better than the individual workers. Designers know solutions to similar developments. Occupational safety personnel can evaluate the potential safety risks of problematic jobs and proposed solutions. Human resource actors are also essential. To be successful, ergonomic actions must integrate the participation of all these actors at the workplace in co-​work. Responsibility for development and safety lies with the organization. The main role of occupational health is to inform the company about the health problems they meet through the health problems of workers. Health professionals also have more knowledge of the limitations of human beings than technicians, and can evaluate proposed solutions from the human point of view.

Psychosocial Health Determinants in the Work Environment Scientific literature since the 1980s focuses on the role of working conditions in the pathogenesis of ischemic heart disease (IHD), depression and a growing number of disease outcomes including sick-​leave, stress and psychosocial factors. Although there were theoretical models available for studying psychosocial factors already in the 1960s and 1970s, they were not widely used in epidemiological studies. This changed when Person Environment Fit (PEF) was introduced at the Institute for Social Research in Ann Arbor, Michigan. For the first time it was possible to relate psychosocial factors to a formal theory accepted by several researchers. When the researchers (Kahn 1974; Sales and House 1971) introduced the PEF model, the overriding theory was that the capacity and competence of the individual has to match the characteristics of the job—​otherwise an elevated risk of illness development would occur. Other researchers

the cornerstone of job quality     235 felt that the use of this approach would lead to an underestimation of the importance of work organization and the possibility that improved work organization may have in health promotion. That kind of approach was stimulated when Karasek introduced his demand control (DC) model in 1979 (Karasek 1979). The demand control model with the addition of social support at work, DCS (Johnson and Hall 1988, Karasek and Theorell 1990) has been extensively used in research over the past 35 years. The DCS theory postulates that combinations of excessive psychological demands and poor possibility for the employee to exert control over and to obtain social support at work give rise to increased employee health risks. A closer look at psychological demands shows many kinds of psychological demands—​cognitive and emotional—​ranging from quantitative (how many tasks per time unit?) to qualitative (how difficult are the tasks?). According to the theory, a work situation is particularly psychologically demanding when the tempo is constantly high and at the same time the tasks are difficult and require constant attention. Similarly, control has several components. In the terminology introduced by Karasek, the concept decision latitude is used for the possibility that the work organization gives the employee to exert control over his/​her situation. Decision latitude has two components, namely decision authority, which is the worker’s possibility to exert control over the work process in daily practice and skill discretion, which is the worker’s possibility to develop skills so that s/​he can take control in unexpected situations. Finally, social support has several components; support from superiors and colleagues as well as emotional and instrumental support. Subsequent alternatives to the DC and DCS models have been utilized. The effort–​ reward imbalance model (ERI) is the most frequently used alternative (Siegrist 1996). It postulates that when there is imbalance between the employee’s effort and the reward s/​ he receives, adverse health risks arise. Although the concepts demand and effort partly overlap, they reflect different aspects of the work situation. While the construction of the questions regarding psychological demands in the DCS model aim at demands the organization makes on the individual, the effort concept addresses the individual´s perception of the effort that s/​he is making for the organization. Similarly, the reward concept addresses the rewards that the individual feels that s/​he gets from the organization. Three kinds of rewards are addressed, namely material (salary), social (promotion and esteem) and psychological (feedback). A third model is the demand resource model which postulates that high demands have adverse health effects only when resources are small for the individual worker. Resources are broadly defined and related both to individual and organizational resources. A more flexible assessment philosophy is used allowing for variations in assessment depending on the organizational context (Demerouti et al. 2001). Several other work-​related psychosocial factors have been studied in relation to illness outcomes. Long working hours (Kivimäki et al. 2015, Kivimäki et al. 2017), job insecurity (Ferrie et al. 2016), shift and night work (Knutsson and Böggild 2000, Knutsson and Kempe 2014) and organizational downsizing (Andreeva et al. 2015, Brenner et al. 2014) have been studied in relation to several disease outcomes such as coronary heart disease, stroke, diabetes 2 and atrial fibrillation and psychiatric depression (Theorell

236    maria albin, chris mathieu, esa-pekka takala and töres theorell et al. 2015, 2016, Fransson et al. 2018). In the majority of studies there are positive findings although the relative risks are relatively low, mostly ranging between 1.2 and 1.8. It should be pointed out these illness outcomes only partly overlap. One has to take into account a total effect on morbidity. Therefore, for instance when a more general illness outcome approach is used (Magnusson Hanson et al. 2018)—​with an analysis of lost years because of any illness—​it becomes obvious that long-​lasting exposure to job strain shortens a healthy life and it also increases the risk of long-​lasting sick-​leave (Mutambudzi et al. 2018). Job strain is the work organization model presently most extensively studied. A new development in job strain research is studying mortality related to exposure to job strain in vulnerable subjects with cardiometabolic disease, for example subjects with manifest coronary heart disease or diabetes. In this group there are particularly strong effects—​ increased mortality after exposure to job strain (Kivimäki et al. 2018). These results point out detrimental effects of long-​term exposure to job strain, but a new interesting development (Pan et al. 2018) is research on other aspects of the demand control model showing that long-​term exposure to the combination of high psychological demands and high control (so-​called active work) seems to protect against cognitive decline in old age, even when adjustments have been made for education level and age. Reviews on the most extensively studied psychiatric illness outcome, development of depressive feelings, show findings similar to those on cardiovascular disease (e.g. Theorell et al. 2015, Siegrist and Wahrendorf 2016). In general, psychosocial work factors increasing the risk of cardiovascular illness development are also established risk factors for depression.

What Mechanisms are Operating? For some chemical work exposures a direct chemical biological effect is likely. Toxicity can be expressed in many different ways. Breakdown of a toxic agent may require large amounts of an enzyme needed for regulation of normal functions. Another example: Recently, immunological mechanisms have been discussed and seem particularly relevant for quartz dust, tobacco smoke and motor exhaust. These kinds of chemical exposures activate the same immunological mechanisms relevant for psychosocial factors—​exposure to inhaled dust, tobacco smoke and motor exhaust activate proinflammatory cytokines which start inflammatory reactions in the vessel walls—​ the exact same mechanism we discuss in relation to cardiovascular long-​term effects of psychosocial stress. Regenerative functions related among other things to the parasympathetic system may in such situations protect the artery wall from long-​term adverse effects. Long-​term stress may inhibit such protective mechanisms. Hence, when there is adverse exposure, the same injuring and protecting biological mechanisms are operating for psychosocial and organizational factors as for adverse chemical exposure (see Theorell et al. 2015 for a more detailed discussion).

the cornerstone of job quality     237 Job insecurity is less established than job strain and poor decision latitude. However, since job insecurity is an important potential risk factor in the modern working world with more and more jobs becoming redundant, more future studies are recommended. The same relates to low support, unfavourable social climate, lack of procedural and relational justice, conflicts with superiors and colleagues and limited skill discretion. It should be noted that some adverse conditions discussed here show some overlap but are also partly unrelated to one another. The implication of this is important since it means that the total effect of adverse working conditions is much greater than each of the relatively small odds ratios indicate. Recent research studies of the association between work environment factors and illness outcomes have become increasingly sophisticated. For instance, Shields (2006), Stansfeld et al. (2012) and De Lange et al. (2002) have examined possible effects of exposure at least twice, or even three times, in follow-​up survey waves on job strain. Their findings indicate that accumulated or increasing job strain has a stronger adverse statistical effect on risk of experiencing increased depressive symptoms during follow-​up than decreasing job strain. Similar observations have been made on the relationship between psychosocial working conditions and coronary heart disease (e.g. Chandola et al. 2008). Accordingly, stronger evidence regarding the influence of working conditions on poor health may be expected in future research with a growing body of studies with such methodology.

Societal Relevance As argued by several authors (e.g. Theorell 2012) improvement of working conditions is not only about eliminating and reducing adverse factors but also about stimulation of regenerative and protective biological processes. In psychosocial terms this refers to improved social support and leadership, and improved decision latitude for employees. Work environment factors for which scientific evidence exists are possible to influence by means of organizational changes. Furthermore, certain psychosocial work conditions make it difficult for employees to protect themselves against adverse physical (e.g. noise) and chemical exposures. Accordingly, psychosocial factors are intertwined also with the chemical and physical exposures. Decision latitude for employees can be improved by analysis of work organization with subsequent goal-​directed organization intervention (Michie and Williams 2003) or by year-​long education of managers about psychosocial factors (Theorell et al. 2001, Romanowska et al. 2011). Reviews of natural experiments designed to reduce work environment psychosocial risks show that such interventions may result in reduced biological stress and improved health (Jauregui and Schnall 2009, Hasson et al. 2012, Montano et al. 2014). Current results suggest that in assessment and treatment plans for patients with already manifest IHD, work environment should be taken into account since work environment appears even more important in secondary prevention than in primary prevention.

238    maria albin, chris mathieu, esa-pekka takala and töres theorell

Conclusion This chapter focused on three central domains of OSH relevant to job quality due to their bearing on somatic and mental health. Though chemicals and biological, ergonomic, and psychosocial risks and illnesses have different sources, they sometimes interact and operate along similar lines. Individuals or categories of workers can be at risk not just from discrete hazards within a given domain, but even multiple risks within, as well as across domains, potentially subjecting them to complex multiplier effects. Total cumulative effects are greater than discrete odds and effects. As our working lives extend (we live and work longer) and become more varied (we increasingly change jobs, occupations, and even branches), our work legacies in terms of hazards and risks we are exposed to and bear with us also become more varied and comprehensive. Additionally, new hazards, risks and diseases are uncovered, and we attain better knowledge about those we have been aware of and their effects. Attention needs to be paid to how risks and illnesses compound and interact, as well as recognizing that the biological, physical, cognitive, mental, and organizational intertwine. That these are domains of specialist knowledge with different assessment methods exacerbates this challenge, making it difficult to deal adequately with them in a comprehensive manner, especially in practice when the next issue, the relationship between the generic and the individual is encountered. The relationship between the generic and the individual is seen in the tension between often needing to work in terms of general standards, developed for a generic individual, and the reality that risks and effects are encountered and experienced by persons with wide ranges of variation in their physiological and mental capabilities and needs. This is compounded by general societal inequalities in health, especially for vulnerable groups, who bear these disparities to their workplaces. Human variation exists in many important parameters—​from size, strength, and reproductive organs/​capacity, to resistance, susceptibility, and exposure. Even for given individuals these are largely dynamic and change over time. These parameters obtain both for detrimental and regenerative or recuperative processes. Furthermore, our lives are not just led in one setting, that is, not just at work, and even if exposures to hazards are similar for employees at work, they may vary in home and leisure settings, or the opposite may obtain. However, as workplaces are more controlled environments, individual-​level monitoring and adapted application of generic standards is possible in this context with sufficient will and responsibility. A related issue evident in this chapter is that sometimes OSH activities focus on the individual, and sometimes on higher systems levels. This was most explicit in discussions of ergonomic models, but can also be seen as the basis of the relationship between health and economic productivity or efficiency. This chapter recurrently displayed the inextricable intertwining of health and economics, sometimes in an adversarial and sometimes in a mutually dependent or reinforcing manner. Health is usually conceived as accruing to individuals,

the cornerstone of job quality     239 whereas economic efficiency returns accrue to firms, organizations, or society. Health and productivity often align as productivity relies on health and human capacity, and therefore investment in protecting and promoting health serves to protect and promote productive capacity. Furthermore, specific aspects of ergonomics and psychosocial workplace interventions explicitly aim to advance productivity by streamlining work-​processes by removing distractive or counterproductive elements. Efficiencies are attained by not overtaxing bodies, minds or social systems, and recognizing the importance of recuperation, regeneration, and decision latitude. But investment in health can be costly and health-​promoting actions can consume time and financial resources. Therefore, they frequently square off in cost–​benefit calculations, the outcomes of which depend on who carries out these calculations, who bears the burdens and who carries the advantages, and who has the power to impact calculations and decision-​making. As in all realms of job quality, benefits and detriments of various types—​in this chapter health, illness, injury and death, are outcomes of decisions. Some are made by workers themselves, but most frequently by managers, employers, and policymakers, who rely on and choose between information and other decision-​support mechanisms. Workplace safety and health has improved greatly, and continues to improve, as the science that underlies OSH improves. But not equally for everyone, as decisions are made in different jurisdictions and at different levels. Again, as in most realms of job quality, our capacity for beneficial improvement is contingent on will, not just knowledge.

References Andreeva, E., Magnusson Hanson, L. L., Westerlund, H., Theorell, T., and Brenner, M. (2015), ‘Depressive Symptoms as a Cause and Effect of Job Loss in Men and Women: Evidence in the Context of Organisational Downsizing from the Swedish Longitudinal Occupational Survey of Health’, BMC Public Health, 12/​15, 1045. Basner M., and Rubinstein J. (2011), ‘Fitness for Duty: A 3-​Minute Version of the Psychomotor Vigilance Test Predicts Fatigue-​ Related Declines in Luggage-​ Screening Performance’, Journal of Occupational and Environmental Medicine, 53/​10, 1146–​1154. Blanc, P. D., Annesi-​Maesano, I., Balmes, J. R., Cummings, K. J., Fishwick, D., Miedinger, D., Murgia, N., Naidoo, R. N., Reynolds, C. J., Sigsgaard, T., Torén, K., Vinnikov, D., and Redlich, C. A. (2019). ‘The Occupational Burden of Nonmalignant Respiratory Diseases. An Official American Thoracic Society and European Respiratory Society Statement’, American Journal of Respiratory and Critical Care Medicine, 199/​11, 1312–​1334. Brenner M. H., Andreeva, E., Theorell, T., Goldberg, M., Westerlund, H., Leineweber, C. et al. (2014), ‘Organizational Downsizing and Depressive Symptoms in the European Recession: The Experience of Workers in France, Hungary, Sweden and the United Kingdom’, PLoS One, (19 May), 9/​5, e97063. Burke, R., and Cooper, C. (eds.) (2018), Violence and Abuse in and around Organisations (London: Routledge).

240    maria albin, chris mathieu, esa-pekka takala and töres theorell Carayon P., Hancock, P., Leveson, N., Noy, I, Sznelwar, L., and van Hootegem, G. (2015), Advancing a Sociotechnical Systems Approach to Workplace Safety—​Developing the Conceptual Framework’, Ergonomics, 58/​4, 548–​564. Carlsten, C., Gulati, M., Hines, S., Rose, C., Scott, K., Tarlo, S. M., Torén, K., Sood, A., and de la Hoz, R. E. (2021), ‘COVID-​19 as an Occupational Disease’, American Journal of Industrial Medicine, 64/​4, 227–​237. https://​doi-​org.proxy.kib.ki.se/​10.1002/​ajim.23222 Chandola, T., Britton, A., Brunner, E., Hemingway, H., Malik, M., Kumari, M. et al. (2008), ‘Work Stress and Coronary Heart Disease: What Are the Mechanisms?’ European Heart Journal, 29, 640–​648. Demerouti, E., Bakker, A. B., Nachreiner, F., and Schaufeli, W. B. (2001), ‘The Job Demands: Resources Model of Burnout’, Journal of Applied Psychology, 86, 499–​512. Donaldson K, and Seaton, A. (2012), ‘A Short History of the Toxicology of Inhaled Particles’, Particle and Fibre Toxicology, 6/​9, 13. EU-​OSHA (2009), ‘Biological Agents and Pandemics: Review of the Literature and National Policies’, European Agency for Safety and Health at Work (EU-​OSHA) European Risk Observatory Working Paper. Brussels. EU-​OSHA (2017), An International Comparison of the Cost of Work-​Related Accidents and Illnesses, https://​osha.europa.eu/​en/​tools-​and-​publications/​publications/​international-​ comparison-​cost-​work-​related-​accidents-​and/​view Fang, S. C., Cassidy, A., and Christiani, D. C. (2010), ‘A Systematic Review of Occupational Exposure to Particulate Matter and Cardiovascular Disease’, International Journal of Environmental Research and Public Health, 7/​4, 1773–​1806. Ferrie, J. E., Virtanen, M., Jokela, M., Madsen, I. E. H., Heikkilä, K., Alfredsson, L. et al. (2016), ’Job Insecurity and Risk of Diabetes: A Meta-​Analysis of Individual Participant Data’, Canadian Medical Association Journal, 188/​17–​18, E447–​E455. Fransson, E. I., Nordin, M., Magnusson Hanson, L. L., Westerlund, H. (2018), Job Strain and Atrial Fibrillation: Results from the Swedish Longitudinal Occupational Survey of Health and Meta-​Analysis of Three Studies’, European Journal of Preventive Cardiology, 25/​11, 1142–​1149. Galinsky T., Swanson, N., Sauter, S., Dunkin, R., Hurrell, J., and Schleifer, L. (2007), ‘Supplementary Breaks and Stretching Exercises for Data Entry Operators: A Follow-​Up Field Study’, American Journal of Industrial Medicine, 50/​7, 519–​527. Halonen J. I., Atkins, S., Hakulinen, H., Pesonen, S., and Uitti, J. (2017), ‘Collaboration between Employers and Occupational Health Service Providers: A Systematic Review of Key Characteristics’, BMC Public Health, 17/​1, 22. Hämäläinen, P., Takala, J., and Kiat, T. (2017), Global Estimates of Occupational Accidents and Work-​Related Illnesses (Singapore: Workplace Safety and Health Institute), http://​www. icoh​web.org/​site/​ima​ges/​news/​pdf/​Rep​ort%20Glo​bal%20Es​tima​tes%20of%20O​ccup​atio​ nal%20Ac​cide​nts%20and%20W​ork-​rela​ted%20Il​lnes​ses%202​017%20r​ev1.pdf. Hasson, H., Gilbert-​Ouimet, M., Baril-​Gingras, G., Brisson, C., Vézina, M., Bourbonnais, R., and Montreuil, S. (2012), ‘Implementation of an Organizational-​Level Intervention on the Psychosocial Environment of Work: Comparison of Managers’ and Employees’ Views’, Journal of Occupational and Environmental Medicine, 54, 85–​91. Holden V. K., and Hines, S. E. (2016), ‘Update on Flavoring-​Induced Lung Disease’, Current Opinion in Pulmonary Medicine, 22/​2, 158–​164. Hutchings S. J., Cherrie, J. W., Van Tongeren, M., and Rushton, L. (2012), ‘Intervening to Reduce the Future Burden of Occupationally-​Related Cancers in Britain: What Could Work?’ Cancer Prevention Research, 5, 1213–​1222.

the cornerstone of job quality     241 IEA, n.d., International Ergonomics Association (IEA), www.iea.cc. ILO (1985), Occupational Health Services Convention, International Labour Organisation (No 161. Convention concerning Occupational Health Services [Entry into force: 17 Feb 1988]) (Geneva: International Labour Organisation). Jauregui, M., and Schnall, P. L. (2009), ‘Work, Psychosocial Stressors and the Bottom Line’, in P. L. Schnall, M. Dobson, and E. Rosskam, eds., Unhealthy Work: Causes, Consequences, Cures (Amityville, NY: Baywood Publishing), 153–​168. Johnson, J. V. and Hall, E. M. (1988), ‘Job Strain, Work Place Social Support, and Cardiovascular Disease: A Cross-​Sectional Study of a Random Sample of the Swedish Working Population’, American Journal of Public Health, 78, 1336–​1342. Johanson, G. and Tinnerberg, H. (2019), ‘Binding Occupational Exposure Limits for Carcinogens in the EU -​Good or Bad? Scandinavian Journal of Work, Environment and Health 45/​3, 213–​214. Kahn R. L. (1974), ‘On the Meaning of Work’, Journal of Occupational Medicine, 16/​11, 716–​7 19. Karasek, R. (1979), ‘Job Demands, Job Decision Latitude and Mental Strain: Implications for Job Redesign’, Administrative Science Quarterly, 24, 285–​307. Karasek, R. A. and Theorell, T. (1990), Healthy Work (New York Basic Books). Kivimäki, M., Jokela, M., Nyberg, S.T., Singh-​Manoux, A., Fransson, E. I., Alfredsson, L. et al. (2015), ’Long Working Hours and Risk of Coronary Heart Disease and Stroke: A Systematic Review and Meta-​Analysis of Published and Unpublished Data for 603 838 Individuals’, Lancet, 386, 1739–​1746. Kivimäki, M., Nyberg, S. T., Batty, G. D., Fransson, E. I., Heikkilä, K., Alfredsson, L. et al. (2012), ‘Job Strain as a Risk Factor for Coronary Heart Disease: A Collaborative Meta-​Analysis of Individual Participant Data’, Lancet, 380, 1491–​1497. Kivimäki, M., Nyberg, S. T., Batty, G. D., Kawachi, I., Jokela, M., Alfredsson, L. et al. (2017), Long Working Hours as a Risk Factor for Atrial Fibrillation: A Multi-​Cohort Study’, European Heart Journal, 38, 2621–​2628. Kivimäki, M., Pentti, J., Ferrie, J. E., Batty, G. D., Nyberg, S. T., Jokela, M. et al. (2018), ‘Work Stress and Risk of Death in Men and Women with and without Cardiometabolic Disease: A Multicohort Study’, Lancet Diabetes & Endocrinology. pii: S2213-​8587, 18, 30140–​30142. Kivimäki, M., and Steptoe, A. (2018), ‘Effects of Stress on the Development and Progression of Cardiovascular Disease’, Nature Reviews Cardiology (April 15), 215–​229. Knutsson, A., and Bøggild, H. (2000), ‘Shiftwork and Cardiovascular Disease: Review of Disease Mechanisms’, Reviews on Environmental Health (Oct–​Dec), 15/​4, 359–​372. Knutsson, A., and Kempe, A. (2014), ‘Shift work and Diabetes-​ -A ​ Systematic Review’, Chronobiology International (Dec.), 31/​10, 1146–​1151. de Lange, A., Taris, T., Kompier, M., Houtman, I., and Bongers, P. (2002), ‘Effects of Stable and Changing Demand-​Control Histories on Worker Health’, Scandinavian Journal of Work, Environment & Health, 28, 94–​108. Lindbohm, M. L., and Sallmén M. (n.d.), Reproductive Effects Caused by Chemical and Biological Agents (European Agency for Safety and Health at Work (EU OSHA): OSH-​ WIKI). https://​oshw​iki.eu/​wiki/​Reproductive_​effects_​caused_​by_​che​mica​l_​an​d_​bi​olog​ ical​_a​ ge​nts. Loomis, D., Guha, N., Hall, A., and Straif, K. (2018), ‘Identifying Occupational Carcinogens: An Update from the IARC Monographs’, Occupational and Environmental Medicine 75, 593–​603.

242    maria albin, chris mathieu, esa-pekka takala and töres theorell Magnusson Hanson, L. L., Westerlund, H., Chungkham, H. S., Vahtera, J., Rod, N. H., Alexanderson, K. et al. (2018), ‘Job Strain and Loss of Healthy Life Years between Ages 50 and 75 by Sex and Occupational Position: Analyses of 64 934 Individuals from Four Prospective Cohort Studies’, Occupational and Environmental Medicine. 75, 486–​493. Michie, S., and Williams, S. (2003), ‘Reducing Work Related Psychological Ill Health and Sickness Absence: A Systematic Literature Review’, Occupational and Environmental Medicine, 60, 3–​9. Montano, D., Hoven, H., and Siegrist J. (2014), ‘Effects of Organisational-​Level Interventions at Work on Employees’ Health: A Systematic Review’, BMC Public Health, 14, 135. Mutambudzi, M., Theorell, T., and Li, J. (2018), ‘Job Strain and Long-​Term Sickness Absence from Work: A Ten-​Year Prospective Study in German Working Population’, Journal of Occupational and Environmental Medicine (Dec 11), doi: 10.1097/​JOM.0000000000001525. NIOSH (1994), Applications Manual for the Revised NIOSH Lifting Equation. DHHS (NIOSH) Publication Number 94-​110, https://​www.cdc.gov/​niosh/​docs/​94-​110/​defa​ult.html. Pan, K. Y., Xu, W., Mangialasche, F., Dekhtyar, S., Fratiglioni, L., and Wang, H. X. (2018), ‘Working Life Psychosocial Conditions in Relation to Late-​ Life Cognitive Decline: A Population-​ Based Cohort Study’, Journal of Alzheimer’s Disease (Dec. 4) doi: 10.3233/​JAD-​180870. Quirce, S., Vandenplas, O., Campo, P., Cruz, M. J., de Blay, F., Koschel, D. et al. (2016), ‘Occupational Hypersensitivity Pneumonitis: An EAACI Position Paper’, Allergy 71/​6, 765–​779. Romanowska J., Larsson, G., Eriksson, M., Wikström, B. M., Westerlund, H., and Theorell, T. (2011), ‘Health Effects on Leaders and Co-​Workers of an Art-​Based Leadership Development Program’, Psychotherapy and Psychosomatics, 80, 78–​87. Rushton L. (2017), ‘The Global Burden of Occupational Disease’, Current Environmental Health 3, 340–​ 348, https://​www.ncbi.nlm.nih.gov/​pmc/​artic​les/​PMC​ Reports (September) 4/​ 5591​369/​. Sales, S. M., and House, J. (1971), ‘Job Dissatisfaction as a Possible Risk Factor in Coronary Heart Disease’, Journal of Chronic Diseases, 23/​12, 861–​873. Schenk L. (2010), ‘Comparison of Data Used for Setting Occupational Exposure Limits’, International Journal of Occupational and Environmental Health, 16/​3, 249–​262. Schulte P. A., Stephenson, C. M., Okun, A. H., Palassis, J., and Biddle, E. (2005), ‘Integrating Occupational Safety and Health Information into Vocational and Technical Education and Other Workforce Preparation Programs’, American Journal of Public Health, 95/​3, 404–​411. Shields, M. (2006), ‘Stress and Depression in the Employed Population’, Health Reports (Canada), 17, 11–​29. Siegrist J. (1996), ‘Adverse Health Effects of High-​Effort/​Low-​Reward Conditions’, Journal of Occupational Health Psychology, 1, 27–​41. Siegrist, J., and Wahrendorf, M. (2016), Work Stress and Health in a Globalized Economy. The Model of Effort-​Reward Imbalance Cham, Switzerland: Springer International Publishing. Sjögren, B., Bigert. C., and Gustavsson, P. (2020), ‘The Nordic Expert Group for Criteria Documentation from Chemicals. 153. Occupational Chemical Expostures and Cardiovascular Disease’, Arbete och Hälsa 54/​2, 1–​428. Available at http://​hdl.handle.net/​ 2077/​66225 Smith M. J., and Sainfort, P. (1989), ‘A Balance Theory of Job Design for Stress Reduction’, International Journal of Industrial Ergonomics, 4/​1, 67–​79.

the cornerstone of job quality     243 Stansfeld, S. A., Shipley, M. J., Head, J., and Fuhrer, R. (2012), ‘Repeated Job Strain and the Risk of Depression: Longitudinal Analyses from the Whitehall II Study’, American Journal of Public Health, 102, 2360–​2366. Takala, J. (2015),‘Eliminating Occupational Cancer’, Industrial Health, 53/​4, 307–​309. https://​ www.ncbi.nlm.nih.gov/​pmc/​artic​les/​PMC​4551​060/​. Theorell, T. (2012), ‘Stress Reduction Programmes for the Workplace’, in R. J. Gatchel, and I. Z. Schultz, eds., Handbook of Occupational Health and Wellness (Cham: Springer Verlag), 383–​403. Theorell, T., Brisson, C., and Vezina, M. (2015), ‘Psychosocial Factors in the Prevention of Cardiovascular Disease’, in S. Gielen, G. de Backer, M. F. Piepoli, and D. Wood, eds., The ESC Textbook of Preventive Cardiology (London: Oxford University Press) Theorell, T., Emdad, R., Arnetz, B., and Weingarten, A. (2001), ‘Employee Effects of an Educational Program for Managers at an Insurance Company’, Psychosomatic Medicine, 63, 724–​733. Theorell, T., Hammarström, A., Aronsson, G., Träskman Bendz, L., Grape, T., Hogstedt, C. et al. (2015), ‘A Systematic Review Including Meta-​Analysis of Work Environment and Depressive Symptoms’, BMC Public Health, 1/​15, 738. doi: 10.1186/​s12889-​015-​1954-​4. Theorell, T., Jood, K., Jarvholm, L. S., Vingård, E., Perk, J., Ostergren, P. O., and Hall, C. (2016), ‘A Systematic Review of Studies in the Contribution of the Work Environment to Ischemic Heart Disease Development’, BMC Public Health, 26, 470–​477. Torén K., and Blanc, P. D. (2009), ‘Asthma Caused by Occupational Exposures is Common: A Systematic Analysis of Estimates of the Population-​Attributable Fraction’, BMC Pulmonary Medicine, 29/​9, 7.

chapter 11

innovation and job quality rafael muñoz de bustillo, rafael grande and enrique fernández-​macías

Introduction Moral and business cases for improving job quality exist (Findlay et al. 2017). In these it is sometimes posited that job quality and innovation might have mutually beneficial effects. For example, a central premise of the European Commission’s Europe 2020 growth strategy was the possibility of stimulating the growth of high innovation, high job quality firms to create more and better jobs (EC, 2012). Nonetheless, the European Commission recognizes that more research is needed. This chapter explores the relationship between job quality and innovation. The chapter has two parts. The first part presents a stylized model highlighting potential bi-​directional links between innovation and job quality, going from innovation to job quality. This discussion also reflects on the potential impact of the new digital technologies on job quality. The second part of the chapter is empirical, exploring the relationship between job quality and innovation using individual data from the 2010 European Working Conditions Survey (EWCS). After briefly introducing the reader to the index of job quality that is used in the analysis as well as to the information regarding innovation included in the 2010 EWCS, the chapter presents descriptive statistics on the relationship between innovation and job quality at the EU and national level. In the same section, we explore in more detail, from a multivariate perspective, the role played by different variables, including technological and organizational innovations, in the determination of job quality. The final section reviews the major conclusions of the chapter and offers comments on potential future developments.

innovation and job quality    245

Technological innovation

Increase in productivity

Change in production processes

Structural change

Job quality Figure 11.1  Relations between technological innovation and job quality Source: Authors’ analysis

Technological Innovation and Job Quality: Transmission Mechanisms As can be seen in Figure 11.1, a priori there are four major transmission mechanisms relating innovation1 and job quality: (1) The first relates to the leading role played by innovation in the determination of productivity, which is in turn a key determinant of job quality. (2) The second mechanism is associated with the impact innovation has on the structure of production and employment, and the implications of such changes on job quality. (3) The third mechanism refers to the direct impact of different technological and organizational innovations on the working environment and the conditions of work, and the subsequent implications for job quality. 1  The chapter adopts a narrow definition of innovation, mostly centred in technological innovation. This focus omits analysis of the other types of innovation considered in the Oslo Manual (OECD 2005), that is, non-​technological innovation related to organizational and marketing changes. Two reasons lie behind this choice. First, there is an issue of data availability: —​the more robust data collected by the EU focuses on technological innovation. Second, despite the rhetoric otherwise, most policy focus within the EU—​nationally and at EC levelalso centres on technological innovation (Mako et al. 2016). We do, nonetheless, briefly address the issue of organizational innovation in the chapter.

246    rafael muñoz de bustillo et al 35,000 30,000

1990 Int. GK$

25,000 20,000 15,000 10,000

0

1870 1876 1882 1888 1894 1900 1906 1912 1918 1924 1930 1936 1942 1948 1954 1960 1966 1972 1978 1984 1990 1996 2002 2008

5,000

Europa (12)

USA

Figure 11.2  GDP per capita in Europe (EU-​12) and USA, 1870–​2010 Source: Author´s analysis from Bolt and van Zanden (2013)

(4) If the three previous mechanisms refer to the effect that productivity may have on job quality, the fourth looks at it the other way round: job quality acting as driver of innovation. Importantly, the above mechanisms cannot be considered in isolation. The relationship between technological innovation and job quality is mediated by a myriad of factors operating at different levels, from trade unions to labour market institutions. The adoption of particular technologies in production and the associated organizational changes are themselves affected by power relations in the workplace and by the wider socio-​economic context.2 However, such effects are beyond the scope of this chapter.

Technological Change, Productivity and Job Quality Increased productivity is the driving force behind the tremendous economic growth experienced by modern developed economies in the last two centuries. From 1870 to 2010 GDP per capita increased tenfold in 12 European countries and twelvefold in the US (Figure 11.2). Even in a context of intensive capital accumulation such an increase in GDP would have not been possible without the corresponding increase in labour productivity.

2 

For an analysis of how social relations affect innovations, see Noble (2011)

innovation and job quality    247 GDP per capita can be expressed as: GDP pc =



employment GDP GDP hours = x x (2) population hours employment population GDP pc = π j . j. e (3)



Where π j is hourly productivity, j is annual working hours and e is the employment rate (defined here as employment by population). We know that working hours have decreased since the 1800s across all developed countries. Taking the United States as an example, in 1870 the average American labourer worked approximately 63 hours per week (Whaples 1990), compared to 36.5 hours in 2010 (OECD labour statistics). The employment rate has followed a different path. According to the same sources, in 1870 total employment amounted to roughly one-​third of the population, while by 2010 it reached 45.5 per cent. This 40 per cent increase of the employment rate has not been strong enough to compensate for the decrease in working hours, leaving a compound negative impact on GDP pc of -​20 per cent. Therefore, we can say that the increase in GDP per capita in the period is fully explained by the rise in hourly productivity. Although there are many elements behind the rise in productivity (investment in fixed and human capital and the corresponding increase in the capital/​labour ratio, economies of scale, etc.), in the long run most of the increase in productivity is explained by technological change. This causal relation is buttressed by both the economic history literature and by the more specific area of Growth Accounting. From the former, we can quote, among many others, Joel Mokryl (1990), for whom technological change is the ‘lever of riches’, or similarly Kranzberg and Pursell (1967) who in their History of Technology argue that ‘by the end of the 19th century technology had shown its capacity to transform human condition and, in many cases, to contribute to its improvement’ (p. 825). From a different perspective, the methodology of Growth Accounting,3 concludes that for most countries and periods, Total Factor Productivity (related to technical change) explains a substantial part—​as much as half (if not more)—​of total growth (Easterly and Levine 2001).4 Once one acknowledges the role of technological innovation in the growth of nations, its relationship with job quality is fairly obvious. One effect of technical change, through productivity growth, is making compatible a reduction in working time with an increase in labour income. As working time and wages are two major components of job quality, 3 Growth

Accounting theory ascribes economic growth to changes in labour and capital input, following a Cobb-​Douglas production function, considering that the residual, that is, the part not explained by such changes—​Total Factor Productivity—​or, in the words of Abramowitz (1956), the ‘measure of our ignorance’ (p. 11)—​is the result of technical change. 4  For an updated account of the methodology of growth accounting from a theoretical perspective, see Hulten (2009) or Bosworth and Collins (2008), with application to India and China. An account of the allocation of growth to labour and capital inputs and productivity-​technical change for the EU and US since 1980 can be found in Timmer et al. (2003).

248    rafael muñoz de bustillo et al 5.0 y = 0.947x + 0.0637 R2 = 0.8331

Annual wage growth (1970-2006) (%)

4.0 Ireland

Japan

3.0

Iceland

France

Denmark

Spain Norway Germany

United Kingdom

2.0

Switzerland United States

Finland

Netherlands

Italy

Australia Canada

1.0 New Zealand 0.0 0.0

1.0

2.0

3.0

4.0

5.0

Anual growth rate of labour productivity 1970–2006 (%)

Figure 11.3  Wages and productivity growth. OECD countries, 1970–​2006 Source: Based on Sharpe et al. (2008: 49)

the improvement of job quality in these realms is directly related with the productivity dividend of technological innovation. It is far from our intention to argue that such relation was automatic and painless. The struggle for the eight-​hour working day was long and painful (Roediger and Foner 1998). But without an underlying increase in productivity it would have been much more difficult to reach the eight-​hour working day, as it would have required a direct and drastic redistribution from profits to wages. From equation (3) and the definition of factor distribution of income (the share of wages, t, and profits, b, from total output) we can say that the rate of change of the participation of profits in GDP is equal to:

 (4) b = j + π − w

where the dot above the variables indicates rate of change. From equation (4) we can see how a sufficiently high rate of productivity growth would allow for the squaring of the circle: a reduction of working hours and an increase in wages, without altering the factor distribution of income (that is, with a fixed share of profits from total income). Figure 11.3 shows the tight correlation between growth in productivity and wages.5 5  Wages are defined as labour compensation per hour worked, deflated with an output price deflator or product wages in the denomination used by Sharp et al. (2008).

innovation and job quality    249 Summing up, innovation and technological change, by increasing productivity, is the major source behind two of the major improvements in job quality in the past: the decrease in working time and the increase in wages.6 From this (partial) perspective, technological change is considered to be a progressive force in improving job quality.

Technology, Structural Change and the Structure of Employment A key feature of economic growth, known as structural change, is the changing importance of the different economic activities in terms of their contribution to GDP as countries transit from low-​to high-​income economies. Low-​income countries are characterized by having a large primary sector, absorbing more than two-​thirds of employment, while on the other end, high-​income economies are basically service economies with nearly 80 per cent of workers employed in tertiary activities. This ongoing process of structural change is explained first by the introduction of new technologies in agriculture and the corresponding emigration of rural workers to urban industrial and service sectors, and second, by the introduction of labour-​ saving technologies in manufacturing and the offshoring of many industrial activities (a process itself facilitated by innovation in the transportation and communication industries). The debate about the impact of these changes on job quality can profit from the distinction between technological changes leading to product innovation and those leading to process innovation.7 When the first type of technological driver is in place, the development of new products might lead to the disappearance of whole industries producing highly substitutive goods (and the jobs and skills related to them). Computers and the development of internet and digital switchboards are an interesting and recent example of technological change that implies the practical disappearance of different occupations in the recent past (Figure 11.4). In the second case, the development of new ways to produce old products, usually associated with the increase in capital/​labour ratios, will lead to a restructuring of employment, both within the sector where the process innovation is taking place, and in the capital goods sector that produces the new technologies. A good example of this

6 To

the extent that technological change (among many other important supply and demand variables) might have an impact on female labour force participation rates, technical change would have another indirect impact on job quality: the development of a new dimension of job quality related to work–​life balance issues. 7 Although following the Oslo manual we are fully aware of the existence of other types of innovation (marketing and organizational innovation), due to the nature and restrictions of the databases used in this chapter we focus mainly on technological innovation, that is, product and process innovations.

250    rafael muñoz de bustillo et al 140 120

Thousands

100 80 60 40 20 0 1871 1881 1891 1901 1911 1921 1931 1951 1961 1971 1981 1991 2001 2011

Figure 11.4  Telephone and telegraph operators in England and Wales Source: Stewart et al. (2015: 7)

H Hotels and restaurants

0.47

A-B Agriculture, hunting, forestry, fishing F Construction C-D Mining, quarrying, Manufacturing I Transport, storage and communication G Wholesale and retail trade; repair of motor... K Real estate activities M-N-O-P-Q Other services E Electricity, gas, and water supply L Public administration and defence;... J Financial intermediation

0.71 0.00

0.10 JQI 2005

0.20

0.30

0.40

0.50

0.60

0.70

0.80

JQI 2010

Figure 11.5  Job quality by major sectors of economic activity; JQI 2005 and 2010, EU-​15 Source: Authors’ analysis from EWCS 2010 microdata

type of process is the automation of warehouses (Chang 2014), or the use of robots in industrial processes of welding and painting, now common in the automobile industry. The impact of technologically driven structural change on job quality in the case of product innovation depends on the job quality of the jobs in the falling and rising industries, while the impact of the second type of change, process innovation, will depend on the type of jobs or tasks subject to substitution by machines and the characteristics of the jobs in the capital goods industry producing the new technologies. Regarding the first issue, Figure 11.5 reproduces the Job Quality Index (JQI) developed by Muñoz de Bustillo et al. (2011) by broad category of economic activity

innovation and job quality    251 for the EU-​15 in 2010. As can be seen, there is a sizeable difference in job quality by sectors, with JQI ranging from 0.47 in Hotels and Restaurants to 0.71 in Financial Intermediation. Any technological-​driven reshuffling of sectors in terms of their relative importance in the economy will have implications in terms of overall job quality. The second issue, the impact of technological change leading to the automatization and robotization of production on job quality, has been the centre of an important debate about the present and future changes in the structure of employment. Depending on the type of jobs substituted by machines, their position in the distribution of jobs according to their quality, technological change (process innovation) might have different implications in terms of job quality. There are two major (and consecutive) hypotheses that dominate the debate regarding this issue. The first, known as Skilled Biased Technical Change (SBTC) and related to the original work of Levy and Murnane (1992), argues that recent changes in the technology of production favour skilled workers over unskilled workers, leading to increased demand for more skilled workers and increased wage inequality (Violante 2008). The second hypothesis, Routine Biased Technical Change (RBTC), argues that the last wave of technical change (more specifically, the application in the 1990s of innovations derived from the IT Revolution of the late 1970s and 1980s) has a strong polarizing impact on the employment structures of most advanced capitalist economies (see Autor et al. 2006; Goos and Manning 2007; Goos et al. 2009; Acemoglu and Autor 2011; see also Goos et al. in the chapter entitled ‘Job Polarization: Its History, an Intuitive Framework and Some Empirical Evidence’ in this volume ). This hypothesis posits that the application of this new IT to production, particularly, routine tasks, tends to substitute labour in the middle of the skills/​wage structure, while simultaneously expanding demand for labour at the top and bottom in high-​and low-​skilled non-​routine activities. The result across all high-​income countries is a hollowing out of the jobs in the middle of the wage distribution, and the corresponding polarization of labour markets in terms of wages, which is an important component of job quality. The polarization thesis as it has become known has been empirically contested by a number of European and US papers (Fernández-​Macías 2012; Oesch 2013; Mishel et al. 2013; Fernández-​Macías and Hurley 2017; Cirillo 2018; Oesch and Piccitto 2019, Hunt and Nunn 2022). This body of work defends the existence of diverse patterns of change in the structure of employment—​job quality upgrading, downgrading and polarization—​arguing that such plurality can be better understood by a more open approximation to structural employment change, in which not only technology but also institutions and labour supply shape the pattern of employment and how it changes over time. It is beyond the scope of this chapter to explore the subtleties of this debate. What can be said here is that the change in the structure of employment is explained by the combination of technical change within sectors, and structural change. Such changes do not operate in a vacuum: institutional and supply side factors also play a part and can make a difference in terms of the final impact of technological change on employment structure.

252    rafael muñoz de bustillo et al 0.8

y = –0.2965x + 0.6514 R2 = 0.5762

Job quality index

0.7 0.6 0.5 0.4 0.3 0.2 0

0.2

0.4

0.6

0.8

1

Probability of computerization

Figure 11.6  Probability of computerization and JQI in 39 sectors of economic activity (*) See Appendix 1 for the data of the specific sectors Source: Author’s analysis from Frey and Osborne (2013) and EWCS microdata

The combination of the recent work of Frey and Osborne (2013) estimating the probability of computerization for 702 detailed occupations and our JQI for Europe offer a glimpse into the direct impact of such changes on future job quality by identifying whether the occupations with higher probability of computerization are characterized now by high or low job quality according to our JQI.8 Figure 11.6, reproducing the JQI and the probability of computerization of 39 sectors of activity, shows a clear inverse relation between computerization risk and job quality, according to which those jobs with lower quality face a higher probability of being substituted in the future by machines. In this respect, at least in a partial equilibrium context, and if Frey and Osborne are right regarding the different probabilities of computerization of the occupations reviewed in their research,9 the new wave of technological innovation would result in an increase in the average job quality by reducing the number of low-​quality jobs through the substitution of labour by capital. It is a separate issue of course as to whether the workers no longer demanded in those jobs increasingly computerizated will find employment in other sectors of the economy and what the quality of their new jobs will be.

8  To do this analysis, the 702 occupations analysed by Frey and Osborne (2013) based on the US SOC classification were converted into 39 ISCO occupations. 9  Arntz et al. (2016) argue that the approach followed by Frey and Osborne (2013) in which whole occupations are automated leads to higher risk of automatization compared to when analysis focuses on tasks potentially automatized because high-​risk occupations often contain a substantial share of tasks that are hard to automate. When, as Arntz et al. (2016) do, the focus is on tasks and not whole occupation, the resulting number of jobs at risk is much more modest: around 9 per cent of jobs for a sample of 21 OECD countries compared to the estimate of Frey and Osborne (2013) about 47 per cent of total employment for the US.

innovation and job quality    253

The Microeconomic Impact: Technological Change and Job Quality at the Job Level The previous section outlined two mechanisms of transmission from technological change to job quality that work at the aggregate or macro level. It showed how those mechanisms affect job quality either by the surplus of time or output made possible by the increase in productivity or by changes in the composition of employment between sectors and firms and within firms. In contrast, this section focuses on how changes in technology affect the process of work itself, more specifically, what workers do and the environment in which they do it. Technological change, while improving some aspects of work, might generate new risks and lead to a deterioration of other aspects of job quality. An obvious illustration of this point is to examine the impact of technological change on the most fundamental working condition: the security of one’s own life while at work (see Albin et al. in the chapter entitled ‘The Cornerstone of Job Quality: Occupational Safety and Health’ in this volume). Taking the US as an example, in the last four decades (1970–​ 2011) the national fatality rate (work-​ related deaths per 100,000 workers) decreased by 80 per cent, from 18 to 3.5 (AFL-​CIO, 2013).10 This reduction is the result of many interrelated factors, not just micro-​level technical change. These other factors include structural change: agriculture and mining have fatality rates more than six times higher than the average, and construction more than twice as high, while the fatality rate in education is less than 20 per cent of the average. The development of stricter health and safety regulations is another major factor affecting the reduction in fatality rates. However, these other factors do not mean that technical change plays a marginal role in the reduction of fatality rates. A good example of the role of technological innovation in the reduction of fatality rates is the mining industry in the US. Although the development of new machinery also introduced some new hazards (MMWR 1999), the result of technological change in terms of lower casualties in the sector reproduced in Figure 11.7 below is telling. Technological innovation can also deteriorate working conditions, creating new risks. One of the first (if not the first) economist to draw attention to the negative impact of technological change on job quality was Adam Smith (1776/​1981): In the progress of the division of labour, the employment of the far greater part of those who live by labour, that is, of the great body of the people, comes to be confined to a few very simple operations, frequently to one or two. . . . The man whose whole life is spent in performing a few simple operations, of which the effects are perhaps always the same, or very nearly the same, has no occasion to exert his understanding or to exercise his invention in finding out expedients for removing difficulties which never occur. He naturally loses, therefore, the habit of such exertion, and

10 

The figure for 2011 is calculated in terms of total hours (AFL-​CIO 2013: 41–​42).

254    rafael muñoz de bustillo et al 700

Permissible Explosives + Federal Coal Mine Health & Safety Act

U.S. Bureau of Mines Established

600

Rock Dusting First Used ł Permissible Electrical Equipment First Used

500

×

Rock Dust-Coal Dust Analazer

Permissible Electrical Cap Lamp First Used ¶

400

300

Methane Degasificatior łł

× ×

×

Explosion-Proff Bulkheads¶¶

× Improved ventilation **

200

×

100

×

×

×

×

×

× ×

0 1900

1910

1920

1930

1940

1950

1960

× 1970

×

× 1980

×

× 1990

Figure 11.7  Five-​year averages of annual number of deaths related to coal mine explosions in the United States, 1901–​1995* * Each X represents the 5 years average of the number of deaths related to coal mine explosions; the line is a smoothed regression line through the 5 years average. +​Explosives and equipment that can be used in an explosive methane-​rich environment without causing a methane explosion. ł The process of applying a layer of rock dust over the coal dust, which creates an inert mixture and inhibits a coal dust explosion. Lamp worn on miners’ caps. ** Ventilation improvements, including the use of reversible fans, reduce the concentration of methane and remove the explosion gas from the mine. +​+​A hand-​held monitor that provides instantaneous readings of the rock-​to-​coal dust mixture to ensure that it is inert. łł Techniques to remove methane from the coal before mining the coal. Explosion proof walls used to seal abandoned (mined-​out) areas to protect workers in active parts of the mine. Source: MMWR (1999).

generally becomes as stupid and ignorant as it is possible for a human creature to become. (n.p.)

While Smith referred to the impact of the first industrial revolution on the nature of work, his words remain relevant for later technology-​driven advances in work organization that deepen the division of labour, and increase managerial control and standardization of processes, thus reducing workers´ autonomy and skills. Braverman (1974), most obviously, showed how Taylorist process innovation resulted in a degradation of work over the first half of the twentieth century. Even in the new world of the digital

innovation and job quality    255 economy, the words of Smith resonate with some emerging forms of work such as digital labour platforms. In these, the use of algorithmic control via digital devices allows an extreme division of labour, expanded surveillance and control, and a significant centralization of decision-​making in production processes. Although the implications for job quality can vary quite significantly depending on the kind of tasks and skills required, there is growing evidence that in many cases it can be negative (Pesole et al. 2018; De Groen et al. 2018; Berg et al. 2018).11 As a ‘vector of change’, it has to be acknowledged that platforms affect the coordination of work and so are more an innovation in work organization than a technical innovation. Beyond this vector, two other vectors are apparent in the new digital economy —​automation and digitization (Fernández-​Macías 2018)—​and, as with platforms, both might also affect job quality. Automation affects productivity and the composition of employment because it can replace labour, though which tasks are more susceptible to substitution is delineated by the limits of automation and the type of tasks that comprise any job; digitization affects both productivity and the intrinsic nature of work because it can render parts of the physical production process into digital information and vice versa, impacting workers’ control and privacy.

Job Quality as a Driver of Innovation Job quality can thus be considered (at least partly) as the outcome of technological change. In this section we will radically reverse this perspective, exploring whether job quality might be one among the many variables affecting innovation. At issue is whether good (or bad) jobs lead to faster technological change, and whether job quality is a factor affecting the pace of innovation. There are two different mechanisms that could explain the existence of a positive relation between job quality and innovation from this perspective. The first builds on the role played by job quality in incentivizing productivity. The theory of efficiency wages (Akerloff and Yellen 1986) is probably one of the best-​known approaches relating job quality (specifically its wage dimension) to improvements in output. The theory of efficiency wages argues that higher wages lead to higher productivity through different mechanisms that go from the higher identification of the workers with the firm and the intensification of their effort at work to improving morale, reducing turn-​over and attracting better workers, etc. From this perspective, better working conditions pay for themselves through higher labour productivity (Raff and Summers 1987). The same argument could be applied directly to the generation of innovation: better wages might lead directly to more innovation. Moreover, good working conditions and higher identification of workers with the goals of the firm might incentive small innovations at the plant level. In fact, firms often have specific incentive programmes to facilitate worker involvement in organizational and technological innovations aimed at improving productivity. From the same perspective, it could be argued that good employment conditions (one of the dimensions of job quality), and specifically job security,

11 

For an review of safety and health risk related to digitalization, see EU-​OSHA (2018).

256    rafael muñoz de bustillo et al can make workers less innovation adverse as their jobs will be protected from the potential negative impact of innovation on employment. More recently, the literature on High Performance Work Systems (HPWS) extols the positive impact on firm performance of new non-​hierarchical management systems. According to Pfeffer (1988), HPWS are characterized by seven key elements: (a) employment security; (b) selective hiring of new personnel; (c) self-​managed teams and decentralization of decision-​making as the basic principles of organizational design; (d) comparatively high compensation contingent on organizational performance; (e) extensive training; (f) reduced status distinctions and barriers, including dress, language, office arrangements, and wage differences across levels; (g) extensive sharing of financial and performance information throughout the organization. These elements create ‘an organization based on employee involvement, commitment and empowerment, not employee control’ (Tomer 2001: 64). Many of the constitutive elements of HPWS are positively related with job quality according to our model: autonomy, employment stability, wage, participation, etc. Although most of the empirical analysis of HPWSs has focused on its impact on firms and workers in the short run, with conflicting results,12 there is a number of studies that look at its effect on innovation, for example Hefferenan et al. (2008), Harden et al. (2006), Flood et al. (2008) and Fu et al. (2015) which find a positive relationship between HPWS and innovation. The second mechanism relating job quality and innovation is completely different. From this perspective, good working conditions translate into higher unit labour costs for the firm (not all the increase in labour cost is compensated by increases in productivity), putting pressure on firms to increase productivity through innovation. This second mechanism can exist in countries with high levels of trade union participation rates and labour leverage high enough to set good working standards across firms and industries (such as the Nordic states of the EU). By not allowing working conditions to be tailored to the specificities of low-​productivity firms, this policy blocks low-​ productivity firms from the market, improving working conditions and acting at the same time as a powerful incentive to increase productivity through innovation as the only road to increase firm survival rates and profitability. Summing up, the review of the existing literature on the relationship between innovation and job quality shows the complexity of the nexus between both variables. On the one hand, innovation, working at different levels, has had a clear positive impact on job quality. On the other hand, job quality can affect innovation by improving the identification of employees with the goals of the firm and increasing their cooperation, both at the level of generating and introducing innovations.

12  As Boxall and Macky (2009) point out, some authors argued that HPWS benefit both workers and firms (Appelbaum et al. 2000), others question the gains for firms (Cappelli and Neumark 2001; Way 2002) or for workers (e.g. White et al. 2003), while others question the benefits for both parties (e.g. Godard 2004). For a survey of the literature see Heffernan et al. (2011) or Kaushik and Mukherjee (2021).

innovation and job quality    257 With this framework of reference, the next section explores the interactions between job quality and innovation using the EWCS 2010. In addition to providing rich data on job quality, this survey included a question about innovation at the level of the firm.

Job Quality and Innovation: Data and Results The Data The EWCS is the most important source of information about working conditions at the European level. One key advantage of the EWCS compared to other surveys is that it is funded, designed and coordinated centrally by the European Foundation for the Improvement of Living and Working Conditions (Eurofound).13 This ensures a higher level of country comparability than for other European labour market surveys. Another important advantage of the EWCS is a high degree of transparency and documentation of the whole research process. The sample of the EWCS is representative of all persons in employment in private households in all EU member states (and some European non-​ member States, for example Turkey, Norway and Albania). The fieldwork procedures follow the same principles across Europe: in all countries, the sample is stratified by region and size of settlement, and the interviews are clustered by geographic proximity. The selection of households is based on the random-​walk method, and within the selected household one employed individual is randomly selected. The size of the sample for the EWCS is 1,000 cases per country, with some exceptions.14 This is also the main problem of the EWCS. It allows for analysis at the national level but is too small to go deeper and break down the results within countries by gender, sector, occupation or other variables as the number of cases becomes too small and the estimation unreliable. The analysis of this chapter is restricted to the EU-​15 states to allow for a reasonable manageability and interpretation of the results.15 The EWCS is extremely rich in terms of information about working conditions, allowing the calculation of a multidimensional aggregate index of job quality. The EWCS also includes (for the first and last time) two questions about innovation (technological and organizational innovation), allowing the exploration of the relationship between job quality and innovation, though analysis is at a very basic level due to the bluntness of the information gathered on innovation. The measure of job quality presented in this chapter draws on the model of job quality developed by Muñoz de Bustillo et al. (2011) and Fernández-​Macias et al. (2016). The 13 

The Eurofound is an EU agency tasked with improving living and working conditions. The total number of interviews across all participating countries in 2010 was 43,816. 15  More details of the EWCS, see http://​www.eurofo​und.eur​opa.eu/​surv​eys/​ewcs/​index.htm). 14 

258    rafael muñoz de bustillo et al Table 11.1 Index of job quality Dimension

Variables and questions

1. Pay

– ​Gross monthly wage in Power Purchasing Parity (20%)

2.  Intrinsic quality of work (20%)

– ​Skills (6.6%) [ISCO, q49d, q49e, q49f] – ​Autonomy (6.6%) [q25a, q50b, q50c, q49b] – ​Social support (6.6%) q51a]

3.  Employment quality (20%)

–  Contractual stability (10%) [q6 q7 q12] –  Development opportunities (10%) [q61a, q77c]

4.  Workplace risks (20%)

–  Physical risks (20%); [q23a-​g, 24a, q24c, q24e]

5. Working time and work–​life balance (20%)

–  Duration (6.6%); [q18] –  Scheduling (6.6%); [q32, q33, q34, q35] –  Intensity (6.6%); [q45a, q45b]

Note: The weights of the items and the question number of the EWCS dealing with the item are shown between brackets. Source: Authors’ elaboration from EWCS 2010

proposed Job Quality Index (JQI) is composed of five different dimensions: (1) pay, (2) intrinsic quality of work, (3) employment quality, (4) health and safety, (5) work–​ life balance. In the baseline formulation of the JQI, each dimension receives the same weight (20%) and the aggregation is carried out using a weighted geometric average. The sensitivity of this weighting scheme is assessed in Muñoz de Bustillo et al. (2011), finding that the rank correlation of country results obtained using alternative systems of weights is remarkably high, with very few changes in the ordering of the countries when the weights are adjusted using different formulas. This assessment provides evidence of the robustness of the JQI for international comparisons. The score of each of the five dimensions is computed using an arithmetic average of the values of its lower-​ level components, weighted according to the values shown in Table 11.1, while the aggregate index is obtained using the geometric mean of the five dimensions, as previously mentioned. Formally, for a certain individual i, the JQI responds to the following formula: 5

JQIi = ∏ Xij



15

j =1

where Xij denotes the score received by dimension j for the individual i. Each dimension takes a value between 0 and 100.16

16 The

standardization of the original variables to a 0-​100 scale was carried out according to a normative logic, as explained in Muñoz de Bustillo et al. (2011: 153–​154).

innovation and job quality    259 The JQI exhibits two advantages. First, its tree-​like design enables an aggregate final single job quality indicator without jeopardizing the possibility of studying the role played by the different dimensions, components and sub-​components of the index in its overall value. Second, the JQI is constructed at the level of the individual worker, which allows evaluation of the complementarily or substitution of attributes in the same job and computing the JQI for any group of specific workers (women, youth, etc.) or, in general, measures of dispersion (inequality) of job quality. Other key features of the JQI are the emphasis on results (rather than procedures), the grounding of the model in a detailed discussion of the specialized literature in the traditions of the social and health sciences and the focus on objective (rather than subjective) elements of job quality. In relation to the last item, although there is a large literature exploring job quality from the subjective perspective of the worker, or work satisfaction (see Knox and Wright in the chapter entitled ‘Understanding Job Quality Using Qualitative Research’ in this volume), the JQI focuses (whenever possible) on the objective elements of the job in order to be able to have a single metric of job quality independent of workers’ preferences and characteristics. Table 11.1 presents a summary of the dimensions considered in the Job Quality Index, JQI, used in the rest of the chapter to measure job quality, as well as the questions in the EWCS used to construct the different dimensions. The analysis of innovation is more crude due to the limitations of the EWCS. The EWCS only supplies binary information regarding whether the firm has (or has not) innovated in the last three years (and only in the 2010 wave of the survey), differentiating between Technological Innovation, TI (‘New processes or technologies were introduced’) and Organizational Innovation, OI (‘Substantial restructuring or reorganization was carried out’).17 Although in the previous section we limited our analysis to the different mechanisms relating technological innovation and job quality without mentioning organizational innovation, we have included organizational change in our empirical analysis, as the introduction of new technology is often accompanied by the introduction of innovations in the organization of the firm (the correlation index is 0.90). In order to better grasp this interrelation, we have also considered a stricter indicator of innovation, applied to those firms answering positively to the question of both technological and organizational innovation (Full Innovation, FI). Table 11.2 reproduces these Indexes of Innovation for the EU (27) (percentage of workers declaring that new processes or technologies were introduced). Before proceeding with the analysis of the correlation between job quality and innovation it is important to test to what extent the (simple) indicators of innovation that we use in the analysis offer a reasonable account of the innovation activity of EU firms.18 In 17  We are aware that the way the question is posed in the survey covers any organizational change, and in that sense includes changes such as downsizing, for example, that can be considered out of the scope of what is usually considered organizational innovation, see OECD (2005). 18  The reliability of the JQI was tested in Muñoz de Bustillo et al. (2011).

260    rafael muñoz de bustillo et al Table 11.2 Index of innovation. EU-​27, 2010 New processes or technologies were introduced (TI)

Substantial restructuring or reorganization was carried out (OI)

Both (FI)

Sweden

57,9

53,1

35,0

Finland

55,2

52,0

37,3

Denmark

54,0

50,1

34,3

Malta

48,4

42,4

32,4

Cyprus

47,3

39,6

33,5

United Kingdom

46,7

38,8

30,0

Luxembourg

46,2

34,0

24,9

Netherlands

45,8

37,4

25,0

Belgium

41,9

30,9

19,7

Slovakia

41,4

32,9

22,8

Ireland

41,4

35,4

26,0

Germany

41,3

30,5

22,9

Austria

41,2

30,5

24,1

Latvia

41,0

38,3

22,1

Estonia

39,8

39,0

24,5

39,37

32,66

22,18

Slovenia

38,3

29,5

18,9

Lithuania

37,8

28,1

18,2

Czech Republic

36,3

33,7

20,4

Portugal

35,9

27,1

21,5

Spain

35,1

23,4

14,4

Hungary

33,3

26,0

15,3

Italy

32,8

23,7

16,0

France

32,6

31,9

21,2

Greece

27,8

23,0

18,5

Romania

27,2

29,9

12,2

Poland

26,0

17,4

9,9

Bulgaria

19,6

20,5

12,0

EU27

Source: Authors´ analysis from EWCS 2010 microdata (Q15)

innovation and job quality    261 order to do so, we check whether the picture of innovation activity in European firms according to our innovation index, TI, is similar to that obtained using other more complex instruments such as the EU Summary Innovation Index, SII. Before presenting the results of this comparison, it is important to acknowledge that the aims of both indexes are different, and so we should not expect a perfect match between the results. The SII is a synthetic index of innovation constructed by the aggregation of 25 different indicators, allocated in eight different dimensions and grouped in three blocks: enablers (main drivers of innovation performance external to the firm), firm activities (innovation efforts at the level of the firm) and outputs (effects of firms’ innovation activities). The resulting indicator, built using secondary aggregated data from different sources, is relatively complex and covers many different aspects of innovation activity, from public expenditure in R&D to international scientific publications to new doctoral graduates. In contrast, the indicator of innovation that we use in the analysis of the interrelation of job quality and innovation is more parsimonious. Nevertheless, as Figure 11.8 shows, there is a relatively high correlation between the Technical Innovation Index built from the EWCS microdata and EU Summary Innovation Index. The correlation index between

70

SW

60 FI MT

Technical innovation index %

50 LV

NL

SK

EE

40 LT

HU RO

30

PL

DK

CY

PT ES

SL

AT

UK IE

LU

BE

DE

CZ FR

IT EL

BG

20

10

0 0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Summary innovation index

Figure 11.8  Summary Innovation Index and Technical Innovation Index from the EWCS in the EU-​27, 2010 Source: Author’s analysis from Maastricht Economic and Social Research and Training Centre on Innovation and Technology (UNU-​MERIT) (2011) and EWCS 2010 microdata

262    rafael muñoz de bustillo et al the two indicators is 0,671 when using the global SII, and 0.754 when using only the firm activity dimension. As a first approach to understanding the relationship between job quality and innovation, Table 11.3 presents the correlation coefficient between the Job Quality Index and Technological Innovation and Organizational Innovation Index by sector using individual-​level data from EWCS 2010. From these results we can say that: (1) technological innovation has a stronger relationship with the quality of jobs than organizational innovation (in interpreting this result it is important to be cautious due to the lower quality of the indicator used as it includes all kind of organizational changes); (2) the relationship between innovation and job quality is contingent on the type of activity, being higher in ‘Water supply, sewerage, waste management and remediation activities’, ‘Administrative and support service activities’, ‘Transportation and storage’ or ‘Manufacturing’.

Job Quality and Innovation at the Country Level To evaluate the existence, direction and intensity of the relationship between innovation and job quality in the EU-​15 we start by exploring the relationship between job quality and technical innovation at the country level. As the results reproduced in Figure 11.8 show, at least at the country level, there is a direct relationship between technological innovation and job quality. Furthermore, this correlation is moderately strong with an R2 of 0.336. In order to further understand the nature of this relationship, it is important to explore the extent that it operates through all the dimensions of the JQI. As Figure 11.8 shows, although the five dimensions of the JQI show a direct relationship to technological innovation, it is clear that the intensity of this relationship is very different for each of them. Technological innovation at the country level has a very weak correlation with the wage, probably because wages are more affected by other factors such as seniority, occupation etc. in addition, the correlation between technological innovation and work–​life balance being particularly strong is probably affected at country level by other factors such as labour regulations and public policies as well as the sectoral composition of the economy. However, the correlation between innovation and job quality is high for the dimensions of intrinsic quality of work and quality of employment, with an R-​squared of 0.81 and 0.85 respectively. Observing the location of different EU-​15 countries in terms of the relationship between innovation and job quality in Figure 11.9, we can conclude that there is a clear divide by region, a kind of regional clustering within the EU. On the one hand, the Mediterranean countries, including France, are located in the lower left corner of the graphic—​that is, they have a lower percentage of employees in establishments that have developed technological innovation in the last three years and have a lower quality of employment. Located in the opposite, top right corner are the Nordic countries—​ Denmark, Sweden and Finland—​with high technological innovation and high job quality. For their part, the countries of Central Europe, and the UK and Ireland, are

innovation and job quality    263 Table 11.3 Pearson correlation coefficient among Job Quality Index and index of innovation by sector, EU(15), 2010 Substantial restructuring New processes or or technologies reorganization was carried were introduced (TI) out (OI)

Both Technological and Organizational Innovation were introduced n

EU15

0.20

0.13

0.14

22,136

Agriculture, forestry and fishing

0.12

0.09

0.09

644

Mining and quarrying *

0.36

0.22

0.23

58

Manufacturing

0.20

0.14

0.14

2,578

Electricity, gas, steam and air conditioning supply

0.14

0.13

0.11

196

Water supply; sewerage, waste management

0.36

0.11

0.12

131

Construction

0.13

0.10

0.07

1,522

Wholesale and retail trade; repair of motor

0.11

0.07

0.07

3,600

Transportation and storage

0.21

0.15

0.17

1,097

Accommodation and food services

0.14

0.10

0.10

1,148

Information and communication

0.15

0.11

0.08

579

Financial and insurance activities

0.14

0.10

0.10

753

Real estate activities

0.09

0.07

0.12

197

Professional, scientific and technical

0.10

0.04

0.01

946

Administrative and support service ac

0.23

0.19

0.17

908

Public administration and defense; compulsory

0.14

0.06

0.05

1,382

Education

0.16

0.06

0.09

1,834

Human health and social work activities

0.14

0.04

0.09

2,765

Arts, entertainment and recreation

0.06

–​0.03

–​0.02

448

Other service activities

0.10

0.08

0.06

843

Activities of households

0.14

0.02

0.03

456

* Less than 100 cases Source: Authors´ analysis from EWCS 2010 microdata

264    rafael muñoz de bustillo et al Job Quality Index

Dimension 0 Wage .5

NL

IE

IT

.5

FR

DK GB

.4 FI

SE

DE

ES

.45

FR IT

.3

DK

.8

IE

DK

ES DE

FI SE

GB

.2

.4

FI NL

.75

DE

.7

GR

GR IT

GR

.1 30

40

50

60

30

40

50

Dimension 2 Employment quality .75

DK

IE

.5

IT

FR

ES PT

IE

.7 BE IT

.65

NL FR

.65

DE FI SE LU

ES

60

Dimension 4 Work-life balance

GB

AT

.6

.45

50

.7 DK

Job Quality Index

.55

GB LU

40

Technological Innovation Index %

NL

FI SE

BE DE

30

Dimension 3 Health and safety

.6 AT NL

60

Technological Innovation Index %

.65

FRES PT

.65

PT

Technological Innovation Index %

SE

LU GB

IE BEAT

PT

.35

Job Quality Index

LU

BEAT

LU

PT

Job Quality Index

Job Quality Index

AT

Job Quality Index

BE

.55

Dimension 1 Intrinsic quality of work .85

NL

Job Quality Index

.6

IT

PT

DK

BE

ES AT DE IE

LU GB

FI SE

.6

.55 FR GR

.55

GR

.4 30

40

50

60

Technological Innovation Index %

.5

GR

30

40

50

Technological Innovation Index %

60

30

40

50

60

Technological Innovation Index %

Figure 11.9 Innovation: new processes or technologies introduced in current workplace during the last 3 years by Job Quality Index. EU15 countries, EWCS 2010 JQI R2 =​0.366; Dimension 0 R2 =​0.069; Dimension 1 R2 =​0.812; Dimension 2 R2 =​0.855; Dimension 3 R2 =​0.348; Dimension 1 R2 =​0.187. Source: author’s elaboration from EWCS 2010

located between the two extremes. Again, we note that the differences between countries are more marked regarding the dimensions of intrinsic quality of work, quality of employment and health and safety.

Models of Job Quality and Innovation According to the results presented in the previous section, there is a positive association between job quality, as measured by the JQI, and innovation (technological and organizational) at the country level. To further advance understanding of the relationship between the two variables in this section we will examine if the correlation between innovation and job quality remains after controlling for the effects of other factors affecting job quality. With that aim, we present three different models of linear regression with JQI as the dependent variable and a dummy variable representing technological innovation and organizational innovation, together with different variables related to job quality as the independent variables. In Model 1 we introduce as control variables socio-​demographic

innovation and job quality    265 items (sex, age, education and country). In Model 2 we introduce variables related to the company and type of economic activity (size and sector). Finally, in Model 3 we introduce occupation as a control variable and simultaneously we reinforce our technological innovation variable with two new variables dealing with the level of digitalization of the job (probably the paradigmatic element of technological innovation in the twenty-​first century): use of computers and use of the internet/​email by the workers. The strong association between use of computers and occupation makes it advisable to include both variables simultaneously in the same model. The regression results confirm the existence of a significant and strong correlation between job quality and technological innovation (both the direct variable of TI and the indirect variables of working with computers and using internet), now at an individual level for the EU-​15.19 However, while technological innovation is positively correlated with the JQI in all three models, organizational innovation has no significant effect on job quality when we control for industry or occupation. Overall, Model 3 accounts for nearly 40 per cent of variation in job quality among jobs (R2 =​0.394).

Conclusion In this chapter, we discussed the four mechanisms that link innovation and job quality. First, by increasing labour productivity, innovation can expand the remuneration of labour and/​or reduce working time. Second, by changing the amount of labour necessary for different types of economic activity, innovation alters the composition of employment in terms of sector and occupation, with significant implications for job quality. Third, by changing the nature of economic activity and the labour process, innovation transforms the nature of work in ways which can significantly affect job quality. Fourth, job quality can be itself a driver of innovation, closing a feedback loop that makes the relationship between innovation and job quality even more important in the long run. As economic historians have argued (Freeman and Louca 2001; Pérez 2003), it generally takes decades for technological revolutions to fundamentally transform the nature and structure of economic activity. Historically, and as a result of the four mechanisms, innovation and technical change have tended to have a positive impact on job quality. First, because innovation and technical change have massively expanded the productivity of labour, thus facilitating a significant increase in real wages with a simultaneous equally significant decrease in working hours. Second, innovation and technical change have reduced drastically (via automation) the amount of labour necessary for the most physically intense and routine tasks, while expanding the amount of labour necessary for intellectual (information processing and problem solving) and social tasks. Third, 19  This result is in line with the result obtained by Grande et al. (2020) based on EU firm level data from the Third European Company Survey, according to which job quality and process innovation have a clear positive association at the firm level.

266    rafael muñoz de bustillo et al Table 11.4 Determinants of job quality (linear regression), EU-​15, 2010 Model 1 Coef.

Model 2 Sig.

Coef.

Technological Innovation (Yes)

0.044

***

0.037

Organizational Innovation (Yes)

0.007

*

0.004

Female (male) Age Age-​squared Education

***

Coef. 0.017

Sig. ***

–​0.006

–​0.043

***

–​0.048

***

–​0.043

***

0.014

***

0.013

***

0.012

***

0.000

***

0.000

***

0.000

***

–​0.083

***

–​0.081

***

–​0.047

***

Secondary

Ref. 0.143

Ref. ***

0.126

Ref. ***

0.058

***

Belgium

0.074

***

0.069

***

0.063

***

Denmark

0.057

***

0.055

***

0.034

***

Germany

Ref.

Greece

–​0.093

Spain

–​0.004

France

0.011

Ref. *** *

–​0.092

Ref. ***

–​0.091

–​0.005

–​0.006

0.006

0.004

***

Ireland

0.035

***

0.028

***

0.014

*

Italy

0.078

***

0.073

***

0.058

***

Luxembourg

0.065

***

0.048

***

0.035

***

Netherlands

0.117

***

0.111

***

0.088

***

***

0.097

***

0.079

***

Austria

0.100

Portugal

–​0.006

Finland

0.004

Sweden

0.028

***

United Kingdom

0.033

***

Establishment 1 Size 2–​9 10–​49

Sector (Nace codes)

Sig.

Primary Tertiary

Country

Model 3

–​0.008

–​0.018

0.006 0.020

*

0.007 ***

–​0.009

0.023

***

0.012

*

0.029

***

0.029

***

–​0.005

–​0.011

***

Ref.

Ref.

50–​249

0.014

***

0.015

***

> 250

0.049

***

0.039

***

Agriculture

0.006

Industrial

Ref.

Construction

–​0.015

**

Wholesale, retail, food and accommodation –​0.006 Transport Financial services

–​0.006 0.101

*** (continued)

innovation and job quality    267 Table 11.4  (Continued) Model 1 Coef.

Occupations (Isco-​08)

Model 2 Sig.

Coef.

Model 3 Sig.

Public administration and defence

0.058

***

Education

0.054

***

Health

0.010

*

Other services

0.013

**

Armed forces occupations

Coef.

Sig.

–​0.033 Ref.

Managers Professionals

–​0.002

Technicians and associate professionals

–​0.036

***

Clerical support workers

–​0.087

***

Service and sales workers

–​0.104

***

Skilled agricultural, forestry and fi..

–​0.069

***

Craft and related trades workers

–​0.109

***

Plant and machine operators, and asse..

–​0.122

***

Elementary occupations

–​0.170

***

Working with computers

0.037

***

Using internet/​email

0.057

***

0.230

***

Cons.

0.071

***

0.078

*

Number of obs

21120

20337

20440

F

410.44

277.81

368.41

Prob > F

0

0

0

R-​squared

0.29

0.3164

0.3941

Adj R-​squared

0.2893

0.3164

0.3931

Root MSE

0.17289

0.1695

0.15974

*** significant at 99%; ** significant at 97,5%; * significant at 95% Source: Authors’ elaboration from EWCS 2010

innovation has contributed significantly to a reduction of health and safety hazards at work, and contributed to the growing importance of skills, autonomy and teamwork in many types of work, thus making work more intrinsically rewarding. Those advances in job quality have themselves contributed to further innovation and a smoother introduction of technical changes in productive systems, thus reinforcing the positive effects. Although a broad historical sketch of the relationship between innovation and job quality must acknowledge this positive impact, it would be wrong to assume that in the future this positive impact will continue, and for at least three reasons. First, it is

268    rafael muñoz de bustillo et al important to acknowledge that the improvement of job quality as a result of innovation is by no means automatic or unproblematic. Although innovation facilitates improvements in job quality, in most cases those improvements have only occurred after conflict and struggle (as in the case of the reduction of working time or the increase in real wages) or after significant changes in labour regulation (as in the case of improvements in health and safety standards). Second, there are also historical examples of innovations (or technical changes) with a negative effect on job quality, at least in the short term, for example the terrifying conditions of work of the early industrialization in Britain or in the original Taylorist experiments of the early twentieth century. The fact that they ended well in the long run and working conditions improved had more to do with institutional and political factors than with technology or innovation as such. Third, although history helps to make sense of the present, the future remains unknown. Innovation has tended to contribute positively to job quality in the past but perhaps the future will be different. The three vectors of change identified by Fernández-​Macias (2018) are potentially highly disruptive in terms of job quality. As with previous waves of innovation, whether they end up benefiting many or just a few will ultimately depend on social and political dynamics. This chapter also illustrated some of the previous conceptual and theoretical points with an empirical exploration of the relationship between innovation and job quality drawing on the 2010 EWCS. Individual-​level analysis confirms that there is a clear empirical correlation between indicators of innovation in the firm and the quality of jobs, which is particularly strong for product and process innovation, and is also reflected in a high correlation between use of digital technologies and job quality. At the country level, the relationship is also clear, with more innovative countries having generally higher levels of job quality, in particular for the dimensions of intrinsic quality of work, employment conditions and health and safety. Going forward, given the time lags with previous technological revolutions between their introduction and their impact on job quality—​and which might explain some of the paradoxes of the digital economy, such as its apparent lack of effect on productivity (Gordon 2016)—​it seems reasonable to think that the effects of the new world of the digital economy may only slowly emerge. However, if those effects do emerge, because technological revolutions change the nature of economic activity, it is likely that measurement will be difficult with existing analytical tools. New tools will be needed to capture and understand the future relationship between innovation and job quality in Europe and elsewhere.

Acknowledgement The authors want to thank Martina Bisello for her assistance in some of the analysis featured in this chapter. We acknowledge funding from the project EU H2020-​EURO-​ SOCIETY-​2014 QuInnE: Quality of Jobs and Innovation Generated Employment Outcomes.

innovation and job quality    269

References Abramovitz, M. (1956), ‘Resource and Output Trends in the U.S. since 1870’, American Economic Review, May 1956 (Papers and Proceedings), 46/​2, 5–​23. Acemoglu, D., and Autor, D. H. (2011), ‘Skills, Tasks and Technologies: Implications for Employment and Earnings’, in O. Ashenfelter and D. E. Card, eds., Handbook of Labor Economics. Vol. 4B (Amsterdam: Elsevier), 1043–​1171. AFL-​CIO (2013), Death on the Job: The Toll of Neglect (Connecticut: AFL-​CIO). Akerlof, G., and Yellen, J. (1986), Efficiency Wage Models of the Labor Market (Cambridge: Cambridge University Press). Albin, M., Mathieu, C., Takala, E-​P, and Theorell, T. (2022), ‘The Cornerstone of Job Quality –​ Occupational Safety and Health’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press). Appelbaum, E., Bailey, T., Berg, P., and Kalleberg, A. (2000), Manufacturing Advantage (Ithaca: ILR Press). Arntz, M., Gregory, T., and Zierahn, U. (2016), ‘The Risk of Automation for Jobs in OECD Countries: A Comparative Analysis’, OECD Social, Employment and Migration Working Papers, No. 189. Paris: OECD Publishing. Autor, D. H., Katz, L. F. and Kearney, M. S. (2006), ‘The Polarization of the U.S. Labor Market’, American Economic Review Papers and Proceedings, 96/​2, 189–​194. Berg, J., Furrer, M., Harmon, E., Rani. U., and Six Silberman, M. (2018), Digital Labour Platforms and the Future of Work: Towards Decent Work in the Online World (Geneva: ILO). Bolt, J., and van Zanden, J. L. (2013), ‘The First Update of the Maddison Project; Re-​Estimating Growth Before 1820’, Maddison Project Working Paper 4. Bosworth, B. and Collins, S. M. (2008), ‘Accounting for Growth: Comparing China and India’, Journal of Economic Perspectives, 22/​1, 45–​66. Boxall, P., and Macky, K. (2009), ‘Research and Theory on High-​ Performance Work Systems: Progressing the High Involvement Stream’, Human Resource Management Journal, 19/​1, 3–​23. Braverman, H. (1974), Labor and Monopoly Capital (New York: Monthly Review Press). Cappelli, P., and Neumark, D. (2001), ‘Do ‘High Performance’ Work Practices Improve Establishment Level Outcomes?’ Industrial and Labor Relations Review, 54, 737–​776. Chang, A. (2014), ‘Army of Amazon Robots Ready to Help Fulfill Orders on Cyber Monday’, Los Angeles Times. Technology Now, (http://​www.latimes.com/​business/​technology/​la-​fi-​tn-​ amazon-​warehouse-​cyber-​monday-​20141130-​story.html) Cirillo, V. (2018), ‘Job Polarization in European Industries’, International Labour Review, 157/​ 1, 39–​63. De Groen, W. P., Kilhoffer, Z., Lenaerts, K., and Mandl, I. (2018), Employment and Working Conditions of Selected Types of Platform Work (Dublin: Eurofound). Easterly, W., and Levine, R. (2001), ‘It’s Not Factor Accumulation: Stylized Facts and Growth Models’, World Bank Economic Review, 15/​2, 177–​219. European Agency for Safety and Health at Work (EU-​OSHA) (2018), Foresight on New and Emerging Occupational Safety and Health Risks Associated with Digitalisation by 2025 (Luxembourg: Publications Office of the European Union). European Commission (EC) (2012), Europe 2020 (Brussels: European Commission) http://​ ec.europa.eu/​europe2020

270    rafael muñoz de bustillo et al Fernández-​Macias, E. (2012), ‘Job Polarisation in Europe? Changes in Employment Structure and Job Quality, 1995-​2007’, Work and Occupations, 39/​2, 157–​182. Fernández-​Macías, E. (2018), Automation, Digitalisation and Platforms: Implications for Work and Employment (Dublin: Eurofound). Fernández-​Macías, E., and Hurley, J. (2017), ‘Routine-​Biased Technical Change and Job Polarization in Europe’, Socio-​Economic Review, 15/​3, 563–​585. Fernández-​Macías, E., Muñoz de Bustillo, R., and Antón, J. I. (2016), ‘Job Quality in the First Decade of the 21st Century, MPRA Archive Working Paper. Munich. Findlay, P., Warhurst, C., Keep, E., and Lloyd, C. (2017), ‘Opportunity Knocks? The Possibilities and Levers for Improving Job Quality’, Work and Occupations, 44/​1, 3–​22. Flood, P. C. et al. (2008), New Models of High Performance Work Systems: The Business Case for Strategic HRM (Dublin: Partnership and Diversity and Equality Systems, National Centre for Partnership & Performance). Freeman, C., and Louca, F (2001), As Time Goes By: From the Industrial Revolutions to the Information Revolution (Oxford: Oxford University Press). Frey, B. F., and Osborne, M. A. (2013), ‘The Future of Employment: How Susceptible are Jobs to Computerization?’, published in 2017 in Technological Forecasting and Social Change, 114, 254–​280. Fu, N., Flood, P. C., Bosak, J., Morris, T., and O’Regan, P. (2015), ‘How Do High Performance Work Systems Influence Organizational Innovation in Professional Service Firms?’ Employee Relations, 37/​2, 209–​231. Godard, J. (2004), ‘A Critical Assessment of the High-​Performance Paradigm’, British Journal of Industrial Relations, 42/​2, 349–​378. Goos, M., and Manning, A. (2007), ‘Lousy and Lovely Jobs: The Rising Polarization of Work in Britain’, Review of Economics and Statistics, 89, 119–​133, Goos, M., Manning, A., and Salomons A. (2009), ‘Job Polarization in Europe’, American Economic Review: Papers & Proceedings, 9, 58–​63. Goos, M., Rademakers, E., Salomons, A and Vandeweyer, M. (2022), ‘Job Polarization: Its History, an Intuitive Framework and Some Empirical Evidence’ in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press). Gordon, R. J. (2016), The Rise and Fall of American Growth: The U.S. Standard of Living since the Civil War (Princeton: Princeton University Press). Grande, R., Muñoz De Bustillo Llorente, R., Fernández-​Macías, E. and Anton, J. I. (2020), ‘Innovation and Job Quality. A Firm-​Level Exploration’, Structural Change and Economic Dynamics, 54, 130–​142. Harden E. E., Kruse D. L, Blasi, J. R. (2006), ‘Who Has a Better Idea? Innovation, Shared Capitalism, and Human Resources Policies’, in D. L. Kruse, R. B. Freeman and J. R. Blasi, eds., Shared Capitalism at Work: Employee Ownership, Profit and Gain Sharing, and Broad-​ based Stock Options (Chicago: University of Chicago Press). Hefferenan M., Flood P. C., and Liu, W. (2011), ‘High Performance Work Systems –​ International Evidence of the Impact on Firms and Employees’, in A.-​W. Harzing and A. Pinnington, eds., International Human Resource Management. 3rd. ed. (London: Sage). Hefferenan M., Harney, B., Cafferkey, K., and Dundon, T. (2008), People Management and Innovation in Ireland, Galway Working Series Research Paper No: 27. National University of Ireland. Hulten, C. R. (2009), Growth Accounting, Working Paper 15341. Cambridge Mass: National Bureau of Economic Research, http://​www.nber.org/​pap​ers/​w15​341.

innovation and job quality    271 Hunt, J., and Nunn, R. (2022). ‘Has US Employment Really Polarized? A Critical Reappraisal’, Labour Economics, 75, 102117. Kaushik, D., and Mukherjee, U. (2021), ‘High-​Performance Work System: A Systematic Review of Literature’, International Journal of Organizational Analysis, https://​doi.org/​10.1108/​ IJOA-​07-​2020-​2282 Knox, A., and Wright, S. (2022), ‘Understanding Job Quality Using Qualitative Research’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press). Kranzberg, M., and Pursell, C. W (1967), Technology in Western Civilization. The Emergence of Modern Industrial Society Earlies Times to 1900 (Oxford: Oxford University Press). Levy, F., and Murnane, R. J. (1992), ‘U.S. Earnings and Earnings Inequality: A review of Recent Trends and Proposed Explanations’, Journal of Economic Literature, 30: 1333–​81. Maastricht Economic and Social Research and Training Centre on Innovation and Technology (UNU-​MERIT) (2011), Innovation Union Scoreboard 2010. The Innovation Union’s Performance Scoreboard for Research and Innovation, Innovation Union Scoreboard (IUS) 2010 (Publications Office of the EU), europa.eu. Makó, C., Illéssy, M., and Warhurst, C. (2016), The Evolution of EU Innovation Policy Relevant to Job Quality and Employment, QuInnE Working Paper 2, QuInnE. Mishel, L., Schmitt, J., and Shierholz, H. (2013). Assessing the Job Polarization Explanation of Growing Wage Inequality. Economic Policy Institute. Working Paper. MMWR, Morbidity and Mortality Weekly Report (1999), Achievements in Public Health, 1900-​ 1999: Improvements in Workplace Safety—​United States, 1900-​1999, 48/​22, 461–​469, (http://​ www.cdc.gov/​mmwr/​prev​iew/​mmwrh​tml/​mm482​2a1.htm). Mokryl J. (1990), The Lever of Riches: Technological Creativity and Economic Progress (Oxford: Oxford University Press). Muñoz de Bustillo, R., Fernández-​Macías, E., Antón, J. I., and Esteve, F. (2011), Measuring More than Money (Chelthenham: Edward Elgar). Noble, D. F. (2011), Forces of Production: A Social History of Industrial Automation (New Brunswick, NJ: Transaction Publishers). OECD (2005), Oslo Manual: Guidelines for Collecting and Interpreting Innovation (Data, Paris: OECD). Oesch, D. (2013), Occupational Change in Europe: How Technology and Education Transform the Job Structure (Oxford: Oxford University Press). Oesch, D., and Piccitto, G. (2019), ‘The Polarization Myth: Occupational Upgrading in Germany, Spain, Sweden, and the UK, 1992–​2015’, Work and Occupations, 46/​4, 441–​469. Perez, C. (2003), Technological Revolutions and Financial Capital The Dynamics of Bubbles and Golden Ages (Cheltenham: Edward Elgar). Pesole A., Urzi Brancati, M. C., Fernandez-​Macias, E., Biagi, F., and Gonzalez-​Vazquez, I. (2018), ‘Platform Workers in Europe: Evidence from the COLLEEM Survey’, JRC Working Papers JRC112157. Seville: Joint Research Centre, jrc112157_​pubsy_​platform_​workers_​in_​ europe_​science_​for_​policy.pdf Pfeffer, J. (1998), The Human Equation: Building Profits by Putting People First (Boston: Harvard Business School Press). Raff D. M. G., and Summers, L. H. (1987), ‘Did Henry Ford Pay Efficiency Wages?’ Journal of Labor Economics, 5/​4, Pt 2, S57–​S86. Roediger, D. R., and Sheldon, P. (1998), Our Own Time: A History of American Labor and the Working Day (London: Verso).

272    rafael muñoz de bustillo et al Sharpe A., Arsenault, J. F., and Harrison, P. (2008), The Relationship between Labour Productivity and Real Wage Growth in Canada and OECD Countries, Ottawa: CSLS Research Report No. 2008-​8. Smith, A. (1776/​1981), An Inquiry into the Nature and Causes of the Wealth of Nations., ed. by R. H. Campbell, A. S. Skinner, and W. B. Todd (Indianapolis: Liberty Fund). Stewart, I., De, D., and Cole, A. (2015), Technology and People: The Great Job-​Creating Machine (London: Deloitte). Timmer, M. P., Ypma, G., and Ark, B. van der (2003), ‘IT in the European Union: Driving Productivity Divergence?’ GGDC Research Memorandum 200363, Groningen Growth and Development Centre, University of Groningen. Tomer, J. (2001), ‘Understanding High-​Performance Work Systems: The Joint Contribution of Economics and Human Resource Management’, Journal of Behavioral and Experimental Economics (formerly The Journal of Socio-​Economics), 30/​1, 63–​73. Violante, G. L. (2008), ‘Skill-​Biased Technical Change’, in S. N. Durlauf and L. E. Blume, eds., The New Palgrave Dictionary of Economics Online. 2nd ed. (Basingstoke: Palgrave Macmillan), 1–​6. Way, S. (2002) ‘High Performance Work Systems and Intermediate Indicators of Firm Performance within the US Small Business Sector’, Journal of Management, 28/​6, 765–​785. Whaples, R. (1990), ‘The Shortening of the American Work Week: An Economic and Historical Analysis of Its Context, Causes, and Consequences’. PhD dissertation, University of Pennsylvania. White, M., Hill, S., McGovern, P., Mills, C. and Smeaton, D. (2003), ‘High-​Performance Management Practices, Working Hours and Work-​Life Balance’, British Journal of Industrial Relations, 41/​2, 175–​195.

APPENDIX 1 probability of computerization and jqi in 39 occupations ISCO2

Probability of Computerisation

JQI

Health professionals

0,02173782

0,5911642

Teaching professionals

0,06215696

0,6233893

Chief executives, senior officials and legislators

0,08784149

0,6942878

Science and engineering professionals

0,09353647

0,6269547

Production and specialised services managers

0,12272096

0,6532809

Hospitality, retail and other services managers

0,13093289

0,5545289

Legal, social and cultural professionals

0,13465769

0,6182067

Information and communications technology professionals 0,15570292

0,6691521

Administrative and commercial managers

0,23084382

0,6705127

Business and administration professionals

0,30372510

0,6562188

innovation and job quality    273 Health associate professionals

0,34279804

0,4971154

Protective services workers

0,40273333

0,5272508

Legal, social, cultural and related associate professionals

0,43130939

0,5122097

Personal care workers

0,48106021

0,4086274

Science and engineering associate professionals

0,49869220

0,5626962

Information and communications technicians

0,50780246

0,5878749

Electrical and electronic trades workers

0,54347302

0,5013853

Personal service workers

0,55689109

0,3671352

Business and administration associate professionals

0,55727253

0,5943613

Cleaners and helpers

0,60510232

0,3250253

Handicraft and printing workers

0,61614064

0,4252262

Drivers and mobile plant operators

0,62042025

0,4257194

Sales workers

0,64028209

0,3889237

Labourers in mining, construction, manufacturing and transport

0,69220248

0,3281024

Building and related trades workers, excluding electricians

0,70069898

0,4219076

Metal, machinery and related trades workers

0,70161495

0,4273907

Customer services clerks

0,70916141

0,4702263

Food processing, wood working, garment and other craft and related trades workers

0,73228666

0,3974339

Market-​oriented skilled agricultural workers

0,73944719

0,4378883

Market-​oriented skilled forestry, fishery and hunting workers

0,74060898

0,4781792

Refuse workers and other elementary workers

0,78143061

0,3411435

Other clerical support workers

0,83506667

0,472204

Stationary plant and machine operators

0,85989215

0,3988743

Food preparation assistants

0,86000000

0,2745264

Agricultural, forestry and fishery labourers

0,87333333

0,3682381

Assemblers

0,89875000

0,3818407

Numerical and material recording clerks

0,92793694

0,4972787

Street and related sales and service workers

0,94000000

0,4798225

General and keyboard clerks

0,95820556

0,4913985

Source: Authors’ analysis from Frey and Osborne (2013) and EWCS microdata

chapter 12

immigration and job quality amada armenta and shannon gleeson

Introduction The study of job quality typically considers a range of indicators—​including wages and benefits, opportunity for mobility and worker autonomy. On each of these points, the employment relationship tends to be the focus of analysis for labour and stratification scholars. However, for 244 million migrants worldwide, employment experiences are shaped by their immigration status, their host country, uneven access to workplace protections, and challenges related to language ability and human capital. Fuelled by biased media reporting and anti-​immigrant government campaigns, immigrants may also encounter racism, discrimination, and xenophobia at work and in the communities in which they live. Given that immigrant workers are amongst the most vulnerable labourers in the workforce, understanding their experiences informs our understanding of other marginal workers who either lack key protections or struggle to access them due to their structural positions. As a result, we must pay close attention to the workplace dynamics facing foreign-​born workers. This chapter reviews the interdisciplinary literature on job quality for immigrant workers, and presents data from a case study of Latino immigrant workers in Philadelphia. A re-​emerging immigrant gateway in the US (Singer 2015), Philadelphia represents a relatively positive case study for immigrant integration. In April 2014, the Philadelphia mayor signed an executive order decreeing that Philadelphia was a ‘sanctuary city’ and would limit its cooperation with federal immigration enforcement efforts. Moreover, the city’s compact size and extensive public transportation means that immigrants can move around the city with relative ease. Data from this chapter come from a 2014 pilot survey of 340 Latino immigrants (155 men; 155 women) in South Philadelphia, as well as 75 in-​depth interviews with Mexican immigrant men and women conducted between 2015 and 2017. Through this survey and interviews, we consider how legal status and gender jointly impact immigrant workers. More specifically, we examine the challenges that immigrant women face in obtaining formal

immigration and job quality    275 and informal job opportunities, negotiating power at work, and contesting workplace protections when violations occur. Lastly, we discuss the defining factors shaping their precarious work lives, as well as the moments of agency they exert at work and beyond.

Conceptualizing Job Quality and Immigration Immigrant Labour in the US and Beyond Migration is a worldwide phenomenon, with a third of the world’s migrants residing in Europe (76.1 million), another third in Asia (75.1 million), followed by 46.6 million in the United States of America (United Nations, Population Division, Department of Economic and Social Affairs 2016). Individuals migrate for many reasons, but chief among them is the search ‘to find decent work, and escape poverty, persecution and violence’. Migrants form an important part of the low-​wage workforce in traditional destinations (such as Canada, Australia and Europe), but also in regions of South–​South migration (such as Latin America, Asia and Africa). While the political context of each destination differs significantly, economic precarity, political violence, and the drive to seek new opportunities continue to push migrants beyond their national borders. This movement is one of the many brutal forms of ‘expulsion’ in the global economy (Sassen 2014). For example, in South Africa the government has tightened their border with Mozambique to prevent the entry of migrant workers, in Germany a longstanding Turkish migrant population still faces significant discrimination, and in the United Arab Emirates, guest workers face staggering inequality. While much of the stratification scholarship focuses on the severe conditions facing low-​wage immigrant workers, foreign-​born workers also constitute a significant part of the ‘high-​skilled workforce,’ which faces unequal wages, uneven freedom of speech, and lack of protections from deportability. In 2015, foreign-​born workers constituted 16.7 per cent of the United States labour force, about half of whom identified as Hispanic and another quarter as Asian. Only 16.8 per cent of foreign-​born workers identified as White, and 9.2 as Black. On the whole, foreign-​born workers had slightly higher labour force participation rates than their native-​born counterparts (65.2 versus 62.2). However, the gap in labour force participation for foreign and native-​born men was far higher (78.2 versus 67.3), rates for foreign versus native-​born women are lower (52.9 versus 57.4), and especially so for mothers with minor children (57.8 versus 73.4). Foreign-​born workers also earned 81 cents to the dollar of native-​born workers, reflecting inequities in human capital, distribution across occupation and industry, and regional concentration. However, even controlling for educational attainment, native-​born workers earned more, especially at lower levels of education (Bureau of Labor Statistics 2016).

276    amada armenta and shannon gleeson Though the primary mechanism of immigration to the United States since 1965 is family-​based petitions, for the vast majority of low-​wage workers who aspire to migrate to the US, there is no reasonable path to legal entry. For example, visa wait times for some applicants seeking family sponsorship spans decades (Shusterman Law 2018). Authorized immigrant labour in the United States is shaped largely by a patchwork of policies that favour high-​skilled workers (for example through the H1-​B visa programme, which provides a pathway to citizenship), and agricultural and seasonal labour (through the H2-​A and H2-​B visa programmes, which are exclusively temporary guestworker statuses). While there are also several other forms of temporary relief that provide short-​term work authorization (Heeren 2015), most other immigrants arrive through family sponsorship or via the far more limited refugee or asylum processes. Among the 26.3 million foreign-​born workers in the United States, an estimated 8 million (5 per cent of the civilian labour force) are unauthorized. These workers work across all sectors, but are particularly over-​represented in farming and construction (Passel and Cohn 2016). These workers are particularly likely to work in hazardous workplaces (Orrenius and Zavodny 2009; Hall and Greenman 2014), work in industries with high rates of wage theft (non-​payment of wages) and other workplace violations (Bernhardt et al. 2013), and face barriers to filing claims against their employer (Gomberg-​Muñoz 2011; Gleeson 2016). Non-​citizen workers with temporary legal status also face significant challenges (Goldring and Landolt 2011; Abrego and Lakhani 2015). Under the 1986 Immigration Reform and Control Act, all workers in the United States must show work authorization through the I-​9 process. This process has placed immigration enforcement in the hands of employers, and made the workplace a key site of enforcement actions, largely through audits of employer I-​9 documents (Griffith 2012). Beyond the Department of Homeland Security, these efforts implicate surveillance from a range of other agencies including the Social Security Administration and Internal Revenue Service. While it is widely believed that undocumented immigrants do not pay taxes in the United States, the vast majority of these workers are in the formal sector and pay employment taxes and into benefit systems such as Social Security. In 2010, the Social Security Earnings Suspense File, which includes benefits paid in that cannot be paid out, included over $1 trillion in tax contributions (Kugler et al. 2013). In the United States, immigrant workers are generally eligible for the same workplace protections as their native-​born counterparts, such as for wage and hour standards, health and safety protocols, and prohibitions against most forms of discrimination and sexual harassment (The Legal Aid Society-​Employment Law Center 2015). Important exceptions include national origin discrimination protections for non-​citizen workers. There are also unequal remedies for reinstatement and backpay under the National Labor Relations Act (NLRA) for unauthorized workers, stemming from a 2002 Supreme Court decision in Hoffman Plastic Compounds, Inc. v. NLRB (Fisk and Wishnie 2005). This ‘immployment’ context (Griffith 2011) has produced a paradoxical set of laws that simultaneously protect workers without regard to their immigration status, but also marks unauthorized immigrants as a target of surveillance at the workplace and beyond.

immigration and job quality    277 The Trump administration in many ways married the devastating spectacle of raids characterized by the Bush administration, and the effective use of employer audits under the Obama-​era ‘silent raids’ (Griffith and Gleeson 2019). These are concerns not only for unauthorized workers, but also other non-​citizen workers whose statuses are subject to political volatility. For example, beneficiaries of the Deferred Action for Childhood Arrivals programme and the Temporary Protected Status programmes—​both of which are afforded short-​term deportation relief and work authorization largely at the discretion of the president—​faced the threat of having their programmes ended during the Trump administration. Beyond federal law, protections for injured unauthorized workers can also vary from state to state in the United States. For example, while some states have curbed protections for unauthorized workers’ access to workers’ compensation, states such as California have strengthened them to include specific prohibitions against employer retaliation on the basis of immigration status (National Employment Law Project 2013). California’s Agricultural Labor Relations Board also provides an important resource for farmworkers (the majority of whom are foreign born) who, along with domestic workers, are explicitly excluded from collective bargaining rights under federal legislation (Griffith 2017). Local governments in the United States too have stepped in to enhance rights for all workers, such as San Francisco’s paid sick leave policy and $15 minimum wage law (de Graauw 2017). However, even when workers are equally covered under the law, their multiple intersectional immigrant identities often render them ‘marginal’ workers who face an uphill battle in formally accessing these rights and remedies (Garcia 2012). Furthermore, the claims process poses a series of bureaucratic hurdles, especially for immigrant workers, even in the most worker-​friendly contexts (Gleeson 2016). Additionally, language and cultural barriers can make access difficult, especially in agencies that lack a diverse outreach and frontline staff. For unauthorized immigrants, lack of access to government-​issued identification, combined with local policing efforts can lead to immigrant detention, even in immigrant-​friendly contexts (Armenta 2017). But even some documented workers face significant challenges if their legal status is temporary and/​or tied to their employment (Griffith and Gleeson 2017). In many places, coalitional civil society efforts have evolved to strengthen worker protections and increase access to existing ones (Fine 2006; Gordon 2007; Gleeson 2012a; de Graauw 2016). Immigrants have also become an integral part of the labour movement in both the US and Europe (Milkman 2011; Adler et al. 2014), but remain relatively under-​represented in unions and face an array of challenges related to the flexibilization of the US labour market, across all sectors and skill levels (Weil 2014).

Operationalizing Good Jobs, Bad Jobs for Immigrants Though a full assessment of job quality requires multiple dimensions, across most all measures, job quality has fallen significantly in the United States (and beyond) since

278    amada armenta and shannon gleeson the mid-​1970s, both in terms of the jobs available to the changing labour force, and the demographic composition of those seeking work. Key dimensions of job quality include both economic reward (pay and other fringe benefits such as health insurance and retirement benefits), as well as other less tangible, though equally important, aspects such as control over one’s schedule and the ability to meet life demands beyond the workplace (Kalleberg 2011). There has been substantial debate regarding the impact of immigrant labour on job quality in various sectors. Milkman (2006) has argued that in some industries a degradation of job conditions preceded or paralleled the significant increase in foreign-​born labour in the United States. To be sure, the same forces of globalization that have shrunk the welfare state and deregulated much of employer activity have also led to mass displacement of workers seeking economic opportunity beyond their place of origin. The degradation of job quality takes on a number of sector-​specific qualities, especially in those jobs where immigrants are over-​represented. For example, Doussard (2013)’s study of the immigrant and day labour-​heavy residential construction industry in Chicago highlights how the uncertain schedules of seasonal contract labour collide with a push to simultaneously work fast in hopes of being retained for the next job, and an hourly pay structure that becomes a race to the bottom in cut-​throat bidding wars. Worker advocates, in their push to hold nefarious employers accountable, must also walk a fine line between forcing compliance of minimum basic standards and even pushing for more, and running smaller less competitive employers out of business without changing the structure of the industry. Residential construction—​in which 20 to 30 per cent of workers are estimated to be undocumented (157)—​thrives on subcontracting and misclassification of independent contractors and a contracting model that has increasingly relied on unskilled labour while pulling in skilled workers for much more delimited aspects of a project (165). As union power has declined, the capacity of labour standards enforcement agencies to hold employers accountable has also weakened, leaving many industries ripe for abuse. This is accentuated further in an era of increased immigration enforcement at the workplace. Agriculture is another major destination for immigrant labour that has received significant attention. Similar to residential construction, an increased reliance on subcontracting and temp agencies has also devolved employer responsibility away from the grower. In his ethnography of a largely indigenous Mexican immigrant workforce in berry picking, Holmes (2013) discusses the physical toll this arduous labour takes on workers’ bodies, and the ways in which workers often adopt a narrative that naturalizes this suffering (Holmes 2007). These narratives may become a survival strategy for immigrants in other high-​ paced environments such as restaurants where immigrant workers may leverage their ability to endure stressful and sometimes dangerous conditions as their comparative advantage (Gleeson 2012b). Aside from the dynamics of industries where immigrant workers are concentrated, there are additional considerations that those seeking to understand immigrant job quality must consider. See Table 12.1. Immigrants must integrate into the racial order of the host society, and often are subject to the existing structures of racism and

Immigration and Job Quality    279 Table 12.1 Job quality and immigration (select dynamics facing immigrant workers) Some Immigration-​Related Considerations

Key Covariates Race/​Ethnicity

Are immigrant workers subject to racism and xenophobia?

Family Status

Does the immigrant worker have transnational family obligations that influence the worker’s labour market strategy?

Language Ability

Does this immigrant speak the same language as the dominant language(s) in the host country/​city?

Human Capital/​Credentials

What are the dynamics of educational self-​selection for this migrant flow? Does the host country recognize/​transfer the immigrant’s credentials?

Relationship to Coercive Arm of State

Does the worker have legal status? Is the immigrant’s legal status considered a surveillance target? Is the immigrant worker deportable?

Regional Differences

Does the immigrant’s place of residence/​work have supportive/​punitive policies targeting immigrants? Is there a history of migration to this place? What role do ethnic niches play in funnelling workers into/​ out of the secondary labour market?

Indicators of Job Quality Employment Status

Do host country immigration laws prohibit undocumented workers from legal employment?

Wages/​Compensation

Do immigrants use a dual frame of reference when negotiating wages?

Unionization

Do these categories of immigrants have collective bargaining protections under federal and state/​ provincial law?

Health and Safety Conditions

Are immigrants concentrated in industries that face particular health and safety concerns (ex: agriculture, construction)

Benefits

Does the host country provide access to welfare state benefits (e.g. social security, national health insurance)?

Worker Autonomy

Does the immigrant have a legal status that impacts her choice to engage in exit, voice or loyalty when considering whether to mobilize her rights at work?

280    amada armenta and shannon gleeson xenophobia (Golash-​Boza 2015), which in turn shape their labour market opportunities and barriers. Immigrants with a spouse and children to support, just as any other breadwinner, may be more likely to make sacrifices at the workplace. However, immigrant workers may also have to juggle transnational obligations that keep them from speaking up about harsh conditions (Gleeson 2015). In either case, immigrant workers may employ a dual frame of reference when considering their options to mobilize their rights (Piore 1979). Immigrant workers also often face linguistic challenges if they don’t speak the dominant host language; on the other hand bilingual immigrants may have a labour market advantage in certain settings (Medvedeva and Portes 2017). In addition to these individual characteristics, immigrant context of reception (both national and sub-​national) matters across a range of factors. These include policies intended to either surveil or integrate immigrants, as well as the labour and employment protection policies and other benefits that shape the benefits and remedies afforded to immigrant workers (Pham and Van 2010).

Methods In this chapter we present survey and interview material drawn from a mixed methods study of immigrant life in Philadelphia to hone in on the dynamics of job quality and immigration. In this section, we describe the choice of Philadelphia as a case study, then describe our empirical approach to data collection, and finally provide a profile of the study sample.

Philadelphia as an Immigrant Destination While Philadelphia was a major receiver of immigrants in 1900, by the middle of the twentieth century, immigrants comprised a small proportion of the population. In 1970, for example, only 5 per cent of residents in Philadelphia’s metropolitan area were foreign born (Singer et al. 2008). Since 2000, the immigrant population in Philadelphia has increased dramatically, comprising an estimated 10 per cent of the population by 2014 (Singer 2015). Latino immigrants, and particularly Mexicans, have fuelled much of the immigrant population growth in the city. In 2000, the US Census estimated there were approximately 6,000 Mexican residents living in Philadelphia and by 2010, these estimates had increased to 20,000. While some migrants arrived directly from Mexico, others initially arrived in New York or New Jersey, and moved to Philadelphia because of its affordable housing costs and expanding construction and service jobs. As is often the case in places with new immigrant communities, the first migrants to arrive were young men, but over time both men and women made the journey north. Many settled in South Philadelphia, an area historically associated with Italian immigrants who arrived in the early 1900s, when

immigration and job quality    281 Philadelphia was a major receiver of immigrants (Singer 2015). Of course, by the time that Mexican and Latino migrants began arriving in South Philadelphia, many Italians and their descendants had moved out of South Philadelphia and into the suburbs. Latino immigrants replaced an ageing and suburbanizing Italian community, opening businesses and restaurants in commercial corridors that had long been vacant. Newcomers arrived in a city that had undergone a major economic and demographic transition. Economically, the city lost its manufacturing sector, and many jobs left the centre of the city and shifted to suburban counties (Brookings Institution 2003; Singer et al. 2008). Demographically, the city lost population between 1980 and 2000, with residents—​particularly White residents—​moving out to the suburbs (Brookings Institution 2003; Singer et al. 2008). As of the result of White out-​migration, Philadelphia is a majority-​minority city, with a large native-​born Black population, a significant native-​born Puerto Rican community, and immigrant residents of all racial and ethnic backgrounds. Today, Philadelphia’s economy is relatively stagnant: for example, while the United States experienced almost 20 per cent employment growth between 1992 and 2002, Philadelphia experienced 13 per cent employment growth during that period (Brookings Institution 2003; Singer et al. 2008). Most of this employment growth occurred in the retail and service sectors, jobs that tend to provide low wages, few benefits and irregular hours. Unlike other new destinations, Philadelphia has a relatively friendly context of reception for unauthorized residents. In April 2014, the Philadelphia mayor signed an executive order decreeing that local police would not honour Immigration and Customs Enforcement (ICE) detainer requests, making Philadelphia a ‘sanctuary city’. Moreover, the city’s compact size and extensive public transportation means that immigrants can move around the city with relative ease, and immigrants contribute 10 per cent of the total economic output in Philadelphia. This makes Philadelphia a positive case study for understanding immigrant integration, in a state where 7.5 per cent of the workforce is foreign-​born (American Immigration Council 2015).

Empirical Approach This chapter draws on quantitative and qualitative data from the South Philly Immigration Project, initiated by researchers at the University of Pennsylvania in 2014. A community-​based research project, it sent trained Latino residents into South Philadelphia neighbourhoods to conduct surveys of foreign-​born Latinos. Eligible participants included men and women between the ages of 18 and 65 who were born in Latin America and resided in South Philadelphia at the time of the survey. With no other sampling frame upon which to draw, South Philadelphia was divided into equally sized quadrants, and enumerators were instructed to knock on doors until they identified eligible respondents living in a household. The survey includes at least two men and women (from separate households) from each quadrant, but more respondents in quadrants that are more densely populated with Latino residents. Surveys collected

282    amada armenta and shannon gleeson information about migration history, household structure, and work. Ultimately, we collected data from 310 Latino immigrant men and women living in South Philadelphia (equally divided between men and women) during 2014 and 2015. The survey results are shown in Table 12.2. The respondents captured in this survey ranged in age from 18 to 55, but most respondents were in their twenties and thirties. Respondents’ length of time in Philadelphia ranged from several months to 24 years. The first respondent arrived in Philadelphia in 1991. On average, respondents had been in Philadelphia about nine years. Half of the respondents had arrived since 2007. While respondents hailed from a number of Latin American countries, 97 per cent of men and 90 per cent of women were Mexican. Most of our respondents were unauthorized: 84 per cent of men and 93 per cent of women lacked work authorization. Consistent with existing research (Bureau of Labor Statistics 2016), the men in this sample are far more likely to be employed than women. Almost 90 per cent of immigrant men worked, whereas only half of women did. Restaurants are important sites

Table 12.2 South Philly Survey descriptive statistics by gender Men

Women

Demographic Characteristics Age (Mean)

33

32

Years in Philadelphia (Mean)

8.6

9.3

Years of Education (Mean)

9

9

89.4

50.7

Employment Characteristics Percent in Labor Force Weekly Salary (Mean)

461

280

48

29

Restaurant

44.4

27.3

Construction

25.0

0.0

Retail

16.0

15.6

1.4

32.5

Weekly Hours Worked (Mean) Employment Sector (Percent)

Cleaning (Janitorial/​Housekeeping) Factory

2.8

7.8

Direct Sales/​Entrepreneurship

4.2

9.1

Other

6.3

3.9

Childcare

0.0

3.9

Source: South Philly Immigration Project.

immigration and job quality    283 of employment; 44 per cent of men and 27 per cent of women worked in restaurants. Immigrants tend to be concentrated in kitchen and support staff jobs, working as prep cooks, line cooks, bussers, dishwashers and bar backs. One quarter of men work in construction, and over 30 per cent of women either clean houses or businesses. Smaller proportions of men and women work in retail and factory jobs. In general, women’s work is more informal than men, that is, they are more concentrated in the domestic sphere and other less regulated work environments, and paid cash. For example, women are more likely to clean houses, take care of people’s children, and earn money by selling prepared food or products. Employed women worked significantly fewer hours than employed men (29 hours a week for women versus 48 hours a week for men). Moreover, while women earned less money than men (women earned an average of $280 per week and men earned an average of $460 a week), these discrepancies were because women worked fewer hours than men, and not because men earned higher wages. In addition to surveys, we drew on interviews with 95 immigrant men and women (65 women, and 30 men) living in South Philadelphia. Respondents were recruited through snowball sampling and asked a variety of questions about migration, work and experiences in Philadelphia. Interviews were conducted between 2015 and 2017. All interview identities were confidential and we used pseudonyms throughout this chapter, per our Institutional Review Board Human Subjects Protection protocol. All female interview respondents were Mexican, predominantly hailing from the Mexican states of Puebla and Guerrero (two relatively more recent sending regions in Mexico) (Riosmena and Massey 2012). The women we interviewed were between the ages of 18 and 54. At the time of the interview, one woman had lived in Philadelphia since 1992, while another had arrived just three months prior; on average, women had lived in Philadelphia for ten years and the average age at arrival was 22. All of the women we interviewed arrived in Philadelphia without authorization. While most are still undocumented, five women have adjusted their status through the U-​Visa programme. The U-​ Visa allows migrants who are victims of particular types of crime to adjust their status if they cooperate with legal authorities (Lakhani 2013). Whereas only half of women we surveyed were employed, over 70 per cent of the women we interviewed worked. With the exception of one Honduran respondent, all of the men we interviewed were also Mexican. They had lived in Philadelphia an average of eight years, and on average, they arrived at 20 years of age. All are currently employed, working in restaurants or construction jobs.

Immigrant Job Quality in South Philly In the sections that follow we draw on interviews to document the work experiences of immigrant men and women in Philadelphia. We find that despite similar aspirations, each experience different types of challenges at work.

284    amada armenta and shannon gleeson

Wage and Hour Conditions Unsurprisingly, wages were a primary concern for immigrant respondents in this study. The federal minimum wage increase in 2009 brought the rate to $7.25. Many states and localities have enacted far higher rates, though at the time of this study, the private-​ sector minimum wage in Philadelphia and the state of Pennsylvania no higher than the federal standard (Bradley 2015; Wage and Hour Division 2015; National Employment Law Project 2015). By law, cities in Pennsylvania have no authority to change the minimum wage, and when costs of living is taken into effect, Philadelphia has effectively the lowest minimum wage compared to other cities who also subscribe to the federal minimum (The Pew Charitable Trusts 2020). The men and women in our sample reported earning an average of $9.60 an hour, compared to the effective living wage of $11.59 for a single adult, or $24.10 for an adult with a spouse and two children, living in Philadelphia in 2015 (Otterbein 2015). Moreover, low-​wage immigrant workers face significant challenges in accessing labour standards enforcement protections (Kerwin and McCabe 2011). Self-​reported rates of wage theft among our South Philadelphia sample were relatively low compared to other surveys of low-​wage workers (Valenzuela et al. 2006; Burnham and Theodore 2012; Petrescu-​Prahova and Spiller 2016). Twenty-​three per cent of men and 18 per cent of women reported ever being a victim of wage theft in Philadelphia; men reported losing an average of $782 and women reported an average of $553. These differences are likely a conservative estimate, in part an artefact of the sectoral differences of these respondents, as well as relying on respondents to assess and report their own losses. The mechanisms driving wage theft vary across sectors. For example, 26-​year-​old Yesenia has lived in Philadelphia for seven years. Her first job was caring for a young girl, so the child’s mother could go to work. Yesenia worked from 6 a.m. to 2 p.m., Monday through Friday, and sometimes worked weekends. Regardless of the number of hours she worked, her employer paid her a fixed payment of $200 per week, bringing Yesenia’s hourly pay to $5 an hour or less. This fixed salary method of payment is a common practice in domestic work, especially for cash employees who may struggle to document and prove wage and hour violations. ‘I didn’t know. When you get here, you don’t know about minimum wage laws, so I went to work. When my brother told me I wasn’t getting paid enough, I left,’ explained Yesenia. Paola also recalls being paid very poorly in the first job she found upon arriving in Philadelphia, explaining that she was paid $4.50 an hour at a factory with Asian immigrant co-​workers with whom she could not communicate. Although the pay these women received was low, it was more than they would have received for similar work in Mexico. As a result, women took these jobs until they could secure better jobs with higher pay. It is not uncommon for undocumented workers to report direct threats from employers who allege that their immigration status precludes them from enjoying rights and benefits. For example, Jose, a 28-​year-​old restaurant worker, recounts how his chef repeatedly deflected Jose’s demands for 160 hours in back wages with threats. ‘I worked

immigration and job quality    285 like 80 hours a week, sometimes I’d work 7 days a week from 8 in the morning until midnight, and they paid me like $8.50 an hour. So I decided that I was going to find another job and I gave them a one week notice, and then the chef started insulting me and told me that this wasn’t my country and he didn’t even know why I was here. . . . I left and when I went to pick up my check, they didn’t want to pay me for the two weeks of work I was owed. I went back another time for my check and they threatened me and told me if I came back, they were going to call immigration on me.’ Jose persisted, and eventually received only half the money he was due. The restaurant eventually shut down two years later. Beyond basic minimums, immigrant workers often struggle to demand higher wages, and report constantly evaluating themselves in relation to their peers. For example, Felipe, earned $23 an hour as an electrician, and felt like he would be earning more money if he were American. In his first construction job, over ten years ago, Felipe earned $10 an hour. After years on the job, he gained additional skills, and when he switched construction firms, his new employer hired him as an electrician. As a result, while many of his peers in the restaurant industry cobbled together two shifts per day, Felipe worked one job, explaining that he had no desire to work eighty hours a week. He was proud of the wage he was earning, particularly because Felipe had won it after his employer accidentally revealed that Felipe was paid less than a US citizen who did the same job: My boss handed out the paychecks. One day my boss handed me a check, and I opened it, but it wasn’t mine, it was an American coworker’s. I saw the number of hours he worked and how much he earned, and that’s how I knew that he was earning more, just because he was American. I work harder than him and I do more than him. So when I got the check, I opened it, and then I told my boss ‘wait, this isn’t mine’ and when I gave it back to my boss, he was upset at the mistake. He was like ‘fuck!’ because he knew what I saw, and that I was not going to be happy. He found my check and gave it to me and asked ‘see you tomorrow?’ and I told him ‘of course, see you tomorrow.’ The next day he found me during my coffee break, but I was so mad. We talked, and he explained that he paid the other guy more because the guy had to pay for health insurance. When I didn’t respond, he agreed to give me a raise. I know I earn a lot, but citizens make $27, $30, or up to $35 an hour.

As described above, immigrants have equal rights to equal pay and other protections under federal, state and local law. Even so wage theft and discrimination claims can be incredibly difficult to pursue and settle; employers have significant discretion under the law.

Beyond Wages—​What Else Immigrant Workers Value Though compensation is clearly a top concern for any worker, immigrant workers we spoke to also found value in non-​economic considerations, often related to precarity in other spheres of life (Paret and Gleeson 2016). For example, Javier, an undocumented

286    amada armenta and shannon gleeson 28-​year-​old restaurant worker, appreciated how easily he could traverse Philadelphia by bike. Peers who lived in other cities were forced to drive, a routine task that poses the risk for undocumented workers of arrest and deportation because they are ineligible for driver’s licences in Pennsylvania (Armenta 2016, 2017; Stuesse and Coleman 2014). Though advocates have continued to press for a legislative fix, as in other states, this remains an ongoing legal battle (Sasinovic 2021). Similarly, payment method particularly mattered for many respondents, despite the associated risks. In a survey of three major US cities (Chicago, Los Angeles and New York), authors find that more than a third of hourly low-​wage workers are paid in cash. Ninety-​three per cent of these cash workers did not receive documentation of their earnings and deductions, as is mandatory in these jurisdictions. These cash workers also had double the minimum violation rate as those paid by cheque, and this was even higher for non-​hourly (flat weekly pay) workers. The rate of cash payment is even higher in industries where immigrant workers are concentrated, such as private households and apparel/​textile (Bernhardt et al. 2009; Bernhardt et al. 2013). Yet for many unauthorized immigrant workers, cash payment is desirable. For example, Esperanza left her first restaurant job after two years because her meagre wages were reduced even further after mandatory taxes were deducted from her paycheque. Struggling to pay debts related to her journey to Philadelphia, she looked for a job that paid in cash. Hector, 24 years old, has lived and worked in the US for nine years. He valued his job at a pizzeria, particularly because his employers paid him in cash. For unauthorized immigrants who are keen on avoiding detection and want to minimize interactions with bureaucracies, this is a key benefit. While unauthorized workers do have access to some financial institutions, they are often fearful of banking practices, access to which may be limited and cumbersome (Padua and Doran 2016). Alternative cheque cashing options are costly, and undocumented immigrants may fall victim to predatory financial practices. Cash payment, however, also leads to serious safety concerns, as immigrant workers may be victimized and targeted for robbery (Fussell 2011). Aside from pay, workplace conditions are obviously a concern. Upon arriving in Philadelphia from Michoacán (Mexico), Facundo quickly found a job in construction through a friend. The work was hard and the heat was unbearable in the summer. Still, the job came with a degree of autonomy that he found satisfying. Indeed, Facundo enjoyed the fact that every day he completed his work without excessive supervision or surveillance. He explains, ‘There’s no one here to pressure you or tell you, you have to do this or that, they just let you and you do your job, you do what you have to do.’ According to Kalleberg (Kalleberg 2011, 7), the ability to exercise discretion over one’s work, and shape one’s work environment, is a central non-​economic dimension of job quality. Similarly in her study of claimants, Gleeson (2016) finds that workers—​especially mothers—​often traded better wages for flexibility and a sense of respect from their supervisors. Marcela, who was 42 years old and had arrived in Philadelphia at the age of 34, recalls that her first job was cleaning restaurants. She explained, ‘It wasn’t the best job, but when you get here, you don’t have many options. I arrived and I wanted to work.

immigration and job quality    287 The job had a good schedule. My daughters were young, so I took them to school, then went to work, and I was home to pick them up and take care of them.’ It is important to note that despite their disadvantages, immigrant workers exert agency and seek out aspects of satisfaction and dignity in their jobs. Asuncíon, a 36-​ year-​old South Philadelphia prep cook remarked positively on her easy relationship with her co-​workers, and appreciated that the chef regularly takes note of her work ethic. ‘I like what I do. It’s not hard, I like it. I arrive at work, I prepare the food, and I know what I need to do, and I like the chefs. I work hard, and they know the people who work hard and those who don’t. I’ve never had a problem with them.’ Similarly, 32-​year-​old Ernesto appreciates his flexible schedule and understanding relationship he has with his boss in his job working for a catering and events company. He explains, ‘I’ve had my current job for 7 years. What I like about it is that the schedule is flexible, whenever I need a day off I just ask for it, and they always give it to me. The pay is good and my bosses don’t hassle me like bosses might at other jobs. I like it, it’s calm.’ When asked to elaborate about how other bosses occasionally hassle employees, Ernesto explained, ‘Sometimes when they see that there’s no work, they send you home, or they say, ‘Come clean this’, or they make you do extra work that’s not really your responsibility. They do that very infrequently here.’ To be sure, workers with more skills typically enjoy more autonomy and control at their job (Kalleberg 2011). Many immigrant workers are hyper aware of the skills they lack, and are eager to pursue positions outside their co-​ethnic niches, which research has repeatedly confirmed can frequently lead to stagnant or downward mobility (Catanzarite and Aguilera 2002; Xie and Gough 2011). One immigrant cashier at a university museum café described a supportive manager that hired her despite her limited English, and even accommodated her medical leave during a high-​risk pregnancy. Though she admits her job is simple, she takes pride in the array of tasks she regularly performs. ‘For example, if someone needs to make a sandwich, I go make it. If we’re out of chips, I find them . . . I know the work is simple, but I like it.’ Despite her fear of taking on a position that required regular customer interaction, she felt empowered by her manager’s support and looked forward to returning to her job.

Job Precarity and Fear of Job Loss Surveyed workers frequently relied on friends to help them find work, and feared losing their jobs. For undocumented workers, job loss not only meant losing their economic livelihood but having to navigate the I-​9 form requirements or finding a prospective employer who would overlook their legal status. Social networks were crucial to finding work in Philadelphia’s competitive labour market. Guadalupe explained that finding work had been difficult for her, ‘If you don’t already know someone there, it’s really hard to get a call back or an opportunity to work there. I’ve looked for other jobs but they tell you that, or you don’t have experience, or you don’t have a resume, and Americans ask you for papers and a social security number, and it’s hard.’

288    amada armenta and shannon gleeson Immigrant workers often described their employment as precarious, believing they were at risk of being fired or let go with no provocation. Marisol explained that this left her feeling powerless. ‘I always say-​here, no one is indispensable. They make you feel disposable, like it doesn’t matter how hard you work, someone is going to come behind you and do the same or do it better. So, if you don’t work really hard they’ll fire you, when actually—​either way, they’ll fire you. That’s the most frustrating thing—​you live through these things here and think, maybe I should have stayed in my country—​but then you realize that what you make there isn’t enough to live,’ she explained. Indeed, emergencies or accommodations were often met with dismissal. For example, 29-​year-​old Juana has lived in Philadelphia for 14 years. With the help of friends, she landed a job at a laundromat. She hated it, and eventually left for another laundry job with better work conditions. While Juana liked this job better, she was effectively banned from taking time off or switching shifts, because her employer refused to hire any additional workers. When Juana got sick, after two years of working at the laundromat, she was forced to miss an entire week of work. On Juana’s second sick day, her employers replaced her. Immigrant workers were often also not allowed the same privileges as other staff, resulting in additional surveillance and discipline. In ten years of working and living in Philadelphia, Soledad was fired three times. In her first job, as a prep cook in a kitchen, Soledad was fired after she ate a piece of bacon. ‘In that restaurant, Latinos didn’t have the right to eat’, she explained. ‘They wouldn’t even give Mexicanos a twenty-​minute break to eat. One day, I ate a piece of bacon because I had been working five hours and I was starving. I grabbed a piece as it was coming out of the oven and the manager saw me and fired me on the spot.’ This job instability stems from the basic tenets of at-​will employment in the United States, and are exacerbated for immigrants like Soledad who have few other options for stable work. While Soledad did find another job, she was again fired after a few days for not supplying a social security number. While a clear requirement of federal law, employers follow these provisions selectively and opportunistically. Soledad found work in another kitchen, but after working there for two weeks, they told her she was ‘no longer needed’ because a ‘better’ person with more experience was coming to replace her.

Navigating Hostile Work Environments Respondents’ relationships with their co-​workers also determined whether respondents felt comfortable at work. Scholars of immigrant labour such as Waldinger and Licther (2003) have highlighted that in jobs where African Americans and immigrants work alongside each other, tensions often arise. This has been further documented in new immigrant destinations such as the South where employment prospects for low-​skilled Blacks have had to compete with an influx of new immigrants (Ribas 2015). In this South Philadelphia sample, most respondents reported amicable co-​worker relationships, though conflicts did arise, with few formal options for redress.

immigration and job quality    289 Women, in particular, reported leaving positions because of hostile work environments. Sexual harassment is rampant in many industries, and despite equal protections for immigrants, immigrant women are particularly vulnerable (Vellos 1996; Marshall 2005). For example, Esperanza was 23 years old when she started working as a prep cook in a busy Philadelphia restaurant. At first, she was grateful when a male co-​ worker helped her learn the ropes of her job, but soon his ‘help’ turned into unwanted romantic advances: I was new there and this guy had been working there for a long time, so he was showing me how to do things. It was helpful—​he told me where things were and showed me how to set things up. He was friendly. But then, he started asking me ‘are you married?’ or ‘do you want to go out tonight or tomorrow?’ and when I kept on politely refusing or saying I was busy, he got mad and was like ‘Oh, if you don’t want to go out with me, then you’ll have to pay me in work.’

After Esperanza repeatedly turned down her co-​worker’s advances, he started insisting that she do harder tasks. When she spoke up in protest, he threatened to tell the manager that she did not want to do her job. Eventually, Esperanza left the job because she felt so uncomfortable, explaining that while she needed work, ‘I just didn’t need that kind of job.’ Her experience is not uncommon in the restaurant industry (The Restaurant Opportunities Centers United Forward Together 2014), and immigrant women of colour face particular challenges in confronting such abuse (Welsh et al. 2006). Yet our data confirm that while respondents believed that keeping their jobs was important, they also exercised agency, leaving jobs that where they felt coerced, undervalued, or exploited. Paulina had lived in Philadelphia for four years when she was interviewed. She described leaving various jobs because of hostility that she experienced from male and female co-​workers. In her first job, a female co-​worker spoke down to her, gave her difficult tasks to do, and yelled at her when a male colleague helped her carry a heavy box of tomatoes up a flight of stairs. Paulina recalled her co-​worker’s words, ‘He shouldn’t be doing your job. This is your job—​they’re paying you to do it. If you’re going to have everyone helping you with your work, you’re going to have to pay them.’ Co-​worker relationships were often complex and volatile, shaped by shifting dynamics of competition and solidarity. For example, while Paulina initially felt resentful, she eventually came to understand that her co-​worker was giving her a lesson about how to interact with male co-​workers and stave off sexual harassment. Her co-​worker explained to her bluntly, ‘You don’t understand how it works here. If a man offers to help you it’s because they’re going to charge you later . . . You need to be stronger because here there are many men and all they want to do is sleep with you. You feel new here and maybe you feel special because of that, but you’re not special. Later, someone new will come and the same thing will happen (to them).’ Although Paulina did not experience unwanted advances from male colleagues, her co-​worker’s accusatory attitude made her feel mistreated and unsupported at work. Unhappy with the environment, Paulina left her first job and started working at a restaurant where her cousin was employed.

290    amada armenta and shannon gleeson Unfortunately, her co-​worker’s comments would turn out to be prescient, because in Paulina’s new job, she did experience sexual harassment. She explained: One of my coworkers, he was also Mexican. He was older and married. One day he cornered me in the walk-​in freezer and said, ‘I know you’re a whore. How much do you want—​how much can I pay you to sleep with me?’ I was shocked. I couldn’t—​I couldn’t believe it was happening. I didn’t know whether I should yell or cry. I felt frozen. I finally answered, ‘I think you’ve confused me for someone else. I don’t know why you’re saying this to me—​but if you try anything—​I’m telling the boss.’ He grabbed me and started laughing and said, ‘You don’t have to be here. I’ll take care of you, if you give me what I want.’ I told him I wasn’t single and to get away from me. I tried to tell the boss, but at that time, I didn’t speak English and he didn’t really understand me. My boss said that he couldn’t let anyone go. That’s when I learned that here, you’re on your own. No one is going to look out for you. I left that job, even though my girlfriends told me I should stay because the job paid well. I just couldn’t take it.

Paulina explained that her work experiences left her feeling depressed and hopeless. Though sexual harassment protections would have extended to her, navigating the claims bureaucracy for sexual harassment is cumbersome and without leadership from management, especially so (Marshall 2005). She considered returning home to Mexico, but her family encouraged her to stay and try to better her circumstances. Rather than return to work immediately, Paulina enrolled in English classes. Within a few months, Paulina found a new restaurant job with amicable co-​workers who were happy to help her practise her English skills.

Women in the Informal Economy and Entrepreneurship Over 90 per cent of women whom we interviewed were mothers, with children ranging in age from recently born to adults with children of their own. While men worked consistently, pregnancy and childbirth changed women’s work trajectories significantly. Nationwide, research shows that more recently arrived immigrant women were more likely to cut back on paid employment following birth, but in time, more established immigrant women are actually less likely to quit or move to part time employment (Lu et al. 2019). In our sample, some women chose (or were compelled) to exit the labour market because they wanted to stay home with their children, others could not find affordable childcare, and some women were pressured by their husbands to opt out of paid work. However, as children became school-​aged and women had more free time, some chose to re-​enter the formal labour force, while others opted for informal work. While working informally may not pay as well as formal work, these efforts can be desirable for immigrant women who require flexibility for childcare (Portes and Haller 2010). For example, 33-​year-​old Rafaela sells food in the park while her husband plays soccer. This has provided an important space for socializing and income generation from April

immigration and job quality    291 to November. ‘Everyone comes to watch games, and lots of people sell stuff. I sell food. It’s really nice to go to the park and walk around—​it’s relaxing’ she explained. While social scientists often draw stark distinctions between private and market activity, Rafaela’s example suggests that these boundaries may often blur (López-​Garza 2002). The informal economy provides some immigrant women with a strategy to cope with breaks in formal paid work, often to care for young children. Angelica, 33 years old, occasionally sells food when she has breaks in child care. She explains, ‘When my kids allow me to, I make food and sell it. It brings in a little extra money that helps us.’ For others like Ximena, who came to Philadelphia when she was 15, this supplementary income allows her to send money back to Mexico. She explained, ‘I used to work in a factory, then I worked at a restaurant. I stopped when I had my daughter. I didn’t have to work anymore because her dad worked. Now, I work a little on the side, I sell food and I use that to send money home-​to help my mom and my brothers.’ Some immigrant women experience these entrepreneurial ventures as a way to gain agency over their financial means. However, this can sometimes mean engaging in controversial multilevel marketing schemes (aka ‘direct sales’) such as Herbalife, Amway, and Mary Kay, which have garnered criticism for their exploitative practices (Palomarez 2016). For example, 39-​year-​old Gabriela moved to Philadelphia when she was 23 and sells Herbalife part-​time. She sees this as a pathway to self-​development and a way to build social capital and to allow her to also care for her kids. ‘I sell nutrition shakes for Herbalife. Women come to see me and I get to know them . . . I like the women who see everything positively. I sell shakes and I learn about them because I like it. It’s important to work, go out, sell, and do something with one’s life, and overcome things.’ Similarly, Carmen had a series of jobs before having kids, and appreciates the autonomy her sales venture provides. She learned about Mary Kay from a friend, but admits that it can be challenging to make a profit. ‘[People don’t really want to buy much because they think it’s too expensive, so it’s kind of hard to sell . . . but I like it. It’s . . . it’s an opportunity to go out and sell without someone looking over your shoulder and telling you what to do.’ Self-​employment also seemed an attractive option for mothers whose husbands pressured them to stay home. For example, 37-​year-​old Ana navigated rigid gender roles in her household, including a husband who did not want her to work. She spent most of her time with her children and valued flexibility with her schedule. ‘Once they’re bigger and they’re in school, maybe I’ll do something else or keep selling Mary Kay, but do it for more hours . . . While my kids are young, they’re my priority—​for now, this is my work because it allows me to have control of my schedule.’ These gender relations, however, do not compete in a vacuum. As Espiritu (2003) finds in her study of Asian entrepreneurship, immigrant women’s subordinate position in US society more broadly limits the transformative potential that their newfound agency at work can provide in the home. Many of these schemes also thrive on their transnational presence, and so were very familiar to immigrant communities. FRIDA, a 35-​year-​old mother and saleswoman explains how she was able to balance raising her young son with work. ‘When I arrived I didn’t work, because I came with my son and he was small. Eventually, I started selling Mary Kay because it’s what I did in Mexico. It works well because I don’t have to pay

292    amada armenta and shannon gleeson a baby-​sitter. I take my kids with me when I go out and sell my products—​I’ve been doing it since they were small. I make a little extra money.’ There is generally mixed evidence regarding the sustainability of entrepreneurship as a viable alternative to good jobs in the formal sector for immigrants (Kloosterman and Rath 2001; Portes et al. 2002; Waldinger and Fitzgerald 2004). In sum, for these undocumented women, and many other immigrants, the informal economy and direct sales became a key—​albeit often exploitative—​way to generate income while balancing childcare needs, traditional gender roles, and limited human capital skills. A broader literature on immigrant entrepreneurship also points to important intersectional effects of race, class and gender for understanding the potential of entrepreneurship for social mobility (Valdez 2011).

Conclusion This chapter has highlighted the challenges shaping job quality for immigrants, with a focus on undocumented Latino immigrant workers in Philadelphia. The data we present points to the limits of city ‘sanctuary’ policies insofar as undocumented immigrants experience significant challenges related to wage theft, job precarity and workplace harassment. While the federal government lost its bid to cut off Philadelphia from federal funding after a judge ruled that Philadelphia’s immigration policies were reasonable, new reporting suggests that ICE has responded to a lack of cooperation from municipal governments by targeting those cities for enforcement raids and sweeps. Indeed, the Philadelphia ICE office is amongst the most active in the country (Sontag and Russakoff 2018). Moreover, the federal government has resumed large-​scale workplace raids. As the policy context for immigrant workers continues to evolve, and the prospects of immigration reform disappeared under the Trump administration, additional comparative research is necessary to understand the factors that shape immigrant economic integration and exploitation. While a great deal of research has focused on the United States, Europe and Canada also provide instructive comparisons (Goldring et al. 2009; McKay et al. 2012; Card and Raphael 2013). The issue of legal status in particular deserves additional attention, given its highly politicized nature. However, it is important to understand the economic condition of immigrants beyond the immigration regime, and to also take into account other factors shaping immigrant precarity at both the ‘low’ and ‘high’ skills ranks of the immigrant labour force. In effect, a ‘continuum of exploitation’ exists even for high-​earning immigrants who enjoy preferred legal status (Ontiveros 2016). Immigration politics in the United States remains uncertain for immigrant workers, despite changes in presidential administrations. According to Gleeson and Sampat (2018), the Trump era brought ‘a vast expansion of the groups prioritized for deportation, a more aggressive arrest strategy at places previously considered safe, a plan to hire fifteen thousand more immigration agents, a broad ban on refugees and even basic

immigration and job quality    293 travel (including from several majority-​Muslim countries), and the creation of an office of Victims of Immigrant Crime Enforcement (87).’ As of this writing, the current political landscape does not suggest a likely way forward for substantial judicial and legislative reform. The Trump administration loomed large in the years following this study, and pushed an expansion of the southern border wall, which proponents argued would act as a deterrent to unauthorized migration. A key rationale was the poorly supported premise that immigrant labour unequivocally drives down wages and takes jobs from native-​ born workers (Ottaviano and Peri 2012; Constant 2014). While not unprecedented throughout history, one of the most consistent impacts of such walls and fences are an increase in the deaths of migrants in transit, 5,604 across the world in 2015. In the United States and Europe especially, evidence suggests that the immediate effect of these structures (and the militarized personnel that often come with them) is a shift to more remote crossing points, accounting for the increase in deaths. In the years after the border hardened in Arizona, the coroner recorded on average 200 deaths (a tenfold increase from the decade prior). That same year, deaths in the European Union numbered in the thousands (Jones 2016). Meanwhile, the hardening of interior enforcement mechanisms, which have long operated aggressively at the workplace, paired with insufficient workplace enforcement, and the brutality of capitalist labour relations, have translated into poorer—​and sometimes even fatal—​workplace conditions for immigrant workers (Moyce and Schenker 2018). As such, any analysis of immigrant job quality should also pay close attention to the role of the state in not only regulating workplace protections and holding employers compliant, but also in shaping a public discourse around the disposability of immigrant labour. This is not limited to the United States, but to be sure has become a central platform for nativist political parties in Europe and beyond (Ono and Ono 2014; Hogan and Haltinner 2015). Acting as a counterbalance to this repression are various nodes of immigrant mobilization that have won important gains within the labour movement and around broader issues of immigration reform and well-​being as well (Voss and Bloemraad 2011; Milkman 2011; Barron et al.2016). As social stratification scholars examine the nature of immigrant job quality in a globalizing economy, these state-​ society relations will remain crucial to reshaping both the economic and political power immigrants hold in host societies.

References Abrego, Leisy J., and Lakhani, Sarah M. (2015), ‘Incomplete Inclusion: Legal Violence and Immigrants in Liminal Legal Statuses.’ Law & Policy, 37/​4, 265–​93. Adler, Lee H., Tapia, Maite, and Turner, Lowell (eds.) (2014), Mobilizing against Inequality: Unions, Immigrant Workers, and the Crisis of Capitalism. Frank W. Pierce Memorial Lectureship and Conference Series (Ithaca, NY: Cornell University Press). American Immigration Council (2015), ‘New Americans in Pennsylvania.’ Fact Sheet https:// www.americanimmigrationcouncil.org/research/immigrants-in-pennsylvania.

294    amada armenta and shannon gleeson Armenta, Amada (2016), ‘Between Public Service and Social Control: Policing Dilemmas in the Era of Immigration Enforcement’, Social Problems, 63/​1, 111–​126. Armenta, Amada (2017), ‘Racializing Crimmigration, Structural Racism, Colorblindness, and the Institutional Production of Immigrant Criminality’, Sociology of Race and Ethnicity, 3/​ 1, 82–​95. Barron, Pierre, Bory, Anne, Chauvin, Sébastien, Jounin, Nicolas, and Tourette, Lucie (2016), ‘State Categories and Labour Protest: Migrant Workers and the Fight for Legal Status in France’, Work, Employment & Society, 30/​4, 631–​648. Bernhardt, Annette, Milkman, Ruth, Theodore, Nik, Heckathorn, Douglas, Auer, Mirabai, and DeFilippis James et al. (2009), Broken Laws, Unprotected Workers: Violations of Employment and Labor Laws in America’s Cities, Center for Urban Economic Development, National Employment Law Project, and the UCLA Institute for Research on Labor and Employment, http://​nelp.3cdn.net/​31998​2941​a549​6c74​1_​9q​m6b9​2kg.pdf. Bernhardt, Annette, Spiller, Michael W., and Polson, Diana (2013), ‘All Work and No Pay: Violations of Employment and Labor Laws in Chicago, Los Angeles and New York City’, Social Forces 91/​3, 725–​46. Bradley, David H. (2015), The Federal Minimum Wage: In Brief, Washington, DC: Congressional Research Service, http://​www.fas.org/​sgp/​crs/​misc/​R43​089.pdf. Brookings Institution (2003), Philadelphia in Focus: A Profile from Census 2000, https://​www. brooki​ngs.edu/​resea​rch/​phila​delp​hia-​in-​focus-​a-​prof​i le-​from-​cen​sus-​2000/​. Bureau of Labor Statistics (2016), ‘Labor Force Characteristics of Foreign-​Born Workers Summary’. USDL-​16-​0989. Economic News Release. United States Department of Labor. Burnham, Linda, and Theodore, Nik (2012), Home Economics: The Invisible and Unregulated World of Domestic Work, New York, http://​www.idw​fed.org/​en/​resour​ces/​home-​econom​ ics-​the-​invisi​ble-​and-​unre​gula​ted-​world-​of-​domes​tic-​work/​@@disp​lay-​file/​attac​hmen​t_​1. Card, David, and Raphael, Steven (2013), Immigration, Poverty, and Socioeconomic Inequality (New York: Russell Sage Foundation). Catanzarite, Lisa, and Michael Bernabé Aguilera (2002), ‘Working with Co-​Ethnics: Earnings Penalties for Latino Immigrants at Latino Jobsites’, Social Problems 49/​1, 101–​127. Constant, Amelie F. (2014), Do Migrants Take the Jobs of Native Workers? IZA World of Labor, 10 (May), http://​wol.iza.org/​artic​les/​do-​migra​nts-​take-​the-​jobs-​of-​nat​ive-​work​ers/​long. Doussard, Marc (2013), Degraded Work: The Struggle at the Bottom of the Labor Market (Minneapolis: University of Minnesota Press). Espiritu, Yen Le (2003), ‘Gender and Labor in Asian Immigrant Families’ in Pierrette Hondagneu-​Sotelo, ed., Gender and US Immigration: Contemporary Trends (Berkeley: University of California Press), 81–​100. Fine, Janice (2006), Worker Centers: Organizing Communities at the Edge of the Dream (Ithaca: ILR Press). Fisk, Catherine L., and Wishnie, Michael J. (2005), ‘The Story of Hoffman Plastic Compounds, Inc. v. NLRB: Labor Rights Without Remedies for Undocumented Immigrants’, in Laura J. Cooper and Catherine L. Fisk, eds., Labor Law Stories (New York, NY: Foundation Press), 399–​438 Fussell, Elizabeth (2011), ‘The Deportation Threat Dynamic and Victimization of Latino Migrants: Wage Theft and Robbery’, The Sociological Quarterly 52/​4, 593–​615. Garcia, Ruben J. (2012), Marginal Workers: How Legal Fault Lines Divide Workers and Leave Them Without Protection (New York: NYU Press).

immigration and job quality    295 Gleeson, Shannon (2012a), Conflicting Commitments: The Politics of Enforcing Immigrant Worker Rights in San Jose and Houston (Ithaca, NY: Cornell University Press). Gleeson, Shannon (2012b), ‘Leveraging Health Capital at the Workplace: An Examination of Health Reporting Behavior among Latino Immigrant Restaurant Workers in the United States’, Social Science & Medicine, 75/​12, 2291–​2298. Gleeson, Shannon (2015), ‘Between Support and Shame: The Impacts of Workplace Violations for Immigrant Families’, in Jody Agius Vallejo, ed., Research in the Sociology of Work: Immigration and Work, 27 (Bingley: Emerald Group Publishing), 29–​52. Gleeson, Shannon (2016), Precarious Claims: The Promise and Failure of Workplace Protections in the United States (Oakland, CA: University of California Press). Gleeson, Shannon, and Sampat, Prerna (2018), ‘Immigrant Resistance in the Age of Trump’, New Labor Forum 27/​1, 86–​95. Golash-​ Boza, Tanya Maria (2015), Race and Racisms: A Critical Approach (Oxford, New York: Oxford University Press). Goldring, Luin, Berinstein, Carolina, and Bernhard, Judith K. (2009), ‘Institutionalizing Precarious Migratory Status in Canada’, Citizenship Studies 13/​3, 239–​265. Goldring, Luin, and Landolt, Patricia (2011), ‘Caught in the Work–​Citizenship Matrix: The Lasting Effects of Precarious Legal Status on Work for Toronto Immigrants’, Globalizations 8/​3, 325–​341. Gomberg-​Muñoz, Ruth (2011), Labor and Legality: An Ethnography of a Mexican Immigrant Network (New York: Oxford University Press). Gordon, Jennifer (2007), Suburban Sweatshops: The Fight for Immigrant Rights (Cambridge, MA: Belknap Press). de Graauw, Els (2016), Making Immigrant Rights Real: Nonprofits and the Politics of Integration in San Francisco (Ithaca, NY: Cornell University Press). Griffith, Kati L. (2011), ‘Discovering ‘Immployment’ Law: The Constitutionality of Subfederal Immigration Regulation at Work’, Yale Law and Policy Review, 29, 389–​451. Griffith, Kati L. (2012), ‘Undocumented Workers: Crossing the Borders of Immigration and Workplace Law’, Cornell Journal of Law and Public Policy, 21, 611–​97. Griffith, Kati L. (2017), ‘The Power of a Presumption: California as a Laboratory for Unauthorized Immigrant Workers’ Rights’, UC Davis Law Review, 50, 1279–​1322. Griffith, Kati L., and Gleeson, Shannon M. (2017), ‘The Precarity of Temporality: How Law Inhibits Immigrant Worker Claims’, Comparative Labor Law and Policy Journal, 39/​1, 111–​141. Griffith, Kati L., and Gleeson, Shannon (2019), ‘Immigration Enforcement and the Employment Sphere: Unpacking Trump-​ Era ‘Immployment’ Law’, Southwestern Law Review 48/​3, 475–​501. Hall, Matthew, and Greenman, Emily (2014), ‘The Occupational Cost of Being Illegal in the United States: Legal Status, Job Hazards, and Compensating Differentials’, International Migration Review (April), 49/​2, 1–​37. Heeren, Geoffrey (2015), ‘The Status of Nonstatus’, American University Law Review, 64, 1115–​1181. Hogan, Jackie, and Haltinner, Kristin (2015), ‘Floods, Invaders, and Parasites: Immigration Threat Narratives and Right-​Wing Populism in the USA, UK and Australia’, Journal of Intercultural Studies, 36/​5, 520–​543. Holmes, Seth (2013), Fresh Fruit, Broken Bodies: Migrant Farmworkers in the United States. Vol. 27 - California Series in Public Anthropology. (Berkeley: University of California Press).

296    amada armenta and shannon gleeson Holmes, Seth M. (2007). ‘ “Oaxacans Like to Work Bent Over”: The Naturalization of Social Suffering among Berry Farm Workers’, International Migration 45/​3, 39–​68. Jones, Reece (2016), ‘Borders and Walls: Do Barriers Deter Unauthorized Migration?’ Migration Policy Institute, https://www.migrationpolicy.org/article/borders-and-walls-dobarriers-deter-unauthorized-migration. Kalleberg, Arne L. (2011). Good Jobs, Bad Jobs: The Rise of Polarized and Precarious Employment Systems in the United States, 1970s-​2000s (New York: Russell Sage Foundation). Kerwin, Donald M, and McCabe, Kristen (2011), Labor Standards Enforcement and Low-​Wage Immigrants: Creating an Effective Enforcement System, Washington, DC, Migration Policy Institute, http://​www.migr​atio​npol​icy.org/​pubs/​lab​orst​anda​rds-​2011.pdf. Kloosterman, Robert, and Rath, Jan (2001), ‘Immigrant Entrepreneurs in Advanced Economies: Mixed Embeddedness Further Explored’, Journal of Ethnic and Migration Studies, 27/​2, 189–​201. Kugler, Adriana, Lynch, Robert, and Oakford, Patrick (2013), Improving Lives, Strengthening Finances: The Benefits of Immigration Reform to Social Security, Washington, DC., Center for American Progress, https://​www.ameri​canp​rogr​ess.org/​iss​ues/​immi​grat​ion/​rep​ort/​2013/​ 06/​14/​66543/​improv​ing-​lives-​streng​then​ing-​finan​ces-​the-​benef​its-​of-​immi​grat​ion-​ref​ orm-​to-​soc​ial-​secur​ity/​. Lakhani, Sarah Morando (2013), ‘Producing Immigrant Victims’ “Right” to Legal Status and the Management of Legal Uncertainty’, Law & Social Inquiry, 38/​2, 442–​473. López-​Garza, Marta (2002), ‘Convergence of the Public and Private Spheres: Women in the Informal Economy’, Race, Gender & Class 9/​3, 175–​192. Lu, Yao, Julia Shu-​Huah Wang, and Wen-​Jui Han. (2017), ‘Women’s Short-​Term Employment Trajectories Following Birth: Patterns, Determinants, and Variations by Race/​Ethnicity and Nativity’, Demography, 54/​1, 93–​118. Marshall, Anna-​ Maria (2005), ‘Idle Rights: Employees’ Rights Consciousness and the Construction of Sexual Harassment Policies’, Law & Society Review, 39/​1, 83–​124. McKay, Sonia, Markova, Eugenia, and Paraskevopoulou, Anna (2012), Undocumented Workers’ Transitions: Legal Status, Migration, and Work in Europe (New York, NY: Routledge). Medvedeva, Maria, and Portes, Alejandro (2017), ‘Immigrant Bilingualism in Spain: An Asset or a Liability?’ International Migration Review, 51/​3, 632–​666. Milkman, Ruth (2006), L.A. Story: Immigrant Workers and the Future of the U.S. Labor Movement (New York: Russell Sage Foundation Press). Milkman, Ruth (2011), ‘Immigrant Workers, Precarious Work, and the US Labor Movement’, Globalizations 8/​3, 361–​372. Moyce, Sally C., and Schenker, Marc (2018), ‘Migrant Workers and Their Occupational Health and Safety’, Annual Review of Public Health, 39, 351–​365. National Employment Law Project (2013), California’s New Worker Protections Against Retaliation, http://​www.nelp.org/​page/​-​/J​ ust​ice/​2013/​ca-​wor​ker-​prot​ecti​ons-​agai​nst-​reta​liat​ion.pdf. National Employment Law Project (2015), City Minimum Wage Laws: Recent Trends and Economic Evidence, Minimum Wage Basics, http://​www.nelp.org/​cont​ent/​uplo​ads/​City-​ Mini​mum-​Wage-​Laws-​Rec​ent-​Tre​nds-​Econo​mic-​Evide​nce.pdf. Ono, Hiroshi, and Ono, Hiromi (2014), ‘Anti-​ Immigrant Sentiments and Immigrant Concentration at Work in Contemporary Japan’, in Shirley A. Jackson, ed., Routledge International Handbook of Race, Class, and Gender (New York: Routledge), 61–​72. Ontiveros, Maria Linda (2016), ‘H-​1B Visas, Outsourcing and Body Shops: A Continuum of Exploitation for High Tech Workers’, Berkeley Journal of Employment and Labor Law, 38/​1, 1–​47.

immigration and job quality    297 Orrenius, Pia M., and Zavodny, Madeline (2009), ‘Do Immigrants Work in Riskier Jobs?’ Demography, 46/​3, 535–​551. Ottaviano, Gianmarco I. P., and Peri, Giovanni (2012), ‘Rethinking the Effect of Immigration on Wages’, Journal of the European Economic Association, 10/​1, 152–​197. Otterbein, Holly (2015), ‘Here’s the Gap Between the Minimum Wage and Cost of Living in Philly’, Philadelphia Magazine. September 11, 2015. https://​www.philly​mag.com/​citif​i ed/​ 2015/​09/​11/​mini​mum-​wage-​liv​ing-​wage-​map/​. Padua, Larissa A., and Doran, Joanna K. (2016), ‘From Being Unbanked to Becoming Unbanked or Unbankable: Community Experts Describe Financial Practices of Latinos in East Los Angeles’, Journal of Community Practice, 24, 428–​444. Palomarez, Javier. (2016), FTC Right, Congress Wrong on Multi-​Level Marketing Giants Like Herbalife, December 6, http://​ori​gin-​nyi.theh​ill.com/​blogs/​pund​its-​blog/​hea​lthc​are/​309​ 033-​ftc-​right-​congr​ess-​wrong-​on-​herbal​ife. Paret, Marcel, and Gleeson, Shannon (2016), ‘Precarity and Agency Through a Migration Lens’, Citizenship Studies, 20/​3–​4, 277–​294. Passel, Jeffrey S., and D’vera Cohn. 2016. “Size of U.S. Unauthorized Immigrant Workforce Stable After the Great Recession.” Pew Research Center’s Hispanic Trends Project (blog). November 3, 2016. https://www.pewresearch.org/hispanic/2016/11/03/ size-of-u-s-unauthorized-immigrant-workforce-stable-after-the-great-recession/. Petrescu-​Prahova, Miruna, and Spiller, Michael W. (2016), ‘Women’s Wage Theft Explaining Gender Differences in Violations of Wage and Hour Laws’, Work and Occupations, 43/​4, 371–​400. Pham, Huyen, and Van, Pham Hoang (2010), ‘Economic Impact of Local Immigration Regulation: An Empirical Analysis’, Immigration and Nationality Law Review, 31, 687–​687. Piore, Michael J. (1979), Birds of Passage: Migrant Labor and Industrial Societies (Cambridge: Cambridge University Press). Portes, Alejandro, Guarnizo, Luis E., and Haller, William J. (2002), ‘Transnational Entrepreneurs: An Alternative Form of Immigrant Economic Adaptation’, American Sociological Review, 67/​2, 278–​298. Portes, Alejandro, and Haller, William (2005), ‘The Informal Economy’ in Neil J. Smelser and Richard Swedberg, eds., The Handbook of Economic Sociology, Second Edition (Princeton/​ New York: Princeton University Press/​Russell Sage Foundation), 403–​425. Ribas, Vanesa (2015), On the Line: Slaughterhouse Lives and the Making of the New South (Oakland, CA: University of California Press). Riosmena, Fernando, and Massey, Douglas S (2012), ‘Pathways to El Norte: Origins, Destinations, and Characteristics of Mexican Migrants to the United States’, The International Migration Review, 46/​1, 3–​36. Sasinovic, Miranda (2021), ‘Removing Roadblocks: Alternatives to Lawful Status and Social Security Number Requirements for Pennsylvania Driver’s Licenses’, Dickinson Law Review, 126/​1, 305. Sassen, Saskia (2014), Expulsions: Brutality and Complexity in the Global Economy (Cambridge, MA: Harvard University Press). Shusterman Law (2018), Visa Bulletin for May 2018, Check Your Priority Date, https://​www.shu​ ster​man.com/​stat​edep​artm​entv​isab​ulle​tin/​. Singer, Audrey (2015), ‘Metropolitan Immigrant Gateways Revisited, 2014’, Washington, DC: Brookings Institution https://www.brookings.edu/research/metropolitan-immigrantgateways-revisited-2014/.

298    amada armenta and shannon gleeson Singer, Audrey, Vitiello, Domenic, Katz, Michael, and Park, David (2008), Recent Immigration to Philadelphia: Regional Change in a Re-​Emerging Gateway. Washington, DC: Brookings Institution. https://​www.brooki​ngs.edu/​wp-​cont​ent/​uplo​ads/​2016/​06/​1113​_​imm​igra​tion​_​ sin​ger.pdf . Stuesse, Angela, and Coleman, Mathew (2014), ‘Automobility, Immobility, Altermobility: Surviving and Resisting the Intensification of Immigrant Policing’, City & Society, 26/​1, 51–​72. Sontag, Deborah and Russakoff, Dale (2018), No Sanctuary: In Pennsylvania, It’s Open Season on Undocumented Immigrants, New York, ProPublica, https://​www.pro​publ​ica.org/​arti​cle/​ penns​ylva​nia-​ice-​undoc​umen​ted-​imm​igra​nts-​immi​grat​ion-​enfo​rcem​ent. The Legal Aid Society-​Employment Law Center (2015), Undocumented Workers: Employment Rights, https://​las-​elc.org/​fact-​she​ets/​undoc​umen​ted-​work​ers-​emp​loym​ent-​rig​hts. The Pew Charitable Trusts (2020), ‘How Philadelphia’s Minimum Wage Compares With Other U.S. Cities’, https://​pew.org/​3fOS​0hD. The Restaurant Opportunities Centers United Forward Together (2014), ‘The Glass Floor: Sexual Harassment in the Restaurant Industry.’ New York, NY. U.S. Citizenship and Immigration Services (2017), E-​Verify Facts and Stats: Archive, https://​ www.uscis.gov/​e-​ver​ify/​e-​ver​ify-​facts-​and-​stats-​arch​ive. Valdez, Zulema (2011), The New Entrepreneurs: How Race, Class, and Gender Shape American Enterprise (Stanford: Stanford University Press). Valenzuela, Abel Jr., Theodore, Nik, Melendez, Edwin, and Gonzalez, Ana Luz (2006), ‘On the Corner: Day Labor in the United States’, Los Angeles: UCLA Center for the Study of Urban Poverty, http://​www.urban​econ​omy.org/​sites/​defa​ult/​files/​onthec​orne​r_​da​ylab​orin​US_​3​ 9p_​20 ​ 06.pdf. Vellos, Diana (1996), ‘Immigrant Latina Domestic Workers and Sexual Harassment’, American University Journal of Gender, Social Policy & the Law, 5, 407–​32. Voss, Kim, and Bloemraad, Irene (2011), Rallying for Immigrant Rights (Berkeley: University of California Press). Wage and Hour Division (2015), ‘Minimum Wage Laws in the States -​January 1, 2015’, Washington, DC: United States Department of Labor, http://​www.dol.gov/​whd/​minw​age/​ amer​ica.htm. Waldinger, Roger D, and Fitzgerald, David (2004), ‘Transnationalism in Question’, American Journal of Sociology, 109/​5, 1177–​1195. Waldinger, Roger, and Lichter, Michael (2003), ‘How the Other Half Works: Immigration and the Social Organization of Labor (Chapter 1)’, How the Other Half Works: Immigration and the Social Organization of Labor (Berkeley and Los Angeles: University of California Press). Weil, David (2014), The Fissured Workplace: Why Work Became So Bad for So Many and What Can Be Done to Improve It (Cambridge, MA: Harvard University Press). Welsh, Sandy, Carr, Jacquie, MacQuarrie, Barbara, and Huntley, Audrey (2006), ‘‘I’m Not Thinking of It as Sexual Harassment’: Understanding Harassment across Race and Citizenship’, Gender & Society, 20/​1, 87–​107. Xie, Yu, and Gough, Margaret (2011), ‘Ethnic Enclaves and the Earnings of Immigrants’, Demography, 48/​4, 1293–​1315.

chapter 13

job quality for service and care occupations A feminist perspective orly benjamin

Preface Quality jobs for women in service and care occupations constitute a current central consensual political requirement of feminist scholars around the world supporting the international campaign for a caring economy and a caring economics. Leading the campaign are Nancy Folbre with her recent The Rise and Decline of Patriarchal Systems: An Intersectional Political Economy (2021), Susan Himmelweit with her co-​authored A care-​ led recovery from covid-​19: Investing in high-​quality care to stimulate and rebalance the economy (2021), Catherine Rottenberg with her co-​authored The care manifesto: The politics of interdependence (2020); and Joan Tronto with her Caring Democracy: Markets, Equality, and Justice. I dedicate my chapter to these leading feminist scholars. I deeply thank Rachel Dwyer, a co-​editor of this handbook who invited me to contribute my chapter and extend my gratitude to Hazel Conley and Haya Stier, two important feminist researchers of gender inequality in the labour market and the welfare state, who commented on earlier versions.

Introduction Neo-​liberalization of the Global North is historically concurrent with middle-​class women’s mass participation in the job market. The first indications of both processes began to appear in the mid-​1970s. Three decades later, during the mid-​2000s, three feminist volumes revealed the importance of job quality in women’s lives: Gendering the

300   orly benjamin Knowledge Economy: Comparative Perspectives (Walby 2007), People at Work: Life, Power and Social Inclusion in the New Economy (DeVault 2008) and The Gender Contours of Precarious Employment (Vosko et al. 2009). These volumes which describe women’s employment in a broad range of countries, noted the false nature of the employed/​unemployed dichotomy in women’s work histories. Each chapter, whether or not it actually mentioned the term ‘job quality’, contributed to its understanding and hinted at its ability to shed light on an extensive range of feminist issues, especially in-​employment poverty (Peña-​Casas and Ghailani 2011). In other words, for women, being employed often coexists with dependence on other providers and/​or state allowances and services. Feminist researchers examining this phenomenon show how crucial investigation of job quality is in allowing for cross-​sectional comparisons among women and for situating women’s work lives within broader policy contexts (e.g. Taniguchi and Rosenfeld 2002 and for the impact of austerity policy see Warren and Lyonnette, 2018). Job quality is commonly evaluated by both extrinsic and intrinsic parameters. The intrinsic ones are introduced together with an emphasis on subjective evaluation of job quality. Nevertheless, caring employees keep searching for alternative employment regardless of job satisfaction (Craft-​Morgan et al. 2013), that is shaped within ‘a range of socio-​economic constraints negotiated by workers when they make sense of their working lives’ (Hebson et al. 2015: 328). These persistent findings call for a feminist conceptualization of job quality that is grounded in women’s experiences of the undervaluation of the skills, knowledge and experience they possess that are typical of specific occupations and particular employment arrangements. My feminist perspective on job quality follows Eisenstein’s (2015) argument that economic globalization reproduces the exploitation of women’s labour. I argue that exploitation is accelerated by promoting undervaluation, deskilling and de-​professionalization in service and care occupations. This argument is supported by the notion of racialization developed by Edna Bonacich and colleagues (2008), who claim that typical low-​quality service jobs manifest forms of exploitation that treat women of the Global North in ways that resemble those inflicted on women in the Global South. Racialization requires a feminist perspective on job quality that is anchored in Davies’ (1996: 664) now classical statement: ‘[W]‌omen (or rather the qualities that women represent) are constructed as ‘devalued Other’, as carriers of qualities that thereby remain unacknowledged and denied.’ Job quality must be evaluated, then, in light of specific employment arrangements within occupations according to three dimensions: Remunerations, that is, the extent to which knowledge, skill and experience are reflected in payment scales; certification, namely, the extent to which on-​the-​job training practices allow both certified and uncertified employees to benefit from skill development; and, decommodification—​the benefits ensuring women’s income when unable to work because of caring emergencies, illness or old age. When these three aspects are minimized, low quality of jobs reproduces the gendered hierarchy that impels women towards their position of other in the job market. With the emerging difficulty of protecting recognition of required skill levels in service and care occupations, interest in the gendered nature of job quality is

job quality for service and care occupations    301 enhanced. A gender job quality index used in 2015 to compare men and women in tourism (Santero-​Sanchez et al. 2015) validated the significance of vertical segregation for job quality and showed that objective job security and stability conditions are higher for men. A more feminist approach pinpoints the need to evaluate job quality along gender-​specific dimensions. These are part of the job quality definition introduced and embraced by the EU (European Commission 2001—​cf. Davoine et al. 2008), whose origins lie in the indicators articulated at the EES Laeken Summit as a manifestation of the will to renew the European Social Model. The definition covers ten dimensions that were operationalized (Davoine et al. 2008) as six principal aspects: Socioeconomic security (i.e. decent wages and secure transitions), skills and training opportunities, working conditions (accidents, work intensity, long working hours, health risks), combining work and family life, promotion of gender equality and collective interest representations. This definition accords particular attention to issues of skill recognition and to secure transitions referring to family constraints in the lives of caring and service employees. Furthermore, it is identified explicitly with common conceptualization of gender issues, including promotion and collective interest representation. After reviewing recent theoretical developments in the field, I consider each of the dimensions of job quality included in the gender-​sensitive model proposed by Davoine and her colleagues (2008), with specific attention to employees in service and care occupations.

Job Quality in Service and Care Occupations Calls to build payment scales and raise the minimum job quality required for direct care workers by persistent and mandatory wage increases are voiced frequently (e.g. Dawson 2016). Such demands tend to criticize the prevailing but involuntary part-​time and hourly payment plans, dichotomies of either/​or job market participation, as well as extrinsic vs intrinsic job quality dimensions. These calls for reform raise issues of undervaluation of women’s work and the vulnerability of women dependent on low-​quality positions, leading employers to bully and abuse them—​by extracting unpaid work from them in various ways. Similarly, Rubery (2008) argued that there is a clear need to improve job quality for women employed externally in public services by protecting equal pay and wage grading for all outsourced staff. The paramount importance of employment arrangements in specific occupations is highlighted by Vidal (2013), who perceives the fragmentation that has already occurred in service and care occupations as a significant obstacle reflecting ‘the new dominant logic of externalized employment relations’ (p. 591). Employment arrangements were also emphasized by Australian scholars (Ravenswood and Kaines 2015) and in European studies (Jaehrling et al. 2016), repeatedly associating public procurement with deterioration in job quality.

302   orly benjamin Nevertheless, a recent account of the ways in which outsourcing in the US shapes job quality, published by prominent researchers of labour market processes somehow omitted job quality among service and care employees who typically rely on hourly payment (Bernhardt et al. 2016). This omission followed Kalleberg’s Good Jobs, Bad Jobs . . . (2011), a project that led prominent journals to launch special issues on job quality (Findlay et al. 2013). It noted deterioration of job quality in service and care occupations only indirectly by referring to the overall decline in the number of public-​ sector unions. Moreover, several accounts of job quality that focused on gender differences (e.g. Mühlau 2011; Gallie 2013) neglected to examine emerging employment arrangements in occupations dominated by women, not to mention the differences within them. Instead, Gallie (2013) proposed comparing men’s and women’s subjective work experiences with regard to direct participation. The recently validated association between job quality and level of technology at work (Stier 2014) clarifies: Technology may reduce the number of bad quality care and service jobs, but it cannot replace the need for interpersonal contact and best practice ethical standards (Tronto 1993, 2013). The type of employment arrangement has not been counted as a gender-​relevant dimension of job quality (Mühlau 2011), even though its detrimental impact on job quality, particularly temporary and involuntary part-​time work (Stecy-​Hildebrandt et al., 2019), underscores the importance of type of job contract (Kauhanen and Nätti 2015). Armstrong (2013), however, showed this association to be gendered: in service and care occupations, the quality of jobs depends on employment arrangements that allow for teamwork and autonomous discretion. Without them, jobs cannot withstand the need for constant development through working with other, often more experienced employees. Reinforcing this point, Stier and Yaish (2014) came exceptionally close to covering the unique nature of these fields by examining gender differences in job quality for occupational categories in 27 countries, one of which comprised ‘low blue-​collar’ (primarily unskilled occupations—​p. 230) with a high percentage of women employees and few employees with graduate degrees. Their ability to isolate this category enabled them to argue that ‘women lagged behind men in most employment dimensions’ (p. 241), including salaries, opportunities for advancement, job content, job security, time, autonomy and emotional conditions. Nevertheless, Stier and Yaish’s (2014) account still neglects to take specific employment arrangements into account and therefore remains mute concerning intra-​occupational differences that shed light on the implications of policy changes. The six dimensions reviewed below help elucidate the value of mediated employment within contracted service delivery in addressing these intra-​occupational differences.

Socioeconomic Security Once neo-​liberalism began to influence public-​sector reforms and welfare restructuring across OECD countries, the prevalence of principles of competition, profitability,

job quality for service and care occupations    303 managerialism and efficiency, as well as non-​standard and precarious employment, began to increase (Grimshaw et al. 2005). Marketization and competition have taken the form of reducing workforce expenses to promote services that are deemed ‘cheap’ by governments and users alike. Howes and colleagues (2012) showed that this approach lowered the quality of services and of the individual jobs that comprise them. Furthermore, this approach meant that part-​time work is assigned even to service and care employees who are required full time. Rubery et al. (2015) followed Supiot (2001) in explaining how part-​time employment is exploited to obtain unpaid work. They defined a time-fragmented employment arrangement to describe employers who use strict work scheduling to identify high-​demand jobs and peak hours. The high demand category refers to the distinction between paid time—​actual face-​to-​face engagement in in-​person services—​and non-​paid time, that is, various preparatory and administrative activities. This distinction shifts the risks of changing customer demand to the staff, increases work intensity during each paid working hour and blurs the boundaries between employee and self-​employed status. In the US context, similar scheduling practices were shown to be primary operating principles of care service delivery and were identified as major components in the explanation of class, racial and gender-​ based polarization of wages (Dwyer 2013). One major mechanism that aggravates violations of workers’ rights and reduces income is the use of domestic supply chains in service delivery. Ravenswood and Kaine (2015) have exposed the details of this mechanism in the context of elderly care as procured by Australia and New Zealand: The contracted party continues to outsource parts of the service to additional firms. The authors explain that contracted service deliverers are accredited regardless of their employment practices and that no specification of labour conditions—​other than a minimal reference to registered nurse coverage and nominal training—​are included in the contract. In the context of their study, this laissez-​faire position is translated into lack of regulation, allowing service deliverers to reduce labour standards to the minimum level. The most important aspect of the domestic supply chain for the discussion of job quality is that of fragmentation: employers cannot be compelled to adopt the achievements of industrial struggles. Because of the effect this fragmentation has on pay and working-​time arrangements, it reduces job quality dramatically. Studies from the US (Craft-​Morgan et al. 2013; Dill 2015), UK (Lewis and West 2014; Perrons 2007; Rubery et al. 2015), Canada (Armstrong 2013; Baines 2004; Caragata 2003; Vosko et al. 2009; Vosko 2010), Australia (Ravenswood and Kaines 2015), Germany (Theobald 2012), Norway (Andrews and Waerness 2011) and Israel (Benjamin 2016b) consistently show that cheap services are often based on unpaid time extraction and workload intensification. Consequently, nursing assistants and elderly care employees, together with childcare teachers’ assistants, who are employed by service deliverers, occupy the lowest income rank, with minimum or close to minimum wages. At the same time, certified employees in these fields, often paid by the hour, are also earning low incomes, even though they receive more than the minimum wage, because the part-​time jobs they receive are of brief duration. Low, non-​negotiable wages and little opportunity to participate in work organization and scheduling, contribute

304   orly benjamin to high turnover rates (Craft-​Morgan et al. 2013). Service and care job quality is thus reduced by two forms of job insecurity: First, given few occupational alternatives, practitioners become dependent on bad jobs with low wages and subcontracting relations; second, they face a constant, substantial threat of either job loss or change in employment conditions because of reform, new employers or downsizing.

Skills and Training Opportunities With the current international incremental convergence in organization of elderly and childcare services (Mahon et al. 2012) job quality depends more than ever on local skill recognition systems. Enhanced job quality is essential for service quality (Howes et al. 2012). Staffing levels, as well as staff training and education, were presented by Mor (2014) as an inevitable aspect of any social service regulation system. Staffing levels define the ability of service and care employees to use and professionalize their skills. Armstrong’s (2013) emphasis on working with more experienced staff members is crucial to job quality: Without such opportunity to learn, interpersonal encounters may become frustrating and even violent (Baines 2006), engendering loss of jobs. At-​work training and occupational education replace intuitive (feminine) care and enhance ethical principles, including attentiveness, responsibility, competence and responsiveness (Tronto 1993). Reflexivity and public participation are also important aspects of skill development associated with decreased levels of controlling/​abusive practices towards those in need of care and service (Abbott and Meerabeau 1998). Leone and colleagues (2014) surveyed national regulation systems and reported diverse levels of professionalization as a criterion for regulation. In their international comparison of the UK, Australia and New Zealand, the cases of more rapid liberalization and government procurement show forms of regulation that marginalize professionalization and focus on results and outputs. By analysing contract-​based public-​service delivery in Canada, Ilcan et al. (2003) showed that Canada maintains high levels of professionalization when it comes to governmental functioning, yet follows the UK and New Zealand in reducing focus on professionalization regarding outsourced services. To secure quality jobs and services, feminists have engaged in professionalization of service and care occupations since women began to enter higher education (Dahle 2012). Once policymakers aim at reduced service costs, sometimes phrased as ‘best value for money’, the feminist professionalization project is attacked, resulting in the reduced quality of services associated with low-​quality jobs and high overload (Pavolini and Klenk 2015). Questioning of professionalization is implicit for certified employees and explicit for uncertified ones; it occurs at institutional sites in which service quality standards are negotiated and decisions concerning public funding, are made. In the current service delivery system, typical in mixed welfare countries (Bode 2006), the ‘cheap services’ principle is promoted through deskilling, namely, endorsement of the assumption that little skill, knowledge or experience is required for performing service

job quality for service and care occupations    305 and care work. These principles diminished job quality critically by ‘a partial erosion of the special status of public sector employees, a decentralization of wage-​setting systems and differentiation of pay and working conditions’ (Bach and Borgodna 2011: 2290). As many have shown (Duffy 2011; Benjamin 2016b; Tronto 2013), the cheap services assumption revives the historical belief that femininity alone suffices for the performance of service and care. Professionalization and any related training, that potentially increase remuneration for service and care employees, are rendered redundant by this assumption. Cheap services then reduce job quality not just by lowering income but by generating work processes that defy employees’ ability to control them. Strict Taylorist time regimes, shown by Theobald (2012, 2015) and Pavolini and Klenk (2015), for example, deprive employees of the ability to control their work pace and consequently limit their ability to use their occupational skills. In this way, explained Baines (2004), a skilled (certified) workforce can be easily replaced by volunteers and unskilled (uncertified) workers. The assumption that childcare services must be affordable to low-​income mothers encouraged policymakers in the US, UK and Israel to define childcare employees as an unskilled labour force. Childcare occupational skills are still associated with (given) mothering skills rather than ‘learned and valuable’ (Findlay et al. 2009: 423). The assumption that caring skills do not require time to acquire enabled the development of a system of home-​based childcare services in these countries, whose providers are subject to a set of regulations (Smith 2006). Following this definition, employees lost their right to all but minimal, unorganized training. More significantly, many lost the practical option of labour law protection (Smith 2006). According to Duffy (2011), as long as employees are perceived as providing ‘care’, the current service delivery system refuses to train them. Only when they are considered ‘educators’ or ‘therapists’ are they seen as entitled to training. In 2013, the UK media revealed certain alarming statistics. For example, hundreds of patients at Stafford Hospital died as a result of neglect and abuse between 2005 and 2008. Reports indicated that healthcare assistants had almost no training (in some places, their entire training consisted of viewing a two-​hour video clip). This may not have been an exception. In the context of cheap services delivered on the basis of deskilling, occupational training is doubly problematic. First, it creates the expectation of a payment scale and promotion based on occupational training, all of which runs contrary to income-​flattening mechanisms. Second, occupational training is expensive and takes time away from actual performance. These features are so expensive for contracted service deliverers that they occasionally breach health and safety regulations to save on training expenses.

Working Conditions—​Work Intensity The deskilling of front-​line employees through Taylorization, in which care work is fragmented into measurable tasks (e.g. Theobald 2012; Armstrong 2013), facilitates

306   orly benjamin performance quantification and evaluation based on work unit tallying. In practical terms, Taylorization ignores the meaning of care as work based on a relationship (Duffy et al., 2013). Deskilling of this type also occurs when service and care employees in the public sector suffer from workload intensification and a significant increase in administrative work required, as explained by Baines et al (2014) regarding healthcare employees and Pavolini and Klenk (2015) regarding welfare professionals in Europe. Administrative burdens of this type, confine professional skills to small portions of the work-​day. Intensification of work was further explained by Reay (2000) for contract researchers in the academic world. She showed that within the context of fixed-​term, contract-​based employment and increasingly more periods of time outside the workplace, even research, a highly valued activity, is reconstituted as low-​level work. This devaluation generates intensification of work by practically demanding that hourly-​paid employees would manage workloads which can only be managed by additional (often un-​paid) working hours. The same intensification was found in a range of service and care occupations and was later named a ‘two-​tier workforce’ typical of local governments amongst other public services (Conley, 2006). The expectation of what can be achieved within one time-unit is heightened by this undervaluing perception: conducting field work, writing detailed reports, composing research proposals for funding their next contract and writing for publication all bear the implication that researchers have never done enough. Reay cites Dorothy Smith’s emphasis on the ‘invisible labour’ performed by contract employees that is not counted as billable. The relevance of invisibility as a measure of intensification is clarified further by Conley and Jenkins (2011), who showed how the modernization of teaching in the UK had introduced principles of rationality and productivity that left teachers’ emotional work unnoticed. Furthermore, their study found that teaching work spilled over into home life, extracting unpaid working hours during evenings and weekends. This form of work intensification led some of the mothers responsible for family care to leave teaching altogether. When the school is small, the workload increases because of multiple roles, as one of Conley and Jenkins’ participants remarked: ‘The workload was just ridiculous, to be honest. The more you did, the more was expected from you’ (p. 500).

Combining Work and Family Life In their early account of the intersectionality of gender, race and class of non-​standard employment, Zeytinoglu and Muteshi (2000) drew attention to the ways in which the feminist conception of ‘flexible work time’ has been appropriated and exploited against women. They distinguish the differences between the feminist goal of enabling flexibility, the principle of organizing one’s work day in a manner that enhances control over how caring obligations can be shared with others—​and constraining flexibility, that is, the mechanisms that significantly increase the motherhood penalty (Budig and Misra 2010). Constraining flexibility refers to very small, very low-​quality, part-​time jobs that Weinkopf (2009) in the German context, calls ‘mini-​jobs’. These apparently reproduce

job quality for service and care occupations    307 dependency rather than alleviate women’s poverty. Combining work and family in the context of constraining flexibility suggests that mothers are penalized for undertaking mini-​jobs. Recently, Jacobson and Padavic (2015) showed that the detrimental consequences of constraining flexibility operate primarily via unpredictable schedules, inadequate hours, time theft, punishment and control expressed in hour reduction, all of which are manifestations of the extreme power differential between employees and employers (Vosko et al. 2009). Job quality is reduced by two other mechanisms based on hours: channelling teachers into freelance jobs or work from home and refusing to allow part-​time jobs after maternity leave. These controlling actions are interpreted by researchers as signs of the masculinization of schools (Conley et al. 2011). Levels of job insecurity in these positions are heightened by continuous competition against uncertified employees allegedly offering the same skilled work at a lower cost. Thus, service delivery based on mini-​jobs, freelancing and working from home must be perceived as based on the recommodification of women’s work, in which payment is performancedependent and all other expenses are shouldered by the employee. In this context, Peña-​ Casas and Ghailani (2011) reported that the risk of employment poverty for service and care employees can be avoided only when living with other providers.

Gender Equality Promotion Public-​sector reforms, particularly those related to government procurement as a mode of service delivery, emerged concurrently with another set of reforms related to intersectionality (Krizsán et al. 2012) and the need to adopt anti-​discriminatory policies. In the UK, this additional set of reforms was summarized by the 2010 Equality Act (Conley and Page 2010), while in the Israeli context, the concurrence was even clearer, with about 10 equality laws legislated between 1985 and 2000—​the period during which government procurement of services was routinized. The set of equality policies, however, could not protect service and care employees from nullification of their skills, experience and knowledge in the new forms of employment introduced by government procurement. This crucial limitation is twofold. First, the domestic supply chain mechanism exempts contractors and service deliverers from applying equality policies. Hence, no protection was offered to the employees who needed it most. Second, according to these policies, discrimination is not defined in terms of job quality but rather on a comparative basis (I discuss this point in Benjamin 2013). As such, these policies could not be applied simultaneously with the income-​reduction mechanisms. If men are employed in service and care occupations, they too will usually be subject to similar wage-​flattening techniques. Furthermore, it may be argued that these mandatory equality-​based improvements in women’s lives can be viewed as applicable primarily to women whose life situations enable them to take cases to court, with or without the support of feminist lawyers. Most service and care employees are not in that position, whether because of uncertain citizenship issues or the lack of other types of resources (Parreñas 2015; Gorham 2017). What is more fascinating, however, is

308   orly benjamin the possibility, validating Fraser’s (2013) argument, that this state feminism-​informed surge of obligatory equality legislation has attracted much attention to the front door. However, when women are still struggling at the back door. Women in bad jobs in need of social protectionism, have been left on their own, facing reluctance to recognize their skills. Following Dantico and Jurik (1986), in an early observation concerning the effects of privatization on American women’s employment, Patricia Hill-​Collins (2005) referred to this process as the Closing Doors Policy. She believed it worked primarily to the detriment of African Americans who, in previous generations, could mobilize into middle-​class positions thanks to public-​sector quality jobs. The transformation of public sectors and the disappearance of quality jobs, that could be held even by uncertified employees, are issues that have been subdued by the emergence of women as citizens entitled to sue employers who discriminate against them as individuals. Adoption of these policies was often voluntary, but their greatest drawback is their failure to become more powerful than other implicit processes occurring on the corporate level (Acker 2006). As confirmed by Healy and colleagues (2011), such antidiscriminatory policies cannot overcome minority women’s exclusion. More problematically, once austerity measures were applied in the UK, managers in charge of implementing equality legislation either lost their jobs or the resources crucial to their implementation (Conley and Page 2014). The contract between public agencies and service deliverers, that is applied mostly as part of a set of market-​based solutions to public and social service delivery, has already attracted the attention of legal scholars. If the contract shapes employment conditions, it is suggested that the skill levels, workloads and related payments included in the contract incorporate all the legal requirements for protection of women, including mandatory equality (e.g. Conley and Wright 2015) and social objectives (Sarter 2015). These potential contract-​improvement methods, however, have not been endorsed by legal scholars, who consider the contract itself to be the problem. Epstein (2013) suggested that two biases are built into the contract, imposing low quality service on users: the first is, the low incentive of public agencies to spend more money on quality of services that target populations with little political power; the second is, the dependency of vulnerable populations on a handful of service deliverers, particularly when the ‘market is shallow’ (p. 5) and higher level goals cannot be enforced. The contract, its design and the budgeting considerations it embeds, all emerge as crucial, inhibiting players, institutionally undermining employees’ right to work, as well as their right to provide care in the appropriate services.

Collective Interest Representation Gender issues shape and are themselves shaped by intraprofessional structure, primarily those reflecting the hierarchy between the valued and undervalued in occupational behaviour (Broadbent et al. 2017). In times of managerialization (Pavolini and Klenk 2015; Evetts 2009), such structures are closely related to job quality and to organizational

job quality for service and care occupations    309 options for its negotiation. Howes and colleagues (2012) report that unless caring employees are members of particularly powerful unions, they tend to pay high caring penalties. For example, their income, compared with that of others with their levels of education, experience, seniority and job extent, is 5–​6 per cent lower per hour. Most are willing to accept less income because they depend on their jobs to manage their household obligations. This balancing act between home and family, however, means that many are compelled to compromise regarding their ability to provide for their families. Howes and colleagues (2012) argue that working with others requires energy, enthusiasm and motivation, the nature of which is largely determined by the degree of autonomy and collaboration afforded by the workplace environment. Extensive use of assistant jobs in UK social service delivery (Bach 2011) has led to several attempts at unionization, that can shed considerable light on the association between unionization and job quality in service and care occupations (Simms 2017). For example, outside the social services, the extent of unionization at call centres was found to have some effect on wage levels (Schönauer 2008) but little impact on job quality unless supported by the national context (Doellgast et al. 2009). The authors addressed three dimensions of job quality: dismissal rates, high involvement work practices and performance monitoring. Moreover, organization in caring professions is far too recent and fragmented to enable significant broad-​based achievements. In 2014, for example, Unison in the UK was successful in protecting long-​term elderly care employees from the zero-​hour contract practices of two employers. This major victory, that was so significant in the lives of women employed in this sector, has yet to have an impact on many other long-​term elderly care workers employed by other firms in the UK and around the globe. Lehndorff (2015) demonstrated that when unions suffer limited bargaining power, both minimum wage legislation and collective agreement extension remain dependent on (virtually non-​existent) government enforcement. Similarly, fragmentation has displayed only limited unionization achievements in Ireland. In 2004, the Health Service Executive was still the direct employer of a majority of long-​term care employees. In addition, 37 NGOs were also employing care workers under vastly inferior conditions. The Irish Service Industrial Professional and Technical Union (SIPTU) managed to organize both types of employees with the hope of providing them with decent and secure employment. Murphy and Turner (2014), who reported on this 2004 struggle, explained that despite successful negotiations between union and employer and the agreement achieved in 2004, the improved terms and conditions of employment were not honoured before 2013 ‘when binding LC judgment ultimately secured victory for the union’ (2014: 7). In 2016, the Union of Care Employees in New Zealand, led by Kristine Bartlett, managed to secure a wage increase for 55,000 employees, to be carried out over time. Certified employees are expected to benefit from this achievement far more than uncertified employees (Ravenswood 2017), but the achievement is impressive nonetheless. In Israel, new unionism benefited service and care employees in a range of outsourced social services (Preminger 2018). Thus, unions that have organized women are of immeasurable importance, as they have been throughout history. At the same time, in an age of fragmentation, their effect is limited, as is the effect of union

310   orly benjamin coalitions (Bailey et al. 2014). For example, the Australian feminist coalition that strove for enactment of The Fair Work Act (2009) and was successful in convincing the labour court that care employees’ work was undervalued, remained limited in its achievements by fragmentation within professions dominated by women. Cortis and Meager (2012) explain that while feminist activity was aimed at social and community services, a large group of other service and care employees, including long-​term caregivers for the elderly, was not included among employees entitled to the wage increase so attained.

Feminist Issues Concerning Job Quality Conceptualizations of job quality have been challenged by the low level of job quality and equity payment in service and care occupations since the 1990s (Armstrong 1997) and still raise several significant questions today (Rubery 2013). As already mentioned, one way of dealing with this challenge proposed the subjective focus or intrinsic job quality arguments. Nevertheless, subjective job quality is shaped by institutional and discursive ideas such as expectations regarding the state and evaluation of state efforts to meet them (Meagher et al. 2016). The perspective proposed in this study persists in criticizing the allegedly compensatory value of subjective job quality and follows an account indicating that the nature of subjective job quality has been shaped by cultural constraints (Hebson et al. 2015). Hence, it does not include subjective assessments of job quality. Instead, this account emphasized issues of remuneration, certification and decommodification. Shifting the focus of job quality to these three dimensions contextualizes women’s working lives and elucidates undervaluation of women’s work as manifested in work process organization and in women’s lives outside workplaces. Using studies of employed women, I showed how these three dimensions are systematically defeated by employment arrangements that legitimate hourly-​based payment, part-​time work, mini-​jobs, temporary employment and—​in a more general sense—​ restrict rather than enable flexibility (Zeytinoglu and Muteshi 2000). One major force defeating these three dimensions of job quality is the cheap services orientation, that entrenches a reduced level of funded work time (Armstrong 2013; Benjamin 2015; Rubery et al. 2015). This force was found to lower remuneration and devalue occupational training and certification to a minimum that flattens payment ladders. Job quality under the cheap services orientation cannot be protected, primarily because women are invited to reproduce traditional gender identities by contributing their work. Elsewhere, I referred to this process as a return to doing gender (Benjamin 2016b). In this context, gender is ‘performed’ through enhanced input of feminine intuition, rendering remuneration redundant. A return to doing gender further implies that quality jobs for women can be only ensured by their inclusion in traditionally male occupations such as construction work (Conley and Wright, 2015). The idea that services should be cheap is also salient in promoting deprofessionalization processes that tend to converge, under these practices, with the loss of workers’ dignity (Benjamin 2016a). It is important, however, to pay attention to different employment arrangements

job quality for service and care occupations    311 that at times emphasize differences between employees whose job quality is enhanced by upskilling and professionalization and those for whom it is not. Standardization of upskilling processes, along with unionization, would improve conditions for currently excluded employees while benefiting society at large by creating reflexive quality services. To conceptualize a feminist approach to job quality that entails understanding the economic-​political-​social mechanisms facilitating these practices, I raise three questions. The first is derived from the assumption that the history of social services in education, healthcare and welfare is the history of the feminist struggle for professionalization, skill recognition, certification and occupational training, thereby distancing such women from volunteer, unpaid, intuitive work and care. As shown by Duffy (2011), the gains of these struggles left several skills unrecognized, especially those skills used by women in service occupations perceived as not requiring a feminine dimension, such as cleaning, catering and long-​term care. Consequently, unionization has been crucial for women in all service and care fields as the one essential measure to confine polarization to protected vs unprotected women’s occupations (Cobble 1993). With the attack on unions (Tope and Jacobs 2009), the continuous struggle against the exclusion of women employed in these occupations from the protection measures that women gained in standard employment (Zeytinoglu and Muteshi 2000) became more relevant (Vosko 2010). The question thus focuses on skill recognition, asking what social forces are currently involved in enhancing job quality for women at the bottom of the service and care occupation ladder and how scholarship on job quality joins this struggle. My second question is derived from the feminist ‘the personal is political’ premise, assuming the salience of incorporating an institutional dimension (Walby 2015) into a feminist conceptualization of job quality, that would examine the treatment of an occupation rather than individual jobs. Furthermore, treatment of occupations in the sense of social organization of remuneration, certification and decommodification should be examined on a policy level. The state’s responsibility for these issues is relevant to three categories: its own employees, employees who depend on its funding of outsourced services and other employees exposed to the state’s role in establishing standards of employment. The state, in its relations with the market, civil society and occupational associations (Thatcher and Schmidt 2013), designs policies that may or may not value occupations and employees of these three categories. In this context, I ask how policies of marketization in general, and public procurement in particular, become shaping forces for job quality in service and care occupations, and what measures can be taken to restrain their power to reduce job quality for women. Intersectionality theory constitutes the starting point of my final question, applying Nancy Folbre’s (1996) observation that policymakers, primarily budgeting administrators, persistently promote their idea that care work should be contributed for free or for pin money. Following Acker’s (2006) understanding of inequality regimes, I ask how ethnic–​ racial–​ national–​ religious hierarchies in different countries are reflected in the polarization of job quality. This question raises the possibility that better

312   orly benjamin quality jobs are reserved almost exclusively for women of dominant categories, while others are left to encounter the Closing Doors Policy (Hill-​Collins 2005). To avoid a service delivery system based on the self-​exploitative feminine self (Baines et al. 2014) and on service employees who continue to do more than the prescribed tasks despite the deep emotional wounds inflicted by the failure to receive adequate recognition for their contribution (Dahl 2009), an alternative service delivery model is required, based on high standards of professionalization and regulation, one unified industrial-​relations system (Bach 2012) and direct participation. Protecting job quality within a caring democracy (Tronto 2013) is a feminist project that values the knowledge, experience and skill accumulated by scholarly work and practice alike. Promoting an alternative requires a union/​feminist/​activist/​scholarly coalition operation. Recent coalitions of this type have managed to gain some improvements. Bailey et al. (2013) report that in Australia, a union coalition was successful in promoting a Good Work—​ Decent Wages campaign, comprising regulatory strategies combined with sector-​wide mobilization. These authors indicate the need for a coalition directed at challenging the care penalty that would have to bridge the interests of educated certified employees, uncertified employees, persons dependent on service and care practitioners and their care-​ providing families, state administrators committed to high quality standards in public and social services and scholars studying and investigating care and service delivery budgeting. Together, we should draw attention to the interrelation between job and service quality and possible sources of funding to benefit human society in a caring and democratic manner.

References Abbott, P., and Meerabeau, L. (1998), The Sociology of the Caring Professions (Abingdon (UK): Psychology Press). Acker, J. (2006), ‘Inequality Regimes: Gender, Class, and Race in Organizations’, Gender and Society, 20/​4, 441–​464. Andrews, T. M., and Wærness, K. (2011), ‘Deprofessionalization of a Female Occupation: Challenges for the Sociology of Professions’, Current Sociology, 59/​1, 42–​58. Armstrong, P. (1997), ‘Restructuring Public and Private: Women’s Paid and Unpaid Work’, in S. B. Boyd, ed., Challenging the Public/​Private Divide: Feminism, Law, and Public Policy (Toronto, University of Toronto Press), 37–​61. Armstrong, P. (2013), ‘Puzzling Skills: Feminist Political Economy Approaches’, Canadian Review of Sociology/​Revue Canadienne de Sociologie, 50/​3, 256–​283. Bach, S. (2011), ‘Assistant Roles in a Modernised Public Service: Towards a New Professionalism?’ in S. Graham and S. Corby, eds., Working for the State: Employment Relations in the Public Services (London: Palgrave Macmillan), 129–​146. Bach, S. (2012), ‘Shrinking the State or the Big Society? Public Service Employment Relations in an Era of Austerity’, Industrial Relations Journal, 43/​5, 399–​415. Bach, S., and Bordogna, L. (2011), ‘Varieties of New Public Management or Alternative Models? The Reform of Public Service Employment Relations in Industrialized Democracies’, International Journal of Human Resource Management, 22/​11, 2281–​2294.

job quality for service and care occupations    313 Bailey, J., Robertson, M., and Hulme, L. (2013), ‘Challenging the “Care Penalty”: The Queensland Pay Equity Campaign for Community Services Workers’, Journal of Industrial Relations, 56/​1, 43–​61. Baines, D. (2004), ‘Caring for Nothing: Work Organization and Unwaged Labour in Social Services’, Work, Employment and Society, 18/​2, 267–​295. Baines, D. (2006), Staying with People Who Slap Us Around: Gender, Juggling Responsibilities and Violence in Paid (and Unpaid) Care Work. Gender, Work & Organization, 13’2, 129–​151. Baines, D., Charlesworth, S., Turner, D., and O’Neill, L. (2014), ‘Lean Social Care and Worker Identity: The Role of Outcomes, Supervision and Mission’, Critical Social Policy, 34/​4, 433–​453. Benjamin, O. (2013), ‘From Cooperation to Criticism of Economic Globalization: An Intersectional Concept of Gender Justice’, Zeitschrift für Geschlecht, Kultur und Gesellschaft, 5(21), 45–​63. Benjamin, O. (2015). ‘Time is Money: Deskilling Caring Work through Time Allocation in Services Procurement’, Wagadu: A Journal of Transnational Women’s and Gender Studies, 14, 85–​110. Benjamin, O. (2016a), ‘Gendered Corrosion of Occupational Knowledge: Contracting-​out Israeli Social Services’, Equality, Diversity and Inclusion: An International Journal, 35/​3, 174–​185. Benjamin, O. (2016b), Gendering Israel’s Outsourcing: The Erasure of Employees’ Caring Skills (New York: Palgrave MacMillan). Bernhardt, A., Batt, R., Houseman, S. N., and Appelbaum, E. (2016), ‘Domestic Outsourcing in the United States: A Research Agenda to Assess Trends and Effects on Job Quality’, Upjohn Institute Working Paper 16-​253. Upjohn Institute for Employment Research. http://​dx.doi. org/​10.17848/​wp16-​253. Accessed July 2016. Bode, I. (2006), ‘Disorganized Welfare Mixes: Voluntary Agencies and New Governance Regimes in Western Europe’, Journal of European Social Policy, 16/​4, 346–​359. Broadbent, K., Strachan, G., and Healy, G. (eds.) (2017), Gender and the Professions: International and Contemporary Perspectives (New York: Routledge). Budig, J. M., and Misra, J. (2010), ‘How Care-​Work Employment Shapes Earnings in Cross-​ National Perspective’, International Labour Review, 149/​4, 441–​460. Caragata, L. (2003), ‘Neoconservative Realities: The Social and Economic Marginalization of Canadian Women’, International Sociology, 18/​3, 559–​580. Cobble, D. S. (1993), Women and Unions: Forging a Partnership (Ithaca: ILR Press). Conley, H. (2006), ‘Modernisation or Casualisation? Numerical Flexibility in Public Services’, Capital and Class, 30/​2, 31–​57. Conley, H., and Jenkins, S. (2011), ‘Still “a Good Job for a Woman”? Women Teachers’ Experiences of Modernization in England and Wales’, Gender, Work & Organization, 18/​5, 488–​507. Conley, H., and Page, M. (2010), ‘The Gender Equality Duty in Local Government: The Prospects for Integration’, Industrial Law Journal, 39/​3, 321–​325. Conley, H., and Page, M. (2014), Gender Equality in Public Services: Chasing the Dream (New York: Routledge). Conley, H., and Wright, T. (2015), ‘Making Reflexive Legislation Work: Stakeholder Engagement and Public Procurement in the Public Sector Equality Duty’, in M. Dustin ed., Beyond 2015: Shaping the Future of Equality, Human Rights and Social Justice (London: EDF), 54–​64.

314   orly benjamin Cortis, N., and Meagher, G. (2012), ‘Recognition at Last: Care Work and the Equal Remuneration Case’, Journal of Industrial Relations, 54/​3, 377–​385. Craft-​Morgan, J., Dill, J., and Kalleberg, A. L. (2013), ‘The Quality of Healthcare Jobs: Can Intrinsic Rewards Compensate for Low Extrinsic Rewards?’ Work, Employment and Society, 27/​5, 802–​822. Dahl, H. M. (2009), ‘New Public Management, Care and Struggles about Recognition’, Critical Social Policy, 29/​4, 634–​654. Dahle, R. (2012), ‘Social Work: A History of Gender and Class in the Profession’, Ephemera Articles: Theory & Politics in Organization, 12/​3, 309–​326. Dantico, M., and Jurik, N. (1986), ‘Where Have All the Good Jobs Gone: The Effect of Government Service Privatization on Women Workers’, Contemporary Crises, 10/​4, 421–​439. Davies, C. (1996), ‘The sociology of professions and the profession of gender’, Sociology, 30(4), 661–​678. Davoine, L., Erhel, C., and Guergoat-​ Lariviere, M. (2008), ‘Monitoring Quality in Work: European Employment Strategy Indicators and Beyond’, International Labour Review, 147/​2–​3, 163–​198. Dawson, S. L. (2016), ‘The Direct Care Workforce—​ Raising the Floor of Job Quality’, Generations, 40/​1, 38–​46. DeVault, M. L. (ed.) (2008). People at Work: Life, Power and Social Inclusion in the New Economy (New York: New York University Press). Dill, J. S. (2015), ‘Are Frontline Healthcare Jobs “Good” Jobs? Examining Job Quality across Occupations and Healthcare Settings’, in M. Duffy, A. Armenia, and C. L. Stacey, eds., Caring on the Clock: The Complexities and Contradictions of Paid Care Work (New Brunswick, NJ: Rutgers University Press), 54–​66. Doellgast, V., Holtgrewe, U., and Deery, S. (2009), ‘The Effects of National Institutions and Collective Bargaining Arrangements on Job Quality in Front-​Line Service Workplaces’, Industrial and Labor Relations Review, 62/​4, 489–​509. Duffy, M. (2011), Making Care Count: A Century of Gender, Race, and Paid Care Work (New Brunswick, NJ: Rutgers University Press). Duffy, M., Albelda, R., and Hammonds, C. (2013), ‘Counting Care Work: The Empirical and Policy Applications of Care Theory’, Social Problems, 60/​2, 145–​167. Dwyer, R. E. (2013), ‘The Care Economy? Gender, Economic Restructuring, and Job Polarization in the US Labor Market’, American Sociological Review, 78/​3, 390–​416. Eisenstein, H. (2015), Feminism Seduced: How Global Elites Use Women’s Labor and Ideas to Exploit the World (New York: Routledge). Epstein, W. N. (2013), ‘Contract Theory and the Failures of Public-​Private Contracting’, Cardozo Law Review, 34, 2211. Employment Commission (2001), Employment and Social Policies: A Framework for Investing in Quality. Communication from the Commission to the Council, Brussels, COM (2001) 313 final, 20.6.2001. Evetts, J. (2009), ‘New Professionalism and New Public Management: Changes, Continuities and Consequences’, Comparative Sociology, 8/​2, 247–​266. Findlay, P., Findlay, J., and Stewart, R. (2009), ‘The Consequences of Caring: Skills, Regulation and Reward among Early Years’ Workers’, Work, Employment and Society, 23/​3, 422–​441. Findlay, P., Kalleberg, A. L., and Warhurst, C. (2013), ‘The Challenge of Job Quality’, Human Relations, 66/​4, 441–​445. Folbre, N. (1996), The Economics of the Family (Cheltenham, UK: Edward Elgar Publishing).

job quality for service and care occupations    315 Fraser, N. (2013), Fortunes of Feminism: From State-​Managed Capitalism to Neoliberal Crisis (London and New York: Verso Books). Gallie, D. (2013), ‘Direct Participation and the Quality of Work’, Human Relations, 66/​4, 453–​473. Gorham, U. (2017), Access to Information, Technology, and Justice: A Critical Intersection (Lanham, MD: Rowman & Littlefield). Grimshaw, D. Marchington, M. Rubery, J., and Willmott, H. (2005), ‘Introduction: Fragmenting Work across Organizational Boundaries’, in M. Marchington, D. Grimshaw, J. Rubery, and H. Willmott, eds., Fragmenting Work: Blurring Organizational Boundaries and Disordering Hierarchies (Oxford: Oxford University Press), 1–​31. Healy, G., Bradley, H., and Forson, C. (2011), ‘Intersectional Sensibilities in Analyzing Inequality Regimes in Public Sector Organizations’, Gender, Work and Organization, 18/​5, 467–​487. Hebson, G., Rubery, J., and Grimshaw, D. (2015), ‘Rethinking Job Satisfaction in Care Work: Looking beyond the Care Debates’, Work, Employment and Society, 29/​2, 314–​330. Hill-​Collins, P. (2005), Black Sexual Politics: African-​Americans, Gender and the New Racism (New York: Routledge). Howes, C., Leana, C., and Smith, K. E. (2012), ‘Paid Care Work’, in N. Folbre, ed., For Love or Money: Care Provision in the United States (New York: Russell Sage), 183–​204. Ilcan, S. M., O’Connor, D. M., and Oliver, M. L. (2003), ‘Contract governance and the Canadian public sector’, Revue Relations Industrielles/​Industrial Relations, 58/​4, 620–​643. Jacobs, A. W., and Padavic, I. (2015), ‘Hours, Scheduling and Flexibility for Women in the US Low‐Wage Labor Force’, Gender, Work & Organization, 22/​1, 67–​86. Jaehrling, K., Weinkopf, C., and Wagner, I. (with G. Bosch and T. Kalina) (2016), ‘Reducing Precarious Work in Europe through Social Dialogue: The Case of Germany’, Report for the European Commission. Institute of Work, Skills and Training, University of Duisburg-​Essen. Kalleberg, A. L. (2011), Good Jobs. Bad Jobs: The Rise of Polarized and Precarious Employment Systems in the United States, 1970s to 2000s (New York: Russell Sage Foundation). Kauhanen, M., and Nätti, J. (2015), ‘Involuntary Temporary and Part-​Time Work, Job Quality and Well-​Being at Work’, Social Indicators Research, 120/​3, 783–​799. Klenk, T., and Pavolini, E. (eds.) (2015), Introduction: Restructuring Welfare Governance: Marketisation, Managerialism and Welfare State Professionalism (Cheltenham, UK: Edward Elgar), 1–​8. Krizsán, A., Skjeie, H., and Squires, J. (eds.) (2012), Institutionalizing Intersectionality: The Changing Nature of European Equality Regimes (New York: Palgrave Macmillan). Lehndorff, S. (2015), ‘Institutional Anchors for Job Quality–​The Case of Industrial Relations’, in U. Holtgrewe, M. Ramioul, and V. Kirov, eds., Hard Work in New Jobs: The Quality of Work and Life in European Growth Sectors (London: Palgrave Macmillan), 110–​113. Lewis, J., and West, A. (2014), ‘Re-​ Shaping Social Care Services for Older People in England: Policy Development and the Problem of Achieving ‘Good Care’, Journal of Social Policy, 43/​1, 1–​18. Mahon, R., Anttonen, A., Bergqvist, C., Brennan, D., and Hobson, B. (2012), ‘Convergent Care Regimes? Childcare Arrangements in Australia, Canada, Finland and Sweden’, Journal of European Social Policy, 22/​4, 419–​431.

316   orly benjamin Meagher, G., Szebehely, M., and Mears, J. (2016), ‘How Institutions Matter for Job Characteristics, Quality and Experiences: A Comparison of Home Care Work for Older People in Australia and Sweden’, Work, Employment and Society, 30/​5, 731–​749. Mor, V. (2014), ‘Regulating Quality of Long Term Care –​What Have We Learned’, in V. Mor, T., Leone, and A. Marreso, eds., Regulating Long-​Term Care: An International Comparison (Cambridge: Cambridge University Press), 447–​476. Mühlau, P. (2011), ‘Gender Inequality and Job Quality in Europe’, Management Review, 22/​2, 114–​131. Murphy, C., and Turner, T. (2014), ‘Organising Non‐Standard Workers: Union Recruitment in the Irish Care Sector’, Industrial Relations Journal, 45/​5, 373–​388. Parreñas, R. (2015), Servants of Globalization: Migration and Domestic Work (Stanford, CA: Stanford University Press). Peña-​Casas, R., and Ghailani, D. (2011), ‘Towards Individualizing Gender In-​Work Poverty Risks’, in N. Fraser, R. Gutierrez, and R. Peña-​Casas, eds., Working Poverty in Europe. Work and welfare in Europe Series (New York: Palgrave Macmillan), 202–​231. Perrons, D. (2007), ‘Living and Working Patterns in the New Knowledge Economy: New Opportunities and Old Social Divisions in the Case of New Media and Care-​ Work’, in S. Walby, H. Gottfried, K. Gottschall, and M. Osawa, eds., Gendering the Knowledge Economy: Comparative Perspectives (New York: Palgrave Macmillan), 188–​206. Preminger, J. (2018), Labor in Israel: Beyond Nationalism and Neoliberalism (Ithaca, NY: Cornell University Press). Ravenswood, K. (2017). ‘Improving Care Worker Conditions for Aged Care Workers in New Zealand’. Paper presented at the First Care Work Summit, University of Massachusetts, Lowell, MA. Ravenswood, K., and Kaine, S. (2015), ‘The Role of Government in Influencing Labour Conditions through the Procurement of Services: Some Political Challenges’, Journal of Industrial Relations, 57/​4, 544–​562. Reay, D. (2000), ‘Dim Dross’: Marginalized Women both inside and outside the Academy’, Women’s Studies International Forum, 23/​1, 13–​21. Rubery, J. (2008), ‘Women and Work in the UK: The Need for a Modernisation of Labour Market Institutions’, in J. Scott, S. Dex, and H. Joshi, eds., Women and Employment: Changing Lives and New Challenges (Cheltenham, UK: Edward Elgar), 289–​312. Rubery, J. (2013), ‘Public Sector Adjustment and the Threat to Gender Equality’, in D. Vaughan-​Whitehead, ed., Public Sector Shock: The Impact of Policy Retrenchment in Europe (New York: International Labour Organization), 43–​83. Rubery, J., Grimshaw, D., Hebson, G., and Ugarte, S. M. (2015), ‘It’s All About Time’: Time as Contested Terrain in the Management and Experience of Domiciliary Care Work in England’, Human Resource Management, 54/​5, 753–​772. Santero-​Sanchez, R., Segovia-​Pérez, M., Castro-​Nuñez, B., Figueroa-​Domecq, C., and Talón-​ Ballestero, P. (2015), ‘Gender Differences in the Hospitality Industry: A Job Quality Index’, Tourism Management, 51, 234–​246. Sarter, E. K. (2015), ‘The Legal Framework of Contracting: Gender Equality, the Provision of Services, and European Public Procurement Law’, WAGADU: Journal of Transnational Women’s Studies, 14 (Winter), 55–​83. Schönauer, A. (2008), ‘Reorganising the Front Line: The Case of Public Call Centre Services’, Work Organisation, Labour and Globalisation, 2/​2, 131–​147. Simms, M. (2017). Unions and job quality in the UK: Extending interest representation within regulation institutions. Work and Occupations, 44/1, 47–67.

job quality for service and care occupations    317 Smith, P. R. (2006), Welfare, Child Care, and the People who Care: Union Representation of Family Child Care Providers’, Kansas Law Review, 55/​2, 321–​364. Stecy-​Hildebrandt, N., Fuller, S., and Burns, A. (2019), ‘‘Bad’ Jobs in a ‘Good’ Sector: Examining the Employment Outcomes of Temporary Work in the Canadian Public Sector’, Work, Employment and Society, 33/​4, 560–​579. Stier, H. (2015), ‘The Skill-​Divide in Job Quality: A Cross-​National Analysis of 28 Countries’, Social Science Research, 49, 70–​80. Stier, H., and Yaish, M. (2014), ‘Occupational Segregation and Gender Inequality in Job Quality: A Multi-​Level Approach’, Work, Employment & Society, 28/​2, 225–​246. Supiot, A. (2001), Beyond Employment (Oxford: Oxford University Press). Taniguchi, H., and Rosenfeld, R. A. (2002), ‘Women’s Employment Exit and Reentry: Differences among Whites, Blacks, and Hispanics’, Social Science Research, 31, 432–​471. Thatcher, M., and Schmidt, V. A. (2013), ‘Conclusion: Explaining the Resilience of Neo-​ Liberalism and Possible Pathways Out’, in V.A. Schmidt and M. Thatcher, eds., Resilient Liberalism in Europe’s Political Economy (Cambridge: Cambridge University Press), 403–​431. Theobald, H. (2012), ‘Combining Welfare Mix and New Public Management: The Case of Long-​Term Care Insurance in Germany’, International Journal of Social Welfare, 21, 61–​74. Theobald, H. (2015), ‘Marketization and Managerialization of Long Term Care Policies in Comparative Perspective’, in T. Klenk and E. Pavolini, eds., Restructuring Welfare Governance: Marketisation, Managerialism and Welfare State Professionalism (Cheltenham, UK: Edward Elgar), 27–​45. Tronto, J. C. (1993), Moral Boundaries: A Political Argument for an Ethic of Care (New York: Psychology Press). Tronto, J. C. (2013), Caring Democracy: Markets, Equality, and Justice (New York: NYU Press). Tope, D., and Jacobs, D. (2009), ‘The Politics of Union Decline: The Contingent Determinants of Union Recognition Elections and Victories’, American Sociological Review 74/​5, 842–​864. Vidal, M. (2013), ‘Low Autonomy Work and Bad Jobs in Post-​Fordist Capitalism’, Human Relations, 66/​4, 587–​612. Vosko, L. F. (2010), Managing the Margins: Gender, Citizenship, and the International Regulation of Precarious Employment (Oxford: Oxford University Press). Vosko, L. F., MacDonald, M., and Campbell, I. (2009), ‘Introduction: Gender and the Concept of Precarious Employment’, in L. F. Vosko, M. MacDonald, and I. Campbell, eds., Gender and the Contours of Precarious Employment (New York: Routledge), 1–​25. Walby, S. (2007), ‘Introduction: Theorizing the Gendering of the Knowledge Economy: Comparative Approaches’, in S. Walby, H. Gottfried, K. Gottschall, and M. Osawa, eds., Gendering the Knowledge Economy: Comparative Perspectives (London: Palgrave Macmillan), 3–​50. Walby, S. (2015), Crisis (Hoboken, NJ: John Wiley & Sons). Weinkopf, C. (2009), ‘Precarious Employment and the Rise of Mini-​Jobs’, in L. F. Vosko, M. MacDonald, and I. Campbell, eds., Gender and the Contours of Precarious Employment (New York: Routledge), 177–​193. Warren, T., and Lyonette, C. (2018), ‘Good, Bad and Very Bad Part-​Time Jobs for Women? Re-​ Examining the Importance of Occupational Class for Job Quality Since the ‘Great Recession’ in Britain’, Work, Employment and Society, 32/​4, 747–​767. Zeytinoglu, I., and Muteshi, J. (2000), ‘Gender, Race and Class Dimensions of Nonstandard Work’, Relations Industrielles, 55/​1, 133–​167.

chapter 14

inequality in job quality Class, gender and contract type duncan gallie

Introduction The issue of whether the quality of work has been becoming more or less equal over time has been at the centre of theory and research in the sociology of work since its inception. In the first three decades of the post-​war period the focus was almost entirely upon the issue of class inequality—​with sharply contrasting theories of class convergence and class polarization. From the 1980s, with the growth of theories of labour market segmentation, disadvantages experienced by women in work and employment became increasingly central. From the 1990s, with the growing internationalization of economic activity and the dominance of neo-​liberal views on the benefits of labour market flexibility, the implications of temporary employment contracts became a core focus of debates about inequalities in work and by the end of the 2000s was attracting significant policy concern (e.g. OECD 2008).1 This chapter examines these class, gender and contract-​based inequalities in work. Focusing on the US and European Union countries, it shows that inequalities of job quality based on class, gender and contract type still persist but the different types of inequality vary in terms of the dimensions of job quality involved and the extent to which they are cumulative. The chapter begins by presenting three contrasting theoretical arguments about the general trends with respect to inequalities and job quality and then turns to examine the empirical evidence with respect to inequalities relating to class, gender and contract security.

1  The most recent area of public concern has been with the potential disadvantages in job quality experienced by ethnic minorities and immigrants. However, the empirical data with respect to job quality is still sparse and a well-​grounded assessment will have to await the future.

inequality in job quality    319

Theoretical Scenarios of Inequalities in Job Quality In broad terms, there have been three major traditions: an optimistic tradition that has predicted increased convergence in work and employment conditions between different categories of employee, a pessimistic tradition that has pointed to factors leading to entrenched or even increased disadvantage and a neo-​institutionalist perspective that has argued for relatively stable and distinctive patterns of inequality between societies or types of society.

Scenarios of Improvement in Job Quality The most influential theoretical account of changes in the quality of work in the immediate post-​war period was the theory of industrialism, developed by Clark Kerr and his colleagues (Kerr et al. 1960). Developments in science and technology, it was argued, were leading to a workforce with a wide range of skills and competency broadly distributed throughout the workforce. Rising skill requirements would require inter alia continual training, the replacement of more authoritarian or directive forms of management by an emphasis on personal responsibility for performance and governance structures that would allow workers to share in the establishment and administration of the rules, providing a strong impetus to greater equality in work quality. In subsequent decades, there were a number of influential studies elaborating and updating the underlying ideas of the theory of industrialism, focusing in particular on the implications of technological change for the transformation of traditional forms of manual work through reskilling the workforce (Blauner 1964; Piore and Sabel 1984; Kern and Schuman 1987, 1992). From the mid-​1980s, technologically based arguments predicting a trend to greater equality in job quality were reinforced by new managerial theories that considered it a condition for productivity and competitiveness in an era of intensified international trade and rising demands for product and service quality. Walton (1985) argued that a revolution was under way in the management of work, involving a shift from seeking to increase productivity through detailed control of employees’ task performance to winning their commitment by improving the quality of their jobs. In the new era, employees would work in self-​supervising work teams, would have good opportunities for training, would be given a voice in organizational matters and have higher levels of job security. In a similar vein, Lawler (1986) argued for the benefits to employers as well as employees of systems of ‘High-​Involvement Management’, a general line of thinking which, under diverse names, became increasingly influential (Appelbaum et al. 2000; Lawler et al. 1995; Wood and Wall 2007; also Boxall and Purcell in the chapter entitled ‘Human Resource Management and Job Quality’ in this volume). In short,

320   duncan gallie developments in managerial theory came to reinforce the conclusions of theorists of technology who predicted a reduction of inequalities as a result of the upgrading of the job quality of lower-​level employees.

Worsening Job Quality with Segmentation and Polarization In sharp contrast to these arguments there has been a continuing vein in the literature that offered a much darker vision of long-​term developments in inequality at work—​ pointing to deeply rooted tendencies to segmentation between different types of employee and to growing inequality in job quality. Studies in the labour process tradition that flourished in the wake of Harry Braverman’s (1974) Labor and Monopoly Capital argued that, even with automation, there would be continuing pressures to subdivide work tasks into ever simpler operations and to replace workers’ decision-​making autonomy by close supervisory control. While earlier neo-​Marxian arguments had viewed polarization primarily as a growing divide between manual and non-​manual work, this new scenario also predicted the undercutting of skills and job control among white-​collar employees and hence a downward convergence of lower non-​manual and manual work. A quarter of a century later, a somewhat similar vision of increasing workforce polarization was also to become influential among some economists, albeit with a stronger emphasis on technological determinism. Advocates of routine-​biased technical change emphasized the destructive effects of technical change on the jobs of employees in the intermediary classes (Autor et al. 2003; Goos and Manning 2007). The assumption of a general trend towards the equalization of job quality was also contested by labour market segmentation theorists who depicted the labour market as structured by power relations that created relatively privileged and deprived sectors of employment, with major barriers to mobility, and channelled more vulnerable categories of employee into jobs offering poorer skill, pay and security. Gender differences were seen as a fundamental dimension of this unequal structure of labour markets, with women’s position undermined both by lower trade union protection and a division of labour placing childcare responsibilities predominantly upon the female partner. From the mid-​1990s, a new line of differentiation became increasingly salient—​ the division between those on standard and non-​standard contracts. Intensified international completion and market volatility were predicted to lead to new employers’ workforce strategies that sought to buffer the effects of unpredictable product market change by the creation of a marginal workforce, with short fixed-​term contracts, that could be easily shed in the event of economic downturn (Capelli et al. 1997; Kalleberg 2011). Whereas dual labour market theory had depicted the secondary workforce as primarily low skilled, the scenario of a growing ‘flexible’ workforce suggested that contractual inequalities might well cut across traditional skill divisions.

inequality in job quality    321

Societal Differences in Job Quality While drawing very different conclusions about long-​ term trends, both theories predicting greater equality and those pointing to the aggravation of inequalities assumed that trends would be very similar across countries. A third view has argued that institutional and cultural differences between countries lead to contrasting patterns of job inequality and that these differences are relatively stable across time. An influential perspective has been ‘varieties of capitalism’ perhaps better termed ‘production regime’ theory (Soskice 1999; Hall and Soskice 2001). Its starting point was variation in the degree of employer coordination in different countries and its implications for skill formation. Coordinated market economies included the Nordic and Continental European countries, while Liberal Market Economies, with low levels of employer coordination, were represented most clearly by the US and the UK. Societies where employer coordination was more developed favoured the creation of strong vocational training systems that emphasized occupational and firm-​specific skills. Specific skill systems in turn encouraged forms of work organization that would reduce inequalities by providing better quality jobs for lower-​level employees in terms of control over the work task, workplace voice, continuing training and job security. In contrast, societies where employer decision-​making was decentralized and coordination was largely provided by the market favoured general skills systems that were associated with poorer work conditions (Soskice 1999). An alternative ‘employment regime’ approach draws on the theoretical tradition of power resource theory to argue that job quality depends on the balance of power between employers and organized labour (Korpi 2006; Gallie 2007a). Where labour has sufficient influence, whether through trade unions or political parties, to participate in national-​level decision-​making, it could raise the salience of quality of work issues on the political agenda, encourage provisions for effective workplace voice and ensure that policies were introduced to improve work conditions. This perspective has tended to underline the differences between four types of regime—​the Nordic (Denmark, Finland, Norway and Sweden); the Continental (Austria, Belgium, Germany, the Netherlands and Switzerland), the Mediterranean (Greece, Italy, Portugal and Spain) and the Transitional (East European) regimes.

Empirical Evidence on Inequalities in Job Quality The focus is on evidence from the advanced industrial societies, in particular the European Union and the US, since these have the most detailed long-​term data. Assessing the scenarios of change developed in the theoretical literature requires representative data covering an extended period. Although, since the 1990s, there has been

322   duncan gallie a growth of large-​scale representative surveys on job quality, the time spans covered are limited and are certainly far shorter than is required for rigorous testing of longer-​term scenarios of change. In examining the current state of research about occupational class, gender and contractual inequalities, it is important to keep in mind the still fragmentary state of the evidence. As views of the types of inequality of interest widened, there was also an extension of the concept of job quality. In the early discussions of class inequalities, concern was almost exclusively with issues of skill and job control. The growing awareness of gender inequalities led to a stronger emphasis on pay differentials, training, career opportunities and work–​life balance. With the increased salience of the problematic of flexibility, job insecurity became central to notions of job quality. Research has focused predominantly on specific dimensions of job quality but, more recently, there have been attempts to create broader measures of job quality, whether through aggregated indices of conceptually related dimensions of jobs (Munoz de Bustillo et al. 2011; Green et al. 2013; OECD 2014; European Commission 2015) or through a single proxy indicator—​ such as relative pay position—​thought to provide a more comprehensive picture of the general quality of jobs (Fernandez-​Macias et al. 2012; Osterman and Shulman 2011).

Occupational Class Inequalities in Job Quality Occupational classes can be understood as aggregates of occupations which share similar broad skill levels (Tahlin 2013). Empirically, their relative skill level is usually indexed by the major groups in occupational classification schemes, although some studies have used relative pay as a proxy. Empirical research in both the US and Europe has highlighted the widely varying levels of job quality of employees in different occupational classes. Differences of class have been shown consistently to be associated with wide pay differentials. In some analyses, they are the most significant determinant of pay dispersion (Tahlin 2007), while a study of earnings differences in nine European countries concluded they account for between 40 per cent and 50 per cent of the variance in earnings, with remarkable similarity between countries (Eurofound 2017). Class also has been found to be a major determinant of non-​pecuniary job quality: whether people are in jobs that offer complex tasks, involve problem-​solving (Smith et al. 2008), and provide variety and new learning experiences (Handel 2012). There is evidence of a strong class gradient in job control and wider voice in the workplace (Gallie 2007b; Eurofound 2013a). Meta-​analysis also has confirmed that blue-​collar workers have higher levels of job insecurity (Keim et al. 2014). While evidence of class differences with respect to income, control, learning opportunities and security is both widespread and consistent, there is less evidence about trends. With respect to earnings, findings have varied but some research in the US has found that absolute class inequalities in wages increased substantially over the period 1973–​2005 (Weeden et al. 2007). Similarly, Mouw and Kalleberg (2010) found that about two-​thirds of the increase in wage inequality between 1992 and 2008 was due

inequality in job quality    323 to occupational differences between jobs, with much of the growth reflecting differences across aggregate groups of occupations (or classes). In the recent period, Europe followed a similar path, albeit with generally more modest increases in differentials. There were rising inequalities in earnings between 2006 and 2011 in two-​thirds of EU countries, with inequality increasing both at the top and the bottom of the distribution (Dreger et al. 2014). The increase was particularly sharp in Britain. One study of trends in wage inequality in Britain between the 1970s and the 2000s, has shown that it is inequality between occupations (and primarily inequality between occupational classes) that accounted for the greatest part of the growth of wage inequality (Williams 2013). However, for the period 2005 to 2014, an analysis of nine European countries, including the UK, concluded that, although there was some increase in between-​occupation inequality between 2005 and 2014, the rise in inequality was driven to a greater extent by changes in the distribution of wages within occupations (Eurofound 2017). In early debates on class inequality at work, the critical dimension of job quality was control over the work task. The available evidence on trends is far from consistent. Taking a summary index of work quality, which took account of skill, skill development and task discretion, Green et al. (2013) examined data for 15 countries in the EU over the period 1995–​2010. They reported a widening gap in work quality between professionals, whose jobs were getting better, and all other categories which either showed no substantial change in work quality or, in the case of service workers, a decline. Analyses, however, more specifically focused on task discretion have found diverse patterns of change between countries. In the 1990s, in a study of Britain, France, Germany, Sweden and Spain, only Britain and Spain experienced significant polarization (Gallie 2007b). Over the period 2004–​2010, there was little change in class differentials in European countries, apart from increased polarization in Britain and Ireland (Gallie and Zhou 2013). In other aspects of non-​pecuniary job quality, there is also little evidence of widespread class polarization in European countries. Increasing work intensity affected both higher and lower occupational classes (Green et al. 2013). Physical working conditions were notably better among higher classes (see also Baldry in the chapter entitled The Changing Quality of Office Work in this volume), but the overall class gap between the high skilled and the low skilled did not change. Finally working-​time quality showed a very general improvement across all occupational classes, apart from professionals. In short with respect to non-​financial job quality, it is primarily with respect to skill, skill development and, in some countries, task discretion that there is evidence of growing class inequality in European countries. Research on specific countries or groups of countries, however, points to a distinct ‘societal’ or ‘regime’ effect. A number of studies have shown that the overall quality of jobs is higher in the Nordic countries (Gallie 2003, 2007b; Esser and Olsen 2012; Green et al. 2013). These countries are characterized by higher levels of task discretion and workplace voice (Gallie and Zhou 2020, 2013; Eurofound 2013a), better continuing training provision (Dieckhoff 2013; Thompson and Hatfield 2015) and higher levels of job security (Green 2009). These advantages appear to be very stable over time. There is also some evidence that the Nordic countries had lower class differentials. They had

324   duncan gallie the lowest gap in job control between higher managers and professionals on the one hand and routine employees on the other, and, unlike the Liberal and Continental European countries, there was no tendency for the class gap to widen between 2004 and 2010, despite the potential weakening of the position of the less skilled in a period of severe economic crisis (Gallie and Zhou 2013). A five-​country comparison showed that the raw class gap in receipt of training was lower in Denmark and Finland than in France, the UK and Ireland (Dieckhoff et al. 2007). Moreover, in contrast to most other EU countries, the class gap declined between 2004 and 2010 in Denmark, Norway and Finland in part as a result of an increase in training provision for the least educated (Dieckhoff 2013). Further, Green et al. (2013) noted generally lower dispersion in the Nordic countries for the different dimensions of job quality, reflecting greater equality between different categories of employee.

Gender Inequalities in Job Quality Given the centrality of skill level to general job quality, a first approach is to examine how far women have been able to obtain jobs of a similar skill level to men. If higher-​ class position is taken as a broad indicator of a high skill level of jobs, then there is a clear pattern of women progressively increasing their share of higher-​skilled positions. For the period 1975–​2004, Tahlin (2007) examined changes in class distributions for men and women in five countries (Germany, Spain, France, the UK and Sweden). With the exception of France, the skill demands for women’s jobs rose faster than for men’s. Nevertheless, women remain disadvantaged in terms of access to higher-​skilled jobs. Schäfer et al. (2012), examining European Social Survey (ESS) data for the years 2002–​2008, found a very substantial under-​representation of women in managerial jobs (ISCO Category 1) in all of the twelve European countries examined. Even with respect to the labour market entry of relatively recent cohorts, Dämmrich (2015) found that, although in terms of gross rates women were more likely to enter managerial and professional occupational classes (ISCO 1 and 2), they were disadvantaged once individual differences—​for instance in educational level—​were taken into account. Further, Tahlin (2007) shows that there remained persisting gender differentials in skills within classes. Examining the skill requirements of jobs using a direct measure of skill—​the educational requirements needed to obtain a job—​men in professional and managerial occupations had higher-​skilled jobs than women in those occupations. Moreover, women were shown to have even stronger within-​class disadvantage in terms of firm-​ specific skills. Handel (2012) also found that women were in jobs requiring less education and shorter job learning times even after controlling for education and occupation. In short, despite educational equalization, women still have less access to the more skilled occupational classes and, even when they do enter such jobs, they tend to occupy less skilled positions within them. A core focus of research on gender inequality has been with respect to earnings. Although the extent of the earnings gap is sensitive to the specific measure used, there

inequality in job quality    325 remains a clear earnings gap in all of the OECD and EU countries for which there is data (OECD 2013, 2016). In 2019, women’s gross hourly earnings were on average 14.1 per cent below those of men in the EU (Eurostat 2022). There was a general tendency for the difference to decline between 2001 and 2014 but it remained pervasive. At the beginning of the period, the pay gap could be attributed in substantial part to differences in education but this has become a relatively minor factor with the marked increases in women’s educational attainment. Even at comparable levels of education, the average full-​year earnings of women in the EU countries in 2012/​13 were only around three-​quarters of those of men (OECD 2015). A good deal of the remaining gap can be attributed to the differences in occupational distributions—​women are disproportionately concentrated in relatively lower-​paid, highly-​feminized occupations. Nonetheless, as Levanon et al. (2009) have shown in a longitudinal analysis for the US, this outcome may partly reflect the fact that employers tend to value, and pay less, those occupations with high proportions of female employees. Women are less likely to be disadvantaged in terms of non-​pecuniary job quality. For instance, Smith et al (2008) found that in the EU-​27 in 2005, gender differences with respect to work monotony were not significant. Further the differences between men and women in key aspects of job autonomy such as work methods, the speed of work and the order in which tasks were done were relatively small. Men, however, were more likely to have control over breaks and have autonomy across a wider range of aspects of the work process. Although men had more autonomy in white-​collar jobs, it was women that were more autonomous in blue-​collar jobs. Studies of continuing vocational training have found that, in gross terms, men and women in European countries were about as likely to receive training, but estimates taking account of individual and work characteristics vary in their conclusions about whether it is men or women that are disadvantaged (Dieckhoff and Steiber 2011; Dieckhoff 2013). Women also tend to work in jobs with a better physical working environment, with fewer risks to health and safety (Fagan and Burchell 2002), with lower levels of work intensity and with higher working time quality (Green et al. 2013). A comparative analysis of job insecurity, drawing on data from 2005, found no significant difference between men and women in their worry about job loss, with the exception of the transition countries where women did feel more insecure (Green 2009). A more aggregated picture can be found from analyses based upon Eurofound’s non-​ pecuniary job quality index (NPI). This index showed that, in 2010, although less well paid, women were more likely than men to be in jobs with a high level of non-​pecuniary job quality (Eurofound 2013b). A study by the OECD, which classified job quality dimensions into a number of broad categories—​earnings quality, labour market insecurity and job strain—​reached a broadly similar conclusion for the wider set of economically advanced countries. While women were disadvantaged in terms of earnings quality, there was little difference with respect to insecurity. Moreover, on a composite measure of job strain, which took account of the balance between job demands and job resources, women were less likely to suffer from job strain than men (OECD 2014). An overall picture of gender inequalities in the labour market has to be qualified by the fact that, in many countries, female employment is highly differentiated. In

326   duncan gallie particular, it has been argued that women working part time are in a particularly vulnerable position and may constitute the type of secondary workforce depicted by labour market segmentation theorists (Tam 1997). There are clearly respects in which female part-​time workers experience significant disadvantage. Nonetheless, empirical studies to date present a complex picture. The unadjusted pay penalty for women in part-​time work is considerably greater than that for women in full-​time work, but the part-​time pay gap is also more substantially accounted for by differences in individual characteristics such as education, work experience and tenure. Part-​time workers experience disadvantages in terms of the complexity of the work they do, training, learning opportunities and career chances (Eurofound 2011). However, in the EU as a whole, they have a similar level of control over their work to female full-​timers (Gallie and Zhou 2013). Studies provide inconsistent conclusions as to whether they experience greater job insecurity (compare Green 2009 and Gash and Inanc 2013). There are also aspects of part-​timers’ work conditions that could be viewed as better than that of full-​timers—​for instance with respect to work intensity and opportunities for work–​life balance (Gallie et al. 2016). They also are less exposed to dangerous physical hazards, poor ergonomic conditions and job strain (Eurofound 2006; OECD 2014). There is again much less evidence on change over time, with the exception of the overall gender pay gap which has diminished substantially. Using a ‘relative pay’ proxy of job quality, Stehrer et al (2009) found that, between 1995 and 2005, in almost all EU countries, there was a decrease in women’s over-​representation in the bottom pay job quality quintile but little change in their under-​representation in the top quintile. For the period of the economic crisis, a European-​wide study (Eurofound 2013b) concluded that ‘w‌omen have increased their employment share, particularly in ‘mid-​paid’ and ‘good’ jobs (those in the higher quintiles)’. In effect, employment shifts have tended to be upgrading for women but more polarizing for men. Similarly, examining trends between 1995 and 2010 using more direct indicators of job quality from European Working Conditions Survey (EWCS) data, Green et al. (2013) found that while work quality (in the sense of skills, training and task discretion) was still higher for men at the end of the period, the differences were diminishing over time. In those respects in which women’s jobs were better than men’s in the mid-​1990s—​ namely in lower work intensity and lower physical risks—​gender differences had remained stable over time. Even in the case of female part-​time workers, there is some indication that there may have been an improvement in their position over the period 1995–​2000, at least as judged by a pay quintile proxy of good job quality (Stehrer et al. 2009). Moreover, although there was little change in their distribution in different pay quintiles in the period 2000–​2005, and in several countries a shift downwards, there was a fairly general and marked improvement in their position in the new member states. The overall picture then appears to be one of diminishing gender differentials over time with respect to skill and pay, and, at least for full-​time female employees, the disappearance of overall disadvantage with respect to intrinsic job quality.

inequality in job quality    327 As with class differences, there are marked differences between advanced countries in the extent of gender inequality. Schäfer et al. (2012) found that the proportion of women in managerial positions varied from 22 per cent in Germany to 36.2 per cent in the UK. An important factor accounting for country differences was the size of the service sector—​women had better chances of reaching high status occupations in countries with a large service sector (Charles 2003, 2005; Schäfer et al. 2012). There was no consistent demarcation between the Liberal and Coordinated Market economies, as had been suggested (Estevez-​Abe 2006). Rather there was a difference between Coordinated Market Economies. Women had significantly higher chances of being in a high-​status occupation in the Scandinavian countries, with their extensive welfare state service sector, than in the Continental CMEs—​such as Germany, Belgium, the Netherlands and Switzerland—​where public and private services were less developed. There are also marked country variations in the gender pay gap. In 2014, among Western countries, the unadjusted earnings gap was particularly high in Estonia (27%), Finland (20%), Austria (18%) as well as in the UK and US (17%), whereas it was only 7 per cent in Denmark, and 6 per cent in Norway and Belgium (OECD 2016). In an influential analysis, Blau (2003) argued that the gender earnings gap tended to be lower in countries where the male wage structure was more compressed, collective bargaining more centralized and wider in coverage, and female labour market supply more restricted. She attributed a central role to the strength of trade unions in affecting the extent of wage compression and increasing the extent of collective bargaining coverage. More recently, Schäfer and Gottschall (2015), comparing the full-​time worker gender wage gap in the health, manufacturing and finance industries in 25 EU countries found that countries with more generous minimum wage provisions and economy-​wide coordination of collective bargaining had lower gender wage gaps. However more highly centralized bargaining structures increased rather than reduced gender inequalities. They tentatively attributed this to the possibility that more centralized bargaining institutions may be subject to higher levels of male bias in a bargaining process dominated by trade unions representing primarily the interests of skilled male workers. There has been little systematic comparison of country differences in gender differentials in non-​pecuniary job quality. Gallie (2007b), comparing Britain, France, Germany, Sweden and Spain found that sex differences in task discretion were not evident in Britain and France, while men had slightly higher discretion in Sweden and women in Germany and Spain. The within-​country gender differences in task discretion largely reflected the types of jobs that workers held and there was no evidence of increased gender polarization in any of the countries between the early 1990s and the early 2000s. A wider comparison of EU countries, drawing on data from 2004 and 2010, showed that women had similar levels of task discretion to men in the Liberal countries (Britain and Ireland) and in France but had lower levels of discretion than men in the Nordic, Continental, Southern and Transition countries (Gallie and Zhou 2013). Differences were very stable across time and there was no evidence of either convergence or polarization. With respect to job insecurity, Green (2009), drawing on data for 2005, found that, while in general there was little difference in levels of insecurity

328   duncan gallie between men and women, women felt more insecure in the Nordic countries on the one hand and in the transitional countries on the other. The greater equalization of intrinsic job quality evident in the Nordic countries with respect to class differentials did not then extend to gender differences in job security.

Contract Security and Job Quality The predictions from the 1990s that there would be a substantial expansion of temporary work have received mixed support. In the EU between 2003 and 2019, there was an increase in the temporary work share of employees from 12.6 per cent to 13.6 per cent (Eurostat, 2019). Taking the OECD as a whole, the temporary work share of dependent employment declined a little from 12.7 per cent in 2006 to 11.4 per cent in 2019 (OECD 2010, 2021). The principal emphasis in the critical literature on the growth of the flexible labour market has been on its implications for the growth of job insecurity. The evidence here is very consistent: both male and female temporary workers experienced much higher levels of job insecurity than male standard employment contract workers in all the regions of Europe. The difference remained stable in most countries between 2004 and 2010 (Gash and Inanc 2013). In addition to insecurity, temporary employees have been found to experience a number of other significant disadvantages in their job quality. To begin with, they were disproportionately concentrated in lower-​skilled occupations. Whereas 24 per cent of all employees were in managerial or professional occupations in 2012, this was the case for only 18 per cent of temporary workers. Conversely, 9.2 per cent of the overall EU workforce was in low-​skilled elementary occupations, compared to 17 per cent of temporary employees.2 Similarly, temporary workers were more like to be in low-​wage employment (earning two-​thirds or less of the national median gross national earnings). Nearly a third (31.3%) of employees with a fixed-​term contract were low-​wage earners in 2010, compared to 15.7 per cent of employees with an indefinite contract (Eurostat 2016). There was an overall wage gap of 19 per cent between temporary and permanent employees, with the differential increasing the lower the occupational level (Eurofound 2015). Even with controls for other factors that influence pay (such as educational attainment and economic sector), there was an adjusted wage gap of 6 per cent. Further, temporary workers working full-​time hours, both male and female, were disproportionately exposed to outright pay cuts over the period of the economic crisis as well as to pay cuts that resulted from reductions in their hours of work (Gash and Inanc 2013).

2 Figures for temporary workers from Eurofound 2015 for all EU employed from Eurostat ‘Employment by Sex, Age, Professional Status and Occupation’, http://​app​sso.euros​tat.ec.eur​opa.eu/​nui/​ submit​View​Tabl​eAct​ion.do, accessed 20-​04-​2018.

inequality in job quality    329 The disadvantages experienced by temporary workers extended to non-​pecuniary aspects of job quality. In all of the major regions of Europe, they had lower levels of discretion over how to carry out their job tasks. These differences were generally stable across the period of the economic crisis, although the relative position of temporary workers deteriorated in the Continental European countries (Germany, Belgium and the Netherlands) and improved in the Transition countries (Gallie and Zhou 2013). In a five-​country study, drawing on data from the 1990s, Dieckhoff et al. (2007) found that temporary workers also generally had fewer continuing training opportunities. An examination of data for all EU countries in 2012 concluded that temporary workers were disadvantaged with respect to learning activities outside the education system and particularly with taught learning during working hours (Eurofound 2015). Green et al. (2013) report that since 2005 EU temporary agency workers (but not fixed-​term contract workers) have experienced higher levels of work intensity than permanent employees. An overview of the job quality of temporary work by the OECD (2014) concluded that it was associated with poor job quality on each of its three key job-​quality dimensions, with lower earnings, higher levels of labour market insecurity and higher job strain (in particular due to the combination of greater exposure to physical health-​risk factors at work and workplace intimidation, while having less autonomy, fewer learning opportunities and less social support). As a result of these cumulative disadvantages, temporary work is associated with an increased risk of psychological and physical ill-​health (Benach et al. 2014; Virtanen et al. 2005. In a longitudinal study of Finnish workers, Kivimaki et al. (2003) found that, even controlling for age, occupational status, salary and change in occupational title, overall mortality was 1.2 to 1.6 times higher among male and female temporary employees compared with permanent employees. Similarly, examining the health effects of leaving unemployment for permanent and temporary work in Germany and Spain, Gash et al. (2007) showed that those workers who obtained temporary jobs had less improvement in their health status than those who moved into permanent jobs. An assessment (Stehrer et al. 2009) of trends over time in terms of a pay quintile proxy for overall job quality found a small improvement in the quality of temporary workers jobs in the EU-​15 between 1995 and 2005, although there was very little change in most countries. Moreover, there was no clear pattern of change in the New Member States for the period for which there was data (2000–​2005). Despite the average improvement, there was an increase in the relative number of fixed-​term jobs in the bottom quintile between 2000 and 2005. Using an overall measure of intrinsic work quality (including skill, skill development and task discretion), Green et al. (2013) noted no change for fixed-​ term contract employees, but an improvement over the period 2005–​2010 for those on a temporary employment agency contract. However, temporary agency workers also experienced a particularly sharp increase in work intensity after 2000. There is no evidence of change in the level of insecurity experienced by temporary workers between 2004 and 2010, with the exception of the Liberal countries (the UK and Ireland) where their insecurity grew greater (Gash and Inanc 2013). But there was a general decrease in transition rates in the 2000s, with the proportion of temporary employees moving into

330   duncan gallie permanent positions in two-​thirds of European countries declining from 28 per cent in 2006 to 20 per cent in 2012 (Eurofound 2015). Overall, in the 2000s, there appears to have been little change in the job quality of temporary workers, with the exception of agency workers, but there was an increase in levels of entrapment in this type of work. While the evidence about temporary worker disadvantage is largely consistent across the different dimensions of job quality and across European countries, there are significant differences between countries, both in the prevalence of fixed-​term contracts and their distribution among different age groups. In 2019, temporary workers represented over 20 per cent of the workforce in countries such as Portugal 21%; Spain, 26%; Poland, 22%, whereas they constituted less than 7 per cent of the workforce in Lithuania, 1.5%; Estonia, 3.1%; Latvia, 3.2%; UK, 5.2% (OECD 2021). Countries also differed in the extent to which temporary contracts were concentrated on younger employees or were spread more evenly across the various age cohorts. In the EU as a whole, nearly 40 per cent of employees aged 15–​24 years were in the temporary workforce. However, more than half of the youngest employees were in temporary employment in Germany, Spain, France, Poland, Portugal, Sweden and Slovenia. By contrast, less than one in ten were in temporary employment in Bulgaria, Latvia, Lithuania and Romania (European Commission 2010). In general, there has been a much weaker relationship between age and type of contract in the new member states. Most crucially from the point of inequality, countries differed substantially in terms of the risks of relatively long-​term entrapment in temporary work. Temporary work is sometimes depicted as a useful bridge into permanent employment, but prolonged experience is likely to lead to labour-​market marginalization. There are substantial country variations in absolute transition rates. In 2013 less than a quarter of temporary workers moved within a year to a permanent job in the Southern countries of Italy, Portugal, Greece and Spain (as well as in the Netherlands, Poland and France). In contrast, more than 50 per cent moved to a permanent job in Latvia, Lithuania, the UK and Estonia (European Commission 2015). Transition rates tended to be lower in countries with high rates of temporary employment and in countries where the young constituted a particularly high proportion of temporary workers (Eurofound 2015). Moreover, the decline over time in transition rates was particularly severe among younger workers. The contractual divide thus represents a rather different phenomenon in different parts of Europe—​with relatively high levels of segment closure in Southern Europe and France, medium levels of closure in Continental and Nordic Europe and relatively high levels of fluidity in Eastern Europe and the UK.

Conclusions While the evidence strongly confirms the view that class, gender and employment contract are major sources of inequality in job quality, there are differences in the particular dimensions of job quality that are involved and in the extent to which disadvantages

inequality in job quality    331 are cumulative. Class is associated with differences in job quality across most of the major dimensions examined: earnings, job control, training opportunities and job security. The exceptions are work intensity and working-​time quality, for which there was no clear class gradient. Women (and female part-​timers) are primarily disadvantaged with respect to the skill level and pay of jobs. They have similar levels of job control to men. In some respects—​work intensity, job strain and working time quality—​their job quality is better than that of men. The most consistently disadvantaged are temporary employees: quite apart from their much higher levels of insecurity, they are in lower-​ skilled work, have less control over their jobs, are disadvantaged with respect to training opportunities and (in the case of agency workers) are subject to higher work intensity. The major scenarios of change in patterns of inequality notably fail to provide an account that captures empirical trends with respect to each of the major sources of inequality, although both optimistic and pessimistic theories of change could claim some evidence in support. The clearest case of a measure of convergence is with respect to gender inequalities in the labour market. Women have improved not only their employment rates but also their level of access to higher-​skilled jobs and their relative earnings. Even female part-​time work may have improved a little with respect to the skill level of jobs, although this improvement occurred primarily in the period 1995–​2000. However, despite having higher levels of educational attainment than men, even the most recent cohorts of female full-​timers are still at a relative disadvantage in terms of entry into the highest-​level jobs. Moreover, there is still a significant earnings differential that cannot be accounted for by education. It is only in the non-​pecuniary aspects of job quality that women’s jobs are as good as those of men. The more pessimistic scenarios account better for the evidence of trends in inequalities with respect to class and contract type. There is some evidence that class earnings have diverged in recent decades, as have differences of skill and opportunities for skill development, while differentials in job control have been either stable or grown greater. The job quality disadvantages of temporary work have been mainly stable, even in terms of job insecurity. However, there has been deterioration in the transition rates into permanent employment, which are likely to be highly detrimental to workers’ longer-​term careers. Finally, while the inequalities associated with class, gender and employment contract are very general across the European societies, their extent varies depending on specific national contexts. Employment regime theory, with its emphasis on the balance of power between employers and labour, may help to account to some extent for class differences in job quality—​in particular for the fact that class differences were lower in the Nordic countries than in other ‘regimes’. However, it does not provide an adequate account of gender inequalities: although women’s access to higher-​level jobs was better in the Nordic countries than in the Continental European countries, it was not better than in a Liberal regime country such as the UK. Moreover, the relative disadvantage of women in the Nordic countries with respect to both job control and job security was greater than in the Liberal countries. Production regime theory, with its emphasis on employer coordination and vocational training, provides little explanation of

332   duncan gallie differences in national patterns of job quality. Perhaps the most problematic aspect for both of the existing macro-​institutional theories is that there appear to have been lower inequalities with respect to temporary work in the Liberal and Transition countries than in others. The use of temporary work was more limited, it was less concentrated among younger workers and there were relatively high transition rates into permanent employment. This finding suggests that, rather than assuming that different types of institutional regime will be related to better or worse inequalities in job quality across all groups of employees, it is important to assess whether they have distinctive and potentially varying effects for specific types of workforce stratification.

References Appelbaum, E., Bailey, T., Berg, P., and Kalleberg, A. L. (2000), Manufacturing Advantage (Ithaca: Cornell University Press). Autor, D. H., Levy, F., and Murnane, R. J. (2003), ‘The Skill Content of Recent Technological Change: An Empirical Exploration’, Quarterly Journal of Economics, (November), 1279–​1333. Baldry, C. (2022), ‘The Changing Quality of Office Work’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press). Benach, J., Vives, A., Amable, M., Vanroelen, C., Tarafa, G., and Muntaner, C. (2014), ‘Precarious Employment: Understanding an Emerging Social Determinant of Health’, Annual Review of Public Health, 35, 229–​253. Blau, F. D., and Kahn, L. M. (2003), ‘Understanding International Differences in the Gender Pay Gap’, Journal of Labor Economics, 21, 1, 106–​144. Blauner, R. (1964), Alienation and Freedom (Chicago, University of Chicago Press). Boxall, P., and Purcell, J. (2022), ‘Human Resource Management and Job Quality’, in C. Warhurst, C. Mathieu, and R. E. Dwyer, eds., Oxford Handbook of Job Quality (Oxford: Oxford University Press). Braverman, H. (1974), Labor and Monopoly Capital (New York, Monthly Review Press). Cappelli, P., Bassi, L., Katz, H., Knoke, D., Osterman, P., and Useem, M. (1997), Change at Work (Oxford: Oxford University Press). Charles, M. (2003), ‘Deciphering Sex Segregation: Vertical and Horizontal Inequalities in Ten National Labor Markets’, Acta Sociologica, 46, 4, 267–​287. Charles, M. (2005), ‘National Skill Regimes, Postindustrialism, and Sex Segregation’, Social Politics, 12, 2, 289–​316. Dämmrich, J. (2015), ‘Gendered Labor Market Outcomes at Labor Market Entry and Their Relationship with Country-​ Specific Characteristics’, in H.-​ P. Blossfeld, J. Skopek, M. Triventi, & S. Buchholz, eds., Gender, Education and Employment. An International Comparison of School-​To-​Work Transitions (Cheltenham, UK/​Northampton, MA: Edward Elgar), 60–​82.. Dieckhoff, M. (2013), ‘Continuing Training in Times of Economic Crisis’, in D. Gallie, ed., Economic Crisis, Quality of Work and Social Integration (Oxford: Oxford University Press), 77–​103. Dieckhoff, M., Jungblut, J. M., and O’Connell, P. J. (2007), ‘Job-​ Related Training in Europe: Do Institutions Matter?’ in D. Gallie, ed., Employment Regimes and the Quality of Work (Oxford: Oxford University Press), 88–​114.

inequality in job quality    333 Dieckhoff, M. and Steiber, N. (2011), ‘A Re-​Assessment of Common Theoretical Approaches to Explain Gender Differences in Continuing Training Participation’, British Journal of Industrial Relations, 49, S1, s135–​s157. Dreger, C., Lopez-​Bazo, E., Ramos, R., Royuela, V., and Surinach, J. (2014), Wage and Income Inequality in the European Union (Brussels: Directorate-​General for Internal Policies European Parliament). Esser, J., and Olsen, K. M. (2012), ‘Perceived Job Quality: Autonomy and Job Security within a Multi-​Level Framework’, European Sociological Review, 28, 4, 443–​454. Estevez-​Abe, M. (2006), ‘Gendering the Varieties of Capitalism: A Study of Occupational Segregation by Sex in Advanced Industrial Societies’, World Politics, 59, 142–​175. Eurofound (2006), Part-​ Time Work in Europe (Dublin: European Foundation for the Improvement of Living and Working Conditions). Eurofound: Sándor, E. (2011), Part-​Time Work in Europe. European Company Survey 2009 (Luxembourg, Publications Office of the European Union. Eurofound: Gallie, D., and Zhou, Y. (2013a), Work Organisation and Employee Involvement in Europe. A Report Based on the Fifth European Working Conditions Survey (Luxembourg: Publications Office of the European Union). Eurofound: Hurley, J., Fernandez-​Macias, E., and Storrie, D. (2013b), Employment Polarisation and Job Quality in the Crisis: European Jobs Monitor 2013 (Dublin: Eurofound. Eurofound: Vacas-​ Soriano, C. (2015), Recent Developments in Temporary Employment: Employment Growth, Wages and Transitions (Luxembourg, Publications Office of the European Union). Eurofound: Fernandez-​Macias, E., Hurley, J., and Arranz-​Munoz, J. M. (2017), Occupational Change and Wage Inequality: European Jobs Monitor 2017 (Luxembourg: Publications of the European Union European Commission (2010), Employment in Europe 2010 (Luxembourg, Publications Office of the European Union). European Commission (2015), Employment and Social Developments in Europe 2014 (Luxembourg, Publications Office of the European Union). Eurostat (2019), ‘Temporary Employment in Q2 2019’, Eurostat News (Brussels: European Commission). https://​ec.europa.eu/​eurostat/​web/​products-​eurostat-​ news/​-/​ ​ddn-​20191120-​1 Eurostat (2022), ‘Gender Pay Gap Statistics’, Eurostat Statistics Explained (Brussels: European Commission). https://​ec.europa.eu/​eurostat/​statistics-​explained/​index.php?title=​Gender_​ pay_​gap_​statistics#Gender_​pay_​gap_​levels_​vary_​significantly_​across_​EU Fagan, C., and Burchell, B. (2002), Gender, Jobs and Working Conditions in the European Union (Luxembourg, Office for Official Publications of the European Communities). Fernandez-​Macias, E., Hurley, J., and Storrie, D. (2012), Transformation of the Employment Structure in the EU and USA, 1995-​2007 (Basingstoke: Palgrave Macmillan). Gallie, D. (2003), ‘The Quality of Working Life: Is Scandinavia Different?’ European Sociological Review, 19, 1, 61–​79. Gallie, D. (2007a). ‘Production Regimes, Employment Regimes, and the Quality of Work’, in D. Gallie, ed., Employment Regimes and the Quality of Work (Oxford: Oxford University Press), 1–​33. Gallie, D. (2007b) ‘Task Discretion and Job Quality’, in D. Gallie, ed., Employment Regimes and the Quality of Work (Oxford: Oxford University Press), 105–​136. Gallie, D., Gebel, M., Giesecke, M., Hallden, K., Van Der Meer, P., and Wielers, R. (2016), ‘Quality of Work and Job Satisfaction: Comparing Female Part-​Time Work in Four European Countries’, International Review of Sociology, 26, 3, 457–​481.

334   duncan gallie Gallie, D., and Zhou, Y. (2013), ‘Job Control, Work Intensity and Work Stress’, in D. Gallie, ed., Economic Crisis, Quality of Work and Social Integration: The European Experience (Oxford: Oxford University Press), 115–​141. Gallie, D., and Zhou, Y. (2020), Employee Involvement, Work Engagement and Skill Development (Dublin: European Foundation for the Improvement of Living and Working Conditions). https://​www.eurofound.europa.eu/​sites/​default/​files/​wpef19061.pdf Gash, V., and Inanc, H. (2013), ‘Insecurity and the Peripheral Workforce’, in D. Gallie, ed., Economic Crisis, Quality of Work and Social Integration (Oxford: Oxford University Press), 142–​168. Gash, V., Mertens, A., and Romeu-​Gordo, L. (2007), ‘Are Fixed-​Term Jobs Bad for Your Health? A Comparison of Spain and Germany’, European Societies, 9, 3, 429–​458. Goos, M., and Manning, A. (2007), ‘Lousy and Lovely Jobs: The Rising Polarization of Work in Britain’, Review of Economics and Statistics, 89, 1, 118–​133. Green, F. (2009), ‘Subjective Employment Security around the World’, Cambridge Journal of Regions, Economy and Society, 2, 3, 343–​363. Green, F., Mostafa, T., Parent-​Thirion, A., Vermeylen, G., Van Houten, G., Biletta, I., and Lyly-​ Yrjanainen, M. (2013), ‘Is Job Quality Becoming More Unequal?’ ILRReview, 66, 4, 753–​784. Hall, P. A. and Soskice, D. E. (2001), Varieties of Capitalism: The Institutional Foundations of Comparative Advantage (Oxford, Oxford University Press). Handel, M. J. (2012), ‘Trends in Job Skill Demands in OECD Countries’, OECD Social, Employment and Migration Working Papers, No. 143. Paris: OECD. Kalleberg, A. L. (2011), Good Jobs, Bad Jobs (New York, Russell Sage Foundation). Keim, A. C., Landis, R. S., Pierce, C. A., and Earnest, D. R. (2014), ‘Why Do Employees Worry About Their Jobs? A Meta-​Analytic Review of Predictors of Job Insecurity’, Journal of Applied Health Psychology, 19, 3, 269–​290. Kern, H. and Schumann, M. (1987), ‘Limits of the Division of Labour: New Production and Employment Concepts in West German Industry’, Economic and Industrial Democracy, 8, 1, 151–​170. Kern, H., and Schumann, M. (1992), ‘New Concepts of Production and the Emergence of the Systems Controller’, in P. S. Adler, ed., Technology and the Future of Work (Oxford University Press: New York), 111–​142. Kerr, C., Dunlop, J. T., Harbison, F., and Myers, C. A. (1960), Industrialism and Industrial Man (Cambridge, Mass.: Harvard University Press). Kivimaki, M., Vahtera, J., Virtanen, M., Elovainio, M., Pentii, J., and Ferrie, J. E. (2003), ‘Temporary Employment and Risk of Overall and Cause-​Specific Mortality’, American Journal of Epidemiology, 158, 7, 663–​668. Korpi, W. (2006), ‘Power Resources and Employer-​Centered Approaches in Explanations of Welfare States and Varieties of Capitalism. Protagonists, Consenters and Antagonists’, World Politics, 58, 167–​206. Lawler, E. (1986), High Involvement Management (San Francisco, Josey-​Bass). Lawler, E., Mohrman, S. A. and Ledford, G. E. (1995), Creating High Performance Organizations (San Francisco, Jossey-​Bass). Levanon, A., England, P., and Allison, P. (2009), ‘Occupational Feminization and Pay: Assessing Causal Dynamics Using 1950-​2000 US Census Data’, Social Forces, 88, 2, 865–​891. Mouw, T., and Kalleberg, A. L. (2010), ‘Occupations and the Structure of Wage Inequality in the United States 1980s to 2000s’, American Sociological Review, 75, 3, 402–​431. Munoz De Bustillo, R., Fernandez-​Macias, E., Anton, J.-​I., and Esteve, F. (2011), Measuring More than Money. The Social Economics of Job Quality (Cheltenham, UK, Edward Elgar).

inequality in job quality    335 OECD (2008), Growing Unequal? Income Distribution and Poverty in OECD Countries (Paris: OECD). OECD (2010) Employment Outlook 2010. Moving Beyond the Jobs Crisis. (Paris, OECD). OECD (2013), OECD Employment Outlook 2013 (Paris: OECD). OECD (2014), ‘How Good is Your Job? Measuring and Assessing Job Quality’, in OECD ed., OECD Employment Outlook 2014 (Paris: OECD), 79–​129. OECD (2015), Education at a Glance 2015 OECD Indicators (Paris: OECD). OECD (2016) OECD Employment Outlook 2016 (Paris, OECD) OECD (2021), Employment Outlook 2021 (Paris: OECD). Osterman, P., and Shulman, B. (2011), Good Jobs America (New York: Russell Sage Foundation). Piore, M. J., and Sabel, C. F. (1984), The Second Industrial Divide (New York, Basic Books). Schäfer, A., and Gottschall, K. (2015), ‘From Wage Regulation to Wage Gap: How Wage-​Setting Institutions and Structures Shape the Gender Gap across Three Industries in 24 European Countries and Germany’, Cambridge Journal of Economics, 39, 467–​496. Schäfer, A., Tucci, I., and Gottschall, K. (2012), ‘Top Down or Bottom Up? A Cross-​National Study of Vertical Occupational Sex Segregation in Twelve European Countries’, in F. Engelstad and M. Teigen, eds., Firms, Boards and Gender Quotas (Bingley: Emerald), 3–​43. Smith, M., Burchell, B., Fagan, C., and O’Brien, C. (2008), ‘Job Quality in Europe’, Industrial Relations Journal, 39, 6, 585–​602. Soskice, D. (1999), ‘Divergent Production Regimes: Coordinated and Uncoordinated Market Economies in the 1980s and 1990s’, in H. Kitschelt, P. Lange, G. Marks, and J.D. Stephens, eds., Continuity and Change in Contemporary Capitalism (Cambridge: Cambridge University Press), 101–​134. Stehrer, R., Ward, T., and Fernandez-​ Macias, E. (2009), ‘Changes in the Structure of Employment in the EU and Their Implications for Job Quality’, Research Report 354. Vienna: The Vienna Institute for International Economic Studies. Tahlin, M. (2007), ‘Skills and Wages in European Labour Markets: Structure and Change’, in D. Gallie, ed., Employment Regimes and the Quality of Work (Oxford: Oxford University Press), 35–​76. Tahlin, M. (2013), ‘Distribution in the Downturn’, in D. Gallie, ed., Economic Crisis, Quality of Work and Social Integration (Oxford: Oxford University Press), 35–​76. Tam, M. (1997). Part-​Time Employment: A Bridge or a Trap? (Aldershot: Avebury). Thompson, S., and Hatfield, I. (2015), Employee Progression in European Labour Markets (London: Institute for Public Policy Research). Virtanen, M., Kivimaki, M., Joensuu, M., Virtanen, P., Elovainio, M., and Vahtera, J. (2005), ‘Temporary Employment and Health: A Review’, International Journal of Epidemiology, 34, 3, 610–​622. Walton, R. E. (1985), ‘From Control to Commitment in the Workplace’, Harvard Business Review, 85, 2, 77–​84. Weeden, K. A., Kim, Y. M., Di Carlo, M., and Grusky, D. B. (2007), ‘Social Class and Earnings Inequality’, American Behavioral Scientist, 50, 5, 702–​736. Williams, M. (2013), ‘Occupations and British Wage Inequality, 1970s-​ 2000s’, European Sociological Review, 29, 4, 841–​857. Wood, S., and Wall, T. (2007), ‘Work Enrichment and Employee Voice in Human Resource Management-​Performance Studies’, International Journal of Human Resource Management, 18, 7, 1335–​1372.

part I V

R E G IONA L DE V E L OP M E N T S I N J OB QUA L I T Y

chapter 15

job quality in the United States and Canada arne l. kalleberg, sylvia fuller and ashley pullman

Introduction: What is Job Quality and Why is it Important? The topic of job quality has occupied a prominent place in discussions of the changing nature of work and employment relations for many decades and has always attracted a great deal of attention from a diverse group of social scientists. The notion of job quality communicates that it is the nature of work that is important to workers, not just whether they have any job at all. Writing in 2000, Graham Lowe (2000: ix) maintained that the ‘most pressing issue about the past, present, and future of work is, in a word, quality’, observing that ‘[t]‌oday the European focus is on job quality, whereas in North America the main goal is to generate lots of jobs, regardless of their quality’ (p. 20). These sentiments are even truer today, as concerns about the growth of precarious and low-​ paying jobs have made it even more pressing. As Findlay et al. (2013: 442) note: Job quality is back in vogue among social scientists and policy-​makers because of its potential impact on individual, firm and national well-​being. For economies in trouble, the impacts of better job quality on lower unemployment and higher employment participation means that improving job quality offers the promise of salvation; for economies doing well, the influence of good jobs on innovation and enhanced productivity offers justification for policy.

Worries about the quality of jobs are not new. The prevalence of bad jobs—​such as marginal and irregular work—​was common among the labouring classes in industrial countries in the nineteenth century. The Philadelphia Declaration, adopted by the International Labour Organization (ILO) in 1944, set out a set of workers’ rights that

340    arne l. kalleberg, sylvia fuller and ashley pullman were centred on improving the quality of work. Years later, in 1999, the new ILO director, Juan Somavia, launched the ‘decent work’ initiative that sought to achieve similar goals regarding job quality. The European Union (EU) also adopted the idea of decent work as a goal in the following decade. The concerns expressed by the ILO, EU as well as other policymakers and academics underscore the importance of considering the correlates of job quality, especially inequality in job quality and the presence of bad jobs. This chapter will report on the significant pre-​pandemic trends and developments in job quality in the United States and Canada. These two countries are usually identified as representing liberal market economies that have many similarities, both socially and economically. Both underwent a period of neoliberal reforms beginning in the 1970s and 1980s. Both suffered from the Great Recession (2008–​2009). And in both countries, the ratio of good jobs to bad jobs increased sharply in the twentieth century as unions and governments helped to construct labour market institutions that provided relatively well-​paying and secure jobs. Despite these similarities, these two countries exhibit ‘small differences’ in policies and institutions that matter for considering the question of job quality in these two countries (Card and Freeman 1993). Looking at these differences both advances our theoretical understanding of institutional sources of inequality in job quality as well as provides useful information for policymakers about the quality of work experienced by workers. We first discuss how we conceptualize and measure the concept of job quality. We then consider some of the differences in labour market, demographic and welfare institutions that help to explain differences in the patterns of job quality between the US and Canada. We next provide empirical information on levels and trends in job quality in the two countries. We finally highlight some of the major issues regarding job quality facing policymakers as well as researchers in the US and Canada.

Measuring Job Quality Jobs are complex and can potentially provide workers with a vast array of benefits and utilities that may be regarded as rewards from work. Thus, it is unsurprising that most researchers have proposed a number of multidimensional definitions of job quality. Some analysts have identified a large number of dimensions: the International Labour Organization’s conceptualization of ‘decent work’, for example, includes nearly a dozen components (each comprised of numerous indicators) such as employment opportunities, adequate earnings, decent hours, stability and security of work, arrangements to combine work and family life, fair treatment in employment, a safe work environment, social protections, social dialogue and workplace relations, and characteristics of the economic and social contexts of work (e.g. Ghai 2003). The European Commission’s (2001) related concept of decent work similarly includes ten components, such as intrinsic job quality, skills, gender equality, health and safety at work, flexibility and security, and work–​life balance (see also Green 2006). Researchers

job quality in the United States and Canada    341 have tried to reduce the dimensions of job quality to a more manageable number, using data reduction techniques such as cluster analysis. Holman (2013), for example, combined 38 indicators into five dimensions: work organization, wages and payment system, security and flexibility, skills and development, and representation and engagement (see also de Bustillo et al. 2011). The broadness of the concept of job quality as well as the vagueness and all-​encompassing nature of measures of it (such as those proposed by the ILO and EU) have made it difficult to agree on how to measure job quality and to obtain internationally comparable data on it (Burchell et al. 2013). Researchers thus have made strategic decisions about which dimensions to consider and measure. While it would be good to have an overall, parsimonious measure of job quality in the economy that can be used as a diagnostic tool and social indicator, any such measure will necessarily be a simplification of a complex phenomenon. Thus, Kalleberg (2011), for example, concentrates on earnings and benefits, control over work and intrinsic rewards, and work scheduling and family flexibility. In this chapter, we will also measure job quality in terms of several key dimensions: wages, job security, work–​family flexibility and skill utilization. We will focus especially on the degree of polarization in job quality, especially between high-​and low-​wage jobs and between standard and nonstandard employment relations such as temporary and involuntary part-​time work, as well as job–​person skill mismatches such as overqualification. These components of job quality are often related: high-​paying jobs are usually associated with high utilization of one’s (relatively high) skills and provide workers with opportunities for schedule flexibility, at least compared to low-​wage jobs. Low-​wage jobs are also typically associated with lower skill requirements, though this association likely varies by country (e.g. low-​wage workers tend to be relatively more skilled in Germany than in the United States) and depends on how one defines skills (many ‘unskilled’ jobs require many tacit skills). Fortunately, governmental agencies in Canada and the US use similar methods, surveys and classification systems for their labour force statistics, enhancing comparability in our selected indicators across the two countries. Where possible, we base our analysis on publicly available data organized specifically for international comparisons. In other cases, we generate our own figures and describe how we harmonized the data.

Small Differences that Matter: US and Canadian Labour Market and Welfare Institutions and Demographic Characteristics This section provides an overview of some of the key institutional differences between the US and Canada that help to explain dissimilarities in job quality in the two countries (which we discuss in the next section).

342    arne l. kalleberg, sylvia fuller and ashley pullman

Labour Market Institutions and Regulatory Protections Table 15.1 presents some information on labour market regulation and social spending in the US and Canada. Both countries have relatively flexible labour markets with weak regulations. For example, both countries have low scores compared to other OECD countries for employment protection, with relatively few protections against dismissal for permanent workers (with the United States ranking lowest among OECD countries and Canada third lowest) and few restrictions on the use of temporary workers (here Canada and the United States are both the lowest in the OECD). Although welfare state scholars typically group Canada and the United States together as countries sharing an individualistic market-​oriented approach, there are nonetheless ‘small differences that matter’ in labour market institutions (Card and Freeman 1993: 1). Among OECD countries, the US is ranked among the lowest with respect to expenditures on active labour market policies. Canada devotes relatively more resources on public expenditures on active labour market policies (.21 percent of GDP), though it still falls well below the OECD average of .63 per cent of GDP. In 2019, the United States spent relatively more on passive income supports to out-​of-​work individuals. Canada also provides stronger statutory entitlements for workers with respect to protected time away from employment. All Canadian provinces require employers to provide workers with paid vacation after (typically 12) months of employment, with the entitlement increasing with years of seniority. In addition, employers are required to pay workers for ‘statutory holidays’ (typically around 9 per year) for which they are not scheduled to work or, if scheduled to work on these days, to pay a wage premium. Canadian employees are entitled to additional days of unpaid leave to attend to family responsibilities, as well as at least 15 months of unpaid job-​protected leave after the birth of a child. Government-​provided paid parenting leaves are also available across the country.1 The United States does not have comparable leave entitlements at either

1 Provisions for unpaid job-​ protected parenting leaves vary somewhat across sub-​ national jurisdictions in Canada, with most requiring a period of continuous employment for qualification. Federal maternity and parental leave policy excludes Quebec, which, since 2006, provides leave to parents under the Quebec Parental Insurance Program. For the rest of Canada (ROC), the federal Employment Insurance programme provides paid family leave. For employees outside Quebec who work at least 600 insurable hours in the past year, the federal government will reimburse 55 per cent of earnings (up to a ceiling) for 15 weeks of maternity and 35 weeks of parental leave. Parents can opt to spread parental leave over a longer period for the same total benefit level (61 weeks at 33% of earnings). Low-​income families are eligible for an EI family supplement, extending benefits up to 80 per cent of average earnings. In Québec, benefits are more generous and conditions for entitlement are much less strict. Since 2010, self-​employed workers can opt to pay into the employment insurance programme to access paid maternity and parental leave (for details of Canada’s leave policies see Doucet et al. 2021). Most self-​employed workers, however, do not currently opt into the programme, and more marginally employed part-​time workers and others unable to accumulate enough hours of employment in the prior year also fail to qualify for benefits. In 2014, 25 per cent of recent mothers did not have insurable employment and were unable to access maternity or parental benefits (Statistics Canada 2015). Eligibility is particularly low for youth between the ages of 15 and 24—​only 44 per cent in 2014. Low-​income

job quality in the United States and Canada    343 Table 15.1 Labour market policies and public expenditures Employment Protections, 2019

United States

Canada

Employment protection against individual dismissal (range: 0.7–​4.2)1

0.7

0.7

Employment protection for permanent workers (range: 1.3–​3.0)1

1.3

1.7

1

0.3

0.3

Active measures, 20192

0.11

0.21

Out-​of-​work income maintenance and support, 20192

0.79

0.50

0.90

0.71

7.0

4.9

Employment protection for temporary workers (range: 0.3–4.5)

Public expenditure on labour market policies as a percentage of GDP

Total, 2019

2

Pensions, 20183

1 OECD. 2020d. OECD Employment Outlook 2020: Worker Security and the COVID-​19 Crisis (Paris: OECD

Publishing). https://​doi.org/​10.1787/​1686c​758-​en Note: These indicators capture the cost and difficulty of dismissing groups of workers. 2 OECD. 2022e. Labour market programmes: Expenditure and participants (indicators) (OECD Statistics).

https://​doi.org/​10.1787/​data-​00312-​en Note: Active measures consist of expenditures on PES and administration, training, employment incentives, sheltered and supported employment and rehabilitation, direct job creation, and start-​up incentives. 3 OECD. 2022f. Pension spending (indicators) (OECD Statistics). https://​doi.org/​10.1787/​a041f4ef-​en

the federal or state level. The 1993 Family and Medical Leave Act provides for 12 weeks of job-​protected but unpaid leave to cover dependent care responsibilities or personal illness, but approximately 40 per cent of American workers are ineligible for this leave due to limitations around hours of work and size of firm (Ray et al. 2008). A few states have recently introduced paid family leave, but at levels considerably less generous than are available in Canada. Another important difference between the two countries is in union density (see Figure 15.1), which is reflective of a generally more supportive regulatory environment for union organizing in Canada than the United States. Union density was about 30 per cent in each country in 1960 (indeed, it was slightly higher in the US), but it has progressively declined since the 1960s in the US; in Canada, union density declined in the 1990s after a peak in the early 1980s, though it is now about where it was in 1960. As a result, the proportion of workers belonging to unions is now much higher in Canada (27%) than in the US (10%). Social confidence in unions among the working class is also higher in Canada than in the US (Frangi et al. 2014).

mothers are also much less likely to qualify for maternity and/​or parental benefits, with only 44 per cent of those with household incomes less than $30,000 qualifying in the ROC versus 74 per cent of those with incomes above $60,000. In Quebec, household income matters much less, with 85.4 per cent of the low-​income group qualifying and 95 per cent of the high income group (McKay et al. 2016).

344    arne l. kalleberg, sylvia fuller and ashley pullman Trade union density, 1960-2020 40 35 30 25 20 15 5

1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020

10

Canada

United States

Figure 15.1  Trade union density in the United States and Canada, 1960–​2020 Source: OECD. 2021. Trade unions: Trade union density (indicators) (Paris: OECD Statistics). https://​stats.oecd.org/​Index.aspx?DataSetCode=​TUD

The public sector is considerably larger in Canada than in the US, due in part to the public nature of mainstream medical care. The Canadian public sector constitutes about one third of overall employment; in the US, the government employs about 20 per cent either directly or indirectly. Historically, the public sector has been an important source of ‘good’ jobs in both countries, as public-​sector workers have enjoyed higher than average levels of employment security, pension coverage and workplace benefits, as well as smaller net gender and racial wage gaps (Fuller 2005; Johnson 2002; Ilcan et al. 2003, 2007; Hou and Coulombe 2010; Lucifora and Meurs 2006; Wilson et al. 2015). The public sector is also more powerful in Canada insofar as union density within it is considerably higher than in the United States. Where a bit more than a third (34.4%) of American public-​sector workers were unionized in 2016, over three-​quarters (76.3%) were union members in Canada (Statistics Canada 2017; United States Bureau of Labor Statistics 2017). Resource and extractive industries (agriculture, hunting, forestry, fishing, mining and oil and gas extraction) are also a larger component of the industrial structure in Canada (especially in the western provinces), which contributes to high-​paying jobs for men but also to greater volatility in unemployment during periods of boom and bust (Fortin and Lemieux 2015; Green and Sand 2015). In Canada, resource industries accounted for 15.5 per cent of value added to the Canadian economy and 5.1 per cent of hours worked in 2008, compared to 5 per cent of value added and 3.2 per cent of hours worked in the United States (Tang 2017). Inter-​provincial differences in reliance on the resource sector contribute to uneven trends for wage polarization across Canada (Green and Sand 2015). Although provinces with large resource and energy sectors historically have smaller wage dispersion (Fortin and Lemieux 2015), the effect of downturns in the energy sector have an adverse effect on employment in these regions (Gellatly 2016).

job quality in the United States and Canada    345

Welfare Policies Government intervention is greater in Canada than the United States in some areas of welfare (Block and Roberts 2000; Breitkreuz 2005; Lipset 1990). A major difference between the two countries is the existence of universal health coverage in Canada, but not the US. This makes employer-​provided health insurance coverage a key aspect of job quality in the US, and its absence a major determinant of whether a job is considered ‘bad’ (Hunt-​McCool et al. 2000). While the Affordable Care Act in the US and its subsequent revisions have helped somewhat to alleviate this concern, health insurance in the US is still not universal. Although there are notable gaps in health benefits between those covered by employers who offer extended health benefits and those who are not (Kennedy and Morgan 2006), convergence in the share of public financing is a notable trend between the two countries (Cacace and Schmid 2008). The United States currently spends a higher share of its GDP on public pensions than does Canada and Americans have somewhat higher levels of net income replacement through mandatory government pensions than do their Canadian counterparts (50.5% vs. 46.4%) (OECD 2022a). In both Canada and the United States, employer-​ provided pensions are an important supplement to government provisions. However, there has been a substantial decline in the number of workers with access to employer-​ provided pensions. In addition, there has been a general shift towards less secure ways of structuring pension entitlements, with defined contribution plans becoming relatively more popular for new employees than the old standard of defined-​benefits (Morissette and Johnson 2005; Wiatrowski 2012).

Demographic Characteristics Employment rates for men of prime working age (25–​54) are broadly similar in the United States and Canada. While Canadian rates dropped below those in the United States through the 1980s and 1990s, they converged in the 2000s, and dropped less after the economic crisis in 2008, resulting in higher employment rates for men in Canada. Conventional labour force estimates do not account for individuals in prison and jail. Since the prison population rate for men is much larger in the United States, this likely results in the overestimation of their employment rates in the US (Western and Pettit 2010). The contrast is more pronounced for women. American women had higher employment rates than their Canadian counterparts in 1976, but the situation reversed by the early 2000s. This marks the beginning of a period of marked divergence in favour of Canadian women. The gender gap in employment is now considerably smaller in Canada than in the United States, reflecting higher rates of labour force participation for

346    arne l. kalleberg, sylvia fuller and ashley pullman 95 90 85 80 75

19 95 19 9 19 6 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 20 19 20 20

70

CAN Men

CAN Women

US Men

US Women

Figure 15.2  Employment rate in the US and Canada, total working-​age population 25–​54 Source: OECD. 2020a. Employment/​population ratio: Labour force status by age and sex (indicators) (Paris: OECD Statistics). https://​stats.oecd.org/​Index.aspx?DataSetCode=​lfs_​sexage_​i_​r

mothers in Canada than in the United States (OECD 2022b). Stronger entitlements to job protection and paid family leave strengthen women’s labour force attachment (Blau and Kahn 2013), and likely contribute to the growing Canadian advantage with respect to gender equality in employment rates (Drolet et al. 2016). Women’s labour market opportunities are also enhanced by a larger public sector, which hosts a substantial share of female-​dominated occupations in health, social services, and education. Although a larger share of part-​time workers seek full-​time employment in Canada, as will be discussed further on, opportunities to work part time tend to increase women’s employment rates. In Canada, women workers are considerably more likely to work part time than in the United States (47.7% of the female labour force works part time in the former but only 31.3% in the latter) (See Table 15.2). Both Canada and the United States are ‘settler’ societies, with immigration playing an important role in the historical and contemporary constitution of the labour force. However, contemporary migration flows are quite different. Canada formally admits more than twice as many immigrants per capita than the United States (Kaushal and Lu 2015), while flows of undocumented migrants are much larger in the US, Canadian immigration policy is also more explicitly oriented to selecting immigrants based on their likelihood of contributing to the Canadian labour market. Immigration levels for skilled workers and their immediate families are substantially higher than for those admitted due to humanitarian concerns or family reunification, a pattern that has become more pronounced over time (Immigration, Refugees, and Citizenship Canada 2020). This has resulted in a different make-​up of the Canadian and American immigrant populations. While the immigrant population in the United States is about as likely to have a university degree as the native-​born population (31.3% vs. 32.3% among those 25 and over), in Canada immigrants are more likely to have a university

job quality in the United States and Canada    347 Table 15.2 Employment characteristics United States

Canada

Percentage of female part-​time employment, 20211

31.3

47.7

Percentage of male part-​time employment, 20211

19.9

31.9

4.4

23.3

47.0

47.3

Percentage of male underemployment, 2021

1.1

2.2

Percentage of female underemployment, 20211

0.8

1.7

Share of involuntary part-​timers as % of part-​time employment, 20202 Female share of employment, 20211 1

3

Percentage of temporary employees, 2020

3.3

8.4

Incorporated self-​employment rate, 20194

3.6

6.9

Unincorporated self-​employment rate, 20194

5.8

7.9

Long-​term unemployment rate as percentage of unemployed, 20205

5.6

5.1

1 ILO. 2022. ILOSTAT database (Genève: International Labour Organization). 2 OECD. 2020b. Involuntary part-​time work (indicators) (Paris: OECD Statistics). https://​stats.oecd.org/​

index.aspx?DataSetCode=​INVPT_​D# 3 OECD. 2022c. Temporary employment (indicator) (Paris: OECD Statistics). https://​doi.org/​10.1787/​

75589b8a-​en 4 Percentages generated by authors using the US Current Population Survey and Canadian Labour

Force Survey (microdata). Based on workers aged 18–​64. 5 OECD. 2022g. Long-​term unemployment rate (indicator) (Paris: OECD Statistics). https://​doi.org/​

10.1787/​76471ad5-​en

education (33.6% vs. 21.8% among those 25 and over). Conversely, in Canada those with less than a high school education make up a similar share of the immigrant and the native-​born population (16.2% vs. 15.3% of those 25 and over), whereas in the United States immigrants are much more likely to fall into this group (27.7% vs. 8.7% of those 25 and over).2

2  Authors’ calculations from the American Community Survey (2019, Table S0501) and the 2016 Canadian Census Public Use Microdata File. Although both surveys use comparable measures of educational credential level (based on the International Standard Classification of Education), there are differences in national sampling strategy that may impact population coverage. The American Community Survey replaced long-​form census data, though it remained mandatory. Furthermore, the American survey provides one-​and five-​year estimates, while the Canadian population is only sampled every five years.

348    arne l. kalleberg, sylvia fuller and ashley pullman

Job Quality in the US and Canada: Key Issues Our discussion of patterns of job quality in the US and Canada will focus on several key indicators: nonstandard work arrangements (such as temporary and involuntary part-​time work), wages and earnings quality, labour market insecurity, hours and scheduling and overqualification. Table 15.2 presents information on various employment characteristics for the US and Canada.

Nonstandard Work Arrangements There is a much higher use of temporary workers in Canada than in the US. This is due in large part to the bigger public sector in Canada, which uses a higher share of temporary workers than the private sector (Stecy-​Hildebrandt and Fuller 2019) and, to a lesser degree, the seasonality of demand for workers produced by the Canadian resource industries. Temporary and part-​time work in Canada is especially high among youth. In 2020, 30.2 per cent of 15-​to 24-​year-​olds in Canada had temporary jobs, versus 8.2 per cent in the US; this compares to 8.4 per cent and 3.3 per cent of the prime-​age labour forces in Canada and the United States, respectively (OECD 2022c). The relatively low incidence of temporary work in the United States reflects the very low employment protections associated with permanent workers, which gives employers considerable flexibility in hiring and firing workers. Part-​time work is more common in Canada than in the United States for both women and men, and a higher share of those working part time do so involuntarily in Canada (23.3% of Canadian part-​timers are involuntary versus 4.4% in the US). In both countries, the share of part-​time workers employed involuntarily is tied to the state of the economy, with declines for most prime-​aged workers over the prosperous early 2000s and an uptick in the wake of the Great Recession of 2008–​2009 (see Figures 15.3a and 15.3b). In both Canada and the United States, men are more likely to work part-​time involuntarily than women. Older workers are also more likely to work part-time involuntarily in both countries (with the exception of American women). In Canada, recent immigrants are also increasingly overrepresented in involuntary part-​time employment (Hira-​Frisen 2014). Previous research in Canada has also found that, compared to full-​ time workers, involuntary part-​time workers experience greater wage inequality than voluntary part-​time workers (Barrett and Doiron 2001). Canadians are more likely to be self-​employed than are Americans (14.8 percent vs. 9.4 percent), with the gap larger among those with incorporated businesses. In both countries, part-​time and temporary employment is associated with lower wages and fewer benefits (Bardasi and Gornick 2008; Galarneau 2010, 2005; Jahn and Pozzoli 2013). Self-​employment also comes with more limited access to benefits, regulatory protections and social entitlements (such as employment insurance) (Vosko 2010).

Job Quality in the United States and Canada    349 (a) 60% 50% 40% 30% 20% 10%

19 98 19 90 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 20 19 20 20

0%

CAN: ages 25–54

CAN: ages 15–24

US: ages 25–54

US: ages 15–24

Figure 15.3a  Share of involuntary part time among part-​time workers in the US and Canada, men 1998–​2020 Source: OECD. 2020b. Involuntary part-​time work (indicators) (Paris: OECD Statistics). https://​stats.oecd.org/​index.aspx?DataSetCode=​INVPT_​D#

(b) 40% 35% 30% 25% 20% 15% 10% 5%

CAN: ages 15-24

US: ages 25-54

20

19

20

18

20

17

20

16

20

15

20

14

20

13

20

12

20

11

20

10

20

09

20

08

20

07

20

06

20

05

20

04

20

03

CAN: ages 25-54

20

02

20

01

20

00

20

90

20

19

19

98

0%

US: ages 15-24

Figure 15.3b  Share of involuntary part time among part-​time workers in the US and Canada, women 1998–​2020 Source: OECD. 2020b. Involuntary part-​time work (indicators) (Paris: OECD Statistics). https://​stats.oecd.org/​index.aspx?DataSetCode=​INVPT_​D#

350    arne l. kalleberg, sylvia fuller and ashley pullman Temporary workers are also less likely to be unionized (Fuller and Vosko 2008), though temporary workers may have a higher social wage in Canada than in the US, due to their receipt of basic health benefits. Canada has also increased its use of temporary migrant labour in recent years (O’Donnell and Skuterud 2021). Canadian temporary workers are diverse, entering the country under an array of programs (that have changed substantially over time), and in many different types of occupations. While temporary migrant workers officially enjoy many of the same employment rights and entitlements as Canadian citizens, low-​skilled temporary workers’ ability to exercise these rights in practice is limited. The Federal agency responsible for the temporary foreign worker programme has no regulatory authority to monitor compliance with the programme’s requirements (Fudge and MacPhail 2009). Dissent is muted by limited enforcement, lack of knowledge of rights, fear of being sent home prematurely and the hope of securing a nomination for permanent residence (Polanco 2016).

Earnings Quality The average level of earnings in 2013 was a bit lower in Canada than the United States, as Table 15.3 shows. However, the distribution of earnings is markedly more unequal in the United States. Indeed, while earnings inequality has been rising in recent decades in both the United States and Canada, the increase has been greater in the former (Saez and Veall 2005). Low-​wage workers fare better in Canada relative to the average worker, with the ratio of the minimum to median wage of full-​time workers .49 in Canada and .29 in the United States in 2020. In absolute terms as well, the minimum wage is now higher in Canada than in the United States, a situation that has prevailed since 2011. This reverses a long-​standing historical pattern of higher real minimum wages in the United States (see Figure 15.4). Higher earnings inequality in the United States is also a product of more pronounced income gains for high earners in that country. Between 1982 and 2010 income growth was higher in the United States than in Canada for the top 10 per cent of earners—​it was more than three times higher for the top 0.01 per cent of earners (Lemieux and Riddell 2015). Median CEO pay was also one-​third higher in the United States than in Canada in 2006, and mean CEO pay was 44 per cent higher (Fernandes et al. 2012). The OECD earnings quality measure—​which accounts for both the level and inequality of earnings—​is lower for the United States than for Canada in 2016, while it was equal in 2007. Moreover, the consequences of unequal market incomes for overall income inequality are greater in the United States than in Canada due to more redistributive policies in the latter (Card and Freeman 1993; Levine 1996). Nevertheless, there is evidence that government income transfers are increasingly less powerful in mitigating inequality in Canada (Frenette et al. 2007). Although government transfers partially eliminated income inequality during the 1980s, they had less of an effect during the 1990s, resulting in the growth of inequality in after-​transfer and tax income.

job quality in the United States and Canada    351 Table 15.3 Earnings quality and labour market insecurity in Canada and the United States United States

Canada

2007

Latest available year

2007

Latest available year

Earning quality (USD)1

17.47

18.46 (2016)

17.47

20.09 (2016)

Gross hourly wage (USD-​PPPs) 1

26.17

26.36 (2013)

23.60

26.15 (2013)

Earnings inequality index

0.33

0.35 (2013)

0.26

0.26 (2013)

Labour market insecurity index1

3.39

4.25 (2016)

3.25

3.80 (2016)

Unemployment risk index

4.90

5.49 (2015)

6.45

7.51 (2015)

Unemployment insurance index1

30.83

25.57 (2015)

49.59

48.39 (2015)

Gross average annual wages (USD, PPP)

60,188

69,392 (2020)

48,606

55,342 (2020)

Minimum relative to median wage, full-​time workers3

0.31

0.29 (2020)

0.41

0.49 (2020)

1

1

2

1

OECD. 2016. OECD job quality database (Paris: OECD Statistics). https://​stats.oecd.org/​viewhtml. aspx?datasetcode=​JOBQ&l&ang=​en# 2

OECD. 2022h. Average annual wages (indicators) (Paris: OECD Statistics). https://​stats.oecd.org/​Index. aspx?DataSetCode=​AV_​AN_​WAGE# 3

OECD. 2022i. Minimum relative to average wages of full-​time workers (indicators) (Paris: OECD Statistics). https://​stats.oecd.org/​Index.aspx?DataSetCode=​MIN2AVE

$12.00

$10.00

$8.00

$6.00

19 6 19 5 6 19 7 6 19 9 7 19 1 73 19 7 19 5 7 19 7 7 19 9 8 19 1 8 19 3 8 19 5 87 19 8 19 9 9 19 1 93 19 9 19 5 9 19 7 9 20 9 01 20 03 20 0 20 5 07 20 09 20 1 20 1 13 20 1 20 5 17 20 19

$4.00

Canada

United States

Figure 15.4  Real hourly minimum wage, 1965–​2020 Source: OECD. 2020c. Real minimum wages (indicators) (Paris: OECD Statistics). https://​stats.oecd.org/​index.aspx?DataSetCode=​RMW Description: Real hourly wages are statutory minimum wages converted into common hourly wages among OECD countries. Estimates are deflated by national Consumer Price Indices, with 2020 as the base year, and converted into a common currency (i.e. US $ Purchasing Power Parities)

352    arne l. kalleberg, sylvia fuller and ashley pullman Transfers increased after 2009, but remained below early 1990s levels in 2017 (Mehdi and Murphy 2021).

Job Polarization The extent to which the economy is differentiated into good, high-​paying, and bad, low-​paying jobs, is an important aspect of the job quality story in industrial countries. Figures 15.5a and 15.5b show the trends in the percentage of the labour force (employees aged 18–​64) in the two countries that are employed in high, medium and low-​wage occupations. We computed these figures by ordering occupations (approximately 500 unique categories for both Canada and the United States) in 2003 into the top, middle and bottom third of earnings (weighted by employment). We then graphed the percentage growth in employment in each of these three groups from 2003 to 2015.3 We see that there is a strong pattern of job polarization in the US: both high-​wage and low-​wage occupations increased during this period, while mid-​wage occupations experienced sharp declines, at least until 2011. By contrast, in Canada, there was less of a job polarization pattern: there was an increase in higher-​wage occupations relative to low-​wage jobs as the proportion of low-​wage occupations remained steady; and there was only a very slight decline in mid-​wage occupations4 (see also the discussion in Krahn et al. 2011: 153–​159). These differences in job polarization reflect the institutional differences discussed in the previous section. Greater union density in Canada has helped to hold down the growth in low-​wage jobs. The expansion of resource industries in Canada has facilitated the creation of middle-​wage jobs (even though these may not always be highly skilled). The emphasis on immigration of skilled workers in Canada has contributed to the lack of growth of low-​wage jobs. In the United States, by contrast, the presence of large numbers of low-​skilled workers from Mexico and other Latin American countries has produced a hyper-​vulnerable group that is ripe for exploitation in low-​skilled jobs.

Labour Market Insecurity Labour market insecurity can be thought of as having two components: the risk of unemployment; and the degree to which the consequences of unemployment are mitigated by social policy. Table 15.3 reveals that unemployment risk is higher in Canada. However,

3  Given occupation-​wage differences, quantile cut-​points differ slightly by country: mid-​to-​high cut point is $16.83 (USD) in the United States and $18.68 (CDN) in Canada, and mid-​to-​low cut point is $11.00 (USD) in the United States and $13.73 (CDN) in Canada (in 2003 dollars). 4  Although trends of occupation and wage polarization between the 1980s and 1990s were similar in Canada and the US, by the 2000s growth in low-​wage occupations diminished in Canada (Green and Sand 2015).

Job Quality in the United States and Canada    353 15%

Higher-wage occupations

10%

5%

Lower-wage occupations 0%

-5%

Mid-wage occupations

-10%

15 20

14 20

13 20

20

12

11 20

10 20

09 20

08 20

07 20

20

06

05 20

20

20

03

04

-15%

Figure 15.5a  US occupational percentage shares, indexed to 2003 Source: Weighted percentages generated by authors using the US Current Population Survey and Canadian Labour Force Survey microdata.

15%

Higher-wage occupations

10% 5%

Lower-wage occupations

0% -5%

Mid-wage occupations

-10%

15 20

14 20

13 20

12 20

11 20

10 20

09 20

08 20

07 20

06 20

05 20

04 20

20

03

-15%

Figure 15.5b  Canadian occupational percentage shares, indexed to 2003 Source: Weighted percentages generated by authors using the US Current Population Survey and Canadian Labour Force Survey microdata.

354    arne l. kalleberg, sylvia fuller and ashley pullman the degree of active labour market policies is greater in Canada (see Table 15.1). Hence, the increase in labour market insecurity after the Great Recession was somewhat greater in the United States than in Canada.

Work Hours and Work Scheduling Another important aspect of insecurity and job quality relates to the scheduling of work. Workers with nonstandard schedules perform the majority of their work hours outside the traditional Monday to Friday 9–​5 workdays, such as on weekends as well as evening and night shifts. Nonstandard schedules are relatively common in both the United States and Canada. Twenty per cent of the American labour force worked a nonstandard schedule in 2011 (a percentage that has remained stable since 2004—​Enchautegui 2013), while about 33 per cent of workers in Canada worked a schedule outside of a regular daytime shift.5 A problem for workers of nonstandard schedules is that they are ‘un-​social’ (Henly and Lambert 2014; Pocock and Clark 2005; Presser 2005; Strazdins et al. 2006), since the time available for leisure and care responsibilities often fails to overlap with the schedules of friends and family members, as well as with other institutions. This mismatch can be particularly problematic for parents insofar as school and childcare schedules are rigid.6 Because regulated childcare facilities typically offer services during the regular workday, parents who work nonstandard schedules must rely more heavily on informal care arrangements (Enchautegui et al. 2015; Strazdins et al.2006). The quality of such arrangements is by nature highly variable, with centre-​based care generally leading to better developmental outcomes for children, such as school readiness (Hansen and Hawkes 2009). Informal care arrangements may also be less reliable, which can jeopardize parents’ employment. Yet over one-​fifth of parents with children under 13 work non-​standard schedules in the US, a figure that rises to 30 per cent for low-​income parents (Enchautegui et al. 2015). In Canada, 27% of parents with children under six years old worked a non-​standard schedule in 2017, with those with lower education and job skills more likely to be working non-​standard schedules (Lero et al., 2021). In both the United States and Canada, studies have found that working nonstandard hours adversely affects both workers’ health and children’s well-​being (Chung et al. 2009; de Castro et al. 2010; Presser 2005; Strazdins et al. 2006). Not surprisingly, most of those working nonstandard schedules do so because it is a requirement of the job, not a personal preference (Strazdins et al. 2006).

5  That is, 6.4 per cent were employed in a regular evening or night schedule, and 26.5 percent were employed in a rotating, split, on call, or irregular schedule. (Source: authors’ calculations from 2011 Survey of Labour and Income Dynamics.) 6  In addition to nonstandard schedules, working long hours is likely to create problems balancing work and family needs. According to the OECD’s Better Life Index, 3.7 per cent of Canadian employees work more than 50 hours a week, compared to 11.1 per cent in the United States (OECD 2022d).

job quality in the United States and Canada    355 Nonstandard schedules are also frequently irregular, where the timing and number of hours worked per day or week can vary substantially. For low-​income workers in particular this can be extremely challenging, making it hard to anticipate whether earnings will be sufficient from month-​to-​month. Unpredictability also makes it difficult to arrange childcare, participate in family rituals, pursue education, or coordinate plans with others more generally (Alexander and Haley‐Lock 2015; Henly and Lambert 2014). Measuring the regularity of work hours is challenging insofar as national surveys typically ask workers to estimate their ‘usual’ work hours, which encourages them to report an average regardless of variability. From 1994–​2013, 4–​5 per cent of American workers reported that their hours varied instead of giving an average,7 with the share higher in service industries (Alexander and Haley‐Lock 2015). There is also evidence of increasing schedule variability among American service workers, with 25 per cent reporting that their hours varied in 2013, up from 13 per cent in 1994 (Alexander and Haley‐Lock 2015). Seventy-​four per cent of both hourly and non-​hourly American workers between 26 and 32 years old in 2011–​2012 reported some fluctuations in their hours in the previous month, with average variation of more than a typical workday for both groups (Lambert et al. 2014). Although there is difficulty tracking trends in Canada, in 2020 23% of Canadian workers reported working irregular schedules, with 62% of those with irregular schedules reporting both their schedule and number of hours changed (Statistics Canada, 2020). Variability in work hours is often accompanied by limited advance notice. Among American young workers between 26 and 32, 38 per cent knew their schedule one week or less in advance in 2011–​2012, with lack of advance notice particularly common for those working part-​time (52%) (Lambert et al. 2014). In Canada, schedules were more predictable, with 11.4 per cent of the same age group experiencing similarly short notice in 2005, although this rises to 31 per cent for part-​time workers.8 While it is difficult to generate strictly comparable statistics, it is clear that irregular and non-​standard schedules are quite common in both the United States and Canada. However, some particular patterns of unpredictability and their consequences are muted in Canada by regulatory provisions for ‘show up’ pay. In retail and hospitality jobs in particular, ‘just-​in-​time’ scheduling is increasingly common in the United States. When demand is slow, employees may be sent home before the end of their shift, reducing employers’ costs, but creating considerable insecurity for workers. Canadian workers are more protected from just-​in-​time scheduling than their American counterparts insofar as they are more likely to be covered by employment standards that require employers to pay workers a guaranteed minimum number of hours for employees who show up to their shifts. Such ‘show-​up pay’ or ‘reporting pay provisions’ are present in nine states, but across all Canadian provinces and territories. Although the aim of this policy is

7  This likely gives a conservative measure of variability as employees are still encouraged to give an average if possible (Lambert et al. 2014) 8  Authors’ calculations from 2005 Statistics Canada Workplace and Employee Survey.

356    arne l. kalleberg, sylvia fuller and ashley pullman to disincentivize employees being sent home without pay, Alexander and Haley-​Lock (2015) illustrate that legally mandated minimum hours are difficult to enforce and create incentive for employers to schedule short shifts. Nevertheless, in her comparative study of employees in the same chain restaurants in British Columbia, Chicago and New York, Haley-​Lock (2011) found that employees were much more likely to be sent home in slow periods in the American States where reporting pay requirements did not exist.

Over-​Qualification and Underemployment There is a continuing discussion in both the US (e.g. Kalleberg 2007: Chapter 3) and Canada (e.g. Livingstone 1998; Uppal and LaRochelle-​Côté 2014) about the degree to which workers are over-​qualified (i.e. have levels of education that are greater than that required to perform their work) for their jobs. Workers in both countries are relatively highly educated and the growth of low-​wage, often low-​skilled jobs suggests that many workers will be in jobs that do not make use of their skills. The 2004 Canadian Learning and Work Survey, for example, found that 34 per cent of respondents felt underemployed and the incidence among new immigrants was particularly high (Krahn et al. 2011, 52–​ 53). Likewise, over-​qualification is growing in the US among all groups, with immigrant and non-​white workers reporting higher levels (Vaisey 2006; Zong and Batalova 2016). While underemployment is a problem in both countries, it is more prevalent in Canada. In 2021, OECD estimates of underemployment were almost twice as high for Canadian as American among both men and women (see Table 15.2). The prevalence of over-​qualification in both countries highlights the reality that education may not be enough to secure quality work, particularly for less-​advantaged groups. Nonetheless, it remains generally true that educational credentials tend to increase one’s chances of securing a better job. In Canada, post-​secondary degree holders not only receive a wage premium, which has grown since the 1980s, but also work more hours and weeks annually (Boudarbat et al. 2010). During this same period, workers with a high school diploma or less faced declining wages (Green and Townsend 2010). The wages of longer-​tenured workers were less affected, illustrating the efficacy of employment protection (Green and Townsend 2010). Increasing returns to education are also a notable trend in the US (Hout 2012). The necessity of advanced education in determining access to high-​quality employment is a cause for concern given the high level of stratification within Canadian and US education systems.

Conclusions and Implications Comparing the United States and Canada ‘controls’ for large differences in labour market and welfare institutions. In this sense, it provides a picture of what might be possible with respect to job quality in countries with relatively low levels of employment

job quality in the United States and Canada    357 protection and flexible labour markets in the absence of more politically challenging large-​scale institutional changes. It is easier to look to one’s immediate neighbours for policy inspiration than to countries with much deeper institutional and cultural differences (such as the social-​democratic Nordic countries). It is notable, in this respect, that despite the similarities between the United States and Canada, there are a number of important differences in job quality. The incidence of temporary and part-​time work is higher in Canada, for example, owing mainly to the larger size of the Canadian public sector and Canada’s greater dependence on resource industries. Canadian part-​time workers are also more likely to prefer full-​time employment than their American counterparts. Average earnings are currently similar in the two countries, but the degree of earnings inequality is greater in the United States, which reflects the greater power of unions in Canada as well as the smaller decline in middle-​ and greater stability in low-​wage occupations in Canada. In addition, labour market insecurity is greater in the United States due mainly to the more generous unemployment insurance provided in Canada. Workers in both countries suffer from nonstandard schedules, but Canadian labour laws offer greater protections to workers than in the US. These indicators suggest that although a greater proportion of Canadian workers are in nonstandard employment arrangements, there is greater earning equality, security and social protection. Despite these differences, both countries are facing many similar challenges regarding job quality. In both countries, for example, the rise in precarious work has made jobs more insecure, pointing to the need for wage insurance, more generous unemployment insurance assistance, and more attention to active labour market policies. There are also concerns in both countries with work–​ family balance due to nonstandard work schedules, high levels of involuntary part-​time work, and problematic gaps in earnings and other benefits between temporary and part-​time workers and regular, full-​time workers. In addition, although education may be a determining factor in who obtains high-​quality employment, there are significant proportions of overqualified workers in both countries, especially among recent immigrants. Persistent overqualification among college-​educated workers in both countries can lead workers to question the value of education and result in an underutilization of workers’ skills and potentials. Comparing job quality in these two similar countries provides us with a better sense of how labour market and welfare institutions generate job quality. We have shown how relatively small differences in policies such as health coverage, maternity/​parental leave provisions, and the regulation of work hours help to shape the quality of jobs. Differences in the power of unions clearly matter, especially when it comes to job polarization and the relatively more favourable positioning of low-​wage workers in Canada. Country differences in the mix of public and private sectors and in the dependence on resource as opposed to manufacturing industries also translate into variations in job quality. Although a low-​wage workforce is present in both Canada and the United States, welfare policies, immigration systems, unions, and other institutional factors influence the availability of quality employment.

358    arne l. kalleberg, sylvia fuller and ashley pullman Our discussion of job quality in the United States and Canada has also highlighted several timely research questions. We need to know more about the extent to which differences in labour market and welfare institutions in these two countries mediate and moderate the impact of working in nonstandard jobs on perceptions of job and economic insecurity and other individual outcomes (e.g. stress, satisfaction, health). We also need more systematic analyses of how job polarization has affected intragenerational mobility and careers in the two countries. And, we need a clearer understanding of the degree to which population sub-​groups (defined by gender, race, age, immigration status, education) are confined to low-​wage, bad jobs. These issues are likely to take on added urgency in the years ahead.

Acknowledgements We thank Rachel Dwyer and Elaine McCrate for helpful comments on an earlier version of this chapter. Extra thanks to Manlin Cai for excellent research assistance providing necessary data updates at time of publication (chapter was originally written in 2017).

References Alexander, Charlotte, and Haley-​Lock, Anna (2015), ‘Underwork, Work-​Hour Insecurity Anna (2015), ‘Underwork, Work-​Hour Insecurity, and a New Approach to Wage and Hour Regulation’, Industrial Relations: A Journal of Economy and Society, 54/​4, 695–​7 16. Bardasi, Elena, and Gornick, Janet C. (2008), ‘Working for Less? Women’s Part-​Time Wage Penalties across Countries’, Feminist Economics, 14/​1, 37–​72. Barrett, Garry F., and Doiron, Denise J. (2001), ‘Working Part Time: By Choice or By Constraint’, Canadian Journal of Economics, 34/​4, 1042–​1065. Beaujot, Roderic, Du, Ching J., and Ravanera, Zenaida (2013), ‘Family Policies in Quebec and the Rest of Canada: Implications for Fertility, Child-​Care, Women’s Paid Work, and Child Development Indicators’, Canadian Public Policy, 39/​2, 221–​240. Blau, Francine D., and Kahn, Lawrence M. (2013), ‘Female Labor Supply: Why is the United States Falling Behind?’ The American Economic Review 103/​3, 251–​256. Block, Richard N., and Roberts, Karen (2000), ‘A Comparison of Labour Standards in the United States and Canada’, Industrial Relations, 55/​2, 273–​307. Boudarbat, Brahim, Lemieux, Thomas, and Riddell, Craig W. (2010), ‘The Evolution of the Returns to Human Capital in Canada, 1980-​2005’, Canadian Public Policy, 36/​1, 63–​89. Breitkreuz, Rhonda S. (2005), ‘Engendering Citizenship: A Critical Feminist Analysis of Canadian Welfare-​to-​Work Policies and the Employment Experiences of Lone Mothers’, Journal of Sociology & Social Welfare, 32/​2, 147–​165. Burchell, Brendan, Sehnbruch, Kristen, Piasna, Agnieszka, and Agloni, Nurjk (2013), ‘The Quality of Employment and Decent Work: Definitions, Methodologies, and Ongoing Debates’, Cambridge Journal of Economics, 38/​2, 459–​477. doi:10.1093/​cje/​bet067.

job quality in the United States and Canada    359 de Bustillo, Rafael Munoz, Fernandez-​Macías, Enrique, Esteve, Fernando, and Antón, Jose-​ Ignacio (2011), ‘E Pluribus Unum? A Critical Review of Job Quality Indicators’, Socio-​ Economic Review, 9/​3, 447–​475. Cacace, Mirella, and Schmid, Achim (2008), ‘The Healthcare Systems of the USA and Canada: Forever on Divergent Paths?’ Social Policy & Administration 42/​4, 396–​417. Card, David, and Freeman, Richard B. (1993), Small Differences That Matter: Labor Markets and Income Maintenance in Canada and The United States (Chicago, Illinois: University of Chicago Press). Chung, Sharon A., Wolf, Theresa K., and Shapiro, Colin M. (2009), ‘Sleep and Health Consequences of Shift Work in Women’, Journal of Women’s Health, 18/​7, 965–​977. de Castro, A. Butch, Fujishiro, Kaori, Rue, Tessa, Tagalog, Eularito A., Samaco-​Paquiz, L. P., and Gee, Gilbert C. (2010), ‘Associations Between Work Schedule Characteristics and Occupational Injury and Illness’, International Nursing Review, 57/​2, 188–​194. Drolet, Marie, Uppal, Sharanjit, and LaRochelle-​ Côté, Sebastien (2016), ‘The Canada-​ U.S. Gap in Women’s Labour Market Participation’, Insight on Canadian Society: 1–​13 (Ottawa: Statistics Canada), http://​www.stat​can.gc.ca/​pub/​75-​006-​x/​2016​001/​arti​cle/​14651-​ eng.htm. Doucet, Andrea, Lero, Donna, McKay, Lindsey and Tremblay, Diane-​Gabrielle (2021), ‘Canada Country Note’, in Koslowski, Alison, Blum, Sonja, Dobrotic, Ivana, Kaufman, Gayle, and Moss, Peter, eds., International Review of Leave Policies and Research. Available at: https://​ www.leavenetwork.org/​annual-​review-​reports/​ Enchautegui, María E. (2013), ‘Nonstandard Work Schedules and the Well-​Being of Low-​ Income Families’, Paper 26. Washington, DC: Urban Institute. Enchautegui, María E., Johnson, Martha, and Gelatt, Julia (2015), Who Minds the Kids When Mom Works a Nonstandard Schedule? (Washington, DC: Urban Institute). European Commission (2001), Third European Survey on Working Conditions 2000 (Luxembourg: Office for Official Publications of the European Communities). Fernandes, Nuno, Ferreira, Miguel A., Matos, Pedro, and Murphy, Kevin J. (2012), ‘Are US CEOs Paid More? New International Evidence’, Review of Financial Studies, 26/​2, 323–​367. Findlay, Patricia, Kalleberg, Arne L., and Warhurst, Chris (2013), ‘The Challenge of Job Quality’, Human Relations, 66/​4, 441–​451. Fortin, Nicole M., and Lemieux, Thomas (2015), ‘Changes in Wage Inequality in Canada: An Interprovincial Perspective’, Canadian Journal of Economics, 48/​2, 682–​7 13. Frangi, Lorenzo, Hennebert, Marc-​Antonin, and Memoli, Vincenzo (2014), ‘Social Confidence in Unions: A US‐Canada Comparison’, Canadian Review of Sociology, 51/​2, 170–​188. Frenette, Marc, Green, David A., and Milligan, Kevin (2007), ‘The Tale of The Tails: Canadian Income Inequality in the 1980s and 1990s’, Canadian Journal of Economics, 40/​3, 734–​764. Fudge, Judy, and MacPhail, Fiona (2009), ‘The Temporary Foreign Worker Program in Canada: Low-​Skilled Workers as an Extreme Form of Flexible Labor’, Comparative Labor Law and Policy Journal, 31, 101–​139. Fuller, Sylvia (2005), ‘Public Sector Employment and Gender Wage Inequalities In British Columbia: Assessing The Effects Of A Shrinking Public Sector’, Canadian Journal of Sociology, 30/​4, 405–​439. Fuller, Sylvia, and Vosko, Leah F. (2008), ‘Temporary Employment and Social Inequality in Canada: Exploring Intersections of Gender, Race and Immigration Status’, Social Indicators Research, 88/​1, 31–​50.

360    arne l. kalleberg, sylvia fuller and ashley pullman Galarneau, Diane (2005), ‘Earnings of Temporary Versus Permanent Employees’, Perspectives on Labour and Income, 6/​1, 5–​18. Galarneau, Diane (2010), ‘Temporary Employment in the Downturn’, Perspectives on Labour and Income, 22/​4, 48–​60. Gellatly, Guy (2016), ‘Recent Development in the Canadian Economy’, Statistics Canada, Economic Insights 58 (Spring), 1–​9. Retrieved from http://​www.statcan.gc.ca/​pub/​11-​626-​x/​ 11-​626-​x2016058-​eng.htm Ghai, Dharam (2003), ‘Decent Work: Concept and Indicators’, International Labour Review, 142/​2), 113–​145. Green, David A., and Sand, Benjamin M. (2015), ‘Has the Canadian Labour Market Polarized?’ Canadian Journal of Economics, 48/​2, 612–​646. Green, David A., and Townsend, James (2010), ‘Understanding the Wage Patterns of Canadian Less Skilled Workers: The Role of Implicit Contracts’, Canadian Journal of Economics, 43/​1, 373–​403. Green, Francis (2006), Demanding Work: The Paradox of Job Quality in the Affluent Economy (Princeton: Princeton University Press). Haley-​ Lock, Anna (2011), ‘Place-​ Bound Jobs at the Intersection of Policy and Management: Comparing Employer Practices in US and Canadian Chain Restaurants’, American Behavioral Scientist, 55, 823–​842. Hansen, Kirstine, and Hawkes, Denise (2009), ‘Early Childcare and Child Development’, Journal of Social Policy, 38/​2, 211–​239. Henly, Julia R., and Lambert, Susan J. (2014), ‘Unpredictable Work Timing in Retail Jobs Implications for Employee Work–​Life Conflict’, Industrial & Labor Relations Review, 67, 986–​1016. Holman, David (2013), ‘Job Types and Job Quality in Europe’, Human Relations, 66/​4, 475–​502. Hou, Feng, and Coulombe, Simon (2010), ‘Earnings Gaps for Canadian Born Visible Minorities in the Public and Private Sectors’, Canadian Public Policy, 23/​1, 29–​43. Hout, Michael (2012), ‘Social and Economic Returns to College Education in the United States’, Annual Review of Sociology, 38/​1, 379–​400. Hunt-​McCool, Janet, McCool, Thomas Joseph, and Dor, Avi (2000), ‘Employer-​Provided Verses Publicly Provided Health Insurance: Effect on Hours Worked and Compensation’, in William T. Alpert and Stephen A. Woodbury, eds., Employee Benefits and Labor Markets in Canada and the United States (Kalamazoo, MI: W.E. Upjohn Institute for Employment Research), 325–​348. Ilcan, Suzan M., O’Connor, Daniel M., and Oliver, Marcia L. (2003), ‘Contract Governance and the Canadian Public Sector’, Industrial Relations, 58/​4), 620–​643. Ilcan, Suzan, Oliver, Marcia, and O’Connor, Daniel (2007), ‘Spaces of Governance: Gender and Public Sector Restructuring in Canada’, Gender, Place and Culture 14/​1, 75–​92. Immigration, Citizenship, and Refugees Canada (2020), 2020 Annual Report to Parliament on Immigration (Ottawa: Government of Canada). https://​www.canada.ca/​content/​dam/​ircc/​ migration/​ircc/​english/​pdf/​pub/​annual-​report-​2020-​en.pdf ILO (2022). ILOSTAT database (Genève: International Labour Organization). Jahn, Elke J., and Pozzoli, Dario (2013), ‘The Pay Gap of Temporary Agency Workers—​Does the Temp Sector Experience Pay Off?’ Labour Economics, 24, 48–​57. Johnson, David (2002), Thinking Government: Ideas, Policies, Institutions, and Public-​Sector Management in Canada (Toronto: University of Toronto Press). Kalleberg, Arne L. (2007), The Mismatched Worker (New York: W.W. Norton).

job quality in the United States and Canada    361 Kalleberg, Arne L. (2011), Good Jobs, Bad Jobs: The Rise of Polarized and Precarious Employment Systems in The United States, 1970s to 2000s (New York: Russell Sage Foundation). Kaushal, Neeraj, and Lu, Yao (2015), ‘Recent Immigration to Canada and the United States: A Mixed Tale of Relative Selection’, International Migration Review, 49/​2, 479–​522. Kennedy, Jae, and Morgan, Steve (2006), ‘Health Care Access in Three Nations: Canada, Insured America, and Uninsured America’, International Journal of Health Services, 36/​4, 697–​7 17. Krahn, Harvey, Lowe, Karen D., and Hughes, Graham S. (2011), Work, Industry, and Canadian Society. Sixth Edition (Toronto: Nelson Education). Lambert, Susan J., Fugiel, Peter J., and Henly, Julia R. (2014), Research Brief: Precarious Work Schedules among Early-​Career Employees in the US: A National Snapshot (Chicago: University of Chicago, Employment Instability, Family Well-​ Being, and Social Policy Network (EINet)). Lero, Donna, Prentice, Susan, Friendly, Martha, Richardson, Brooke and Fraser, Ley (2021), Non Standard Work and Child Care in Canada: A Challenge for Parents, Policy Makers, and Child Care Provision. Childcare Resource and Research Unit and the University of Guelph. Lemieux, Thomas, and Riddell, W. Craig (2015), Top Incomes in Canada: Evidence from the Census. No. w21347’ (National Bureau of Economic Research). Levine, Marc V. (1996), ‘Public Policies, Social Institutions, and Earnings Inequality: Canada and the United States, 1970–​1995’, American Review of Canadian Studies, 26/​3, 315–​339. Lipset, Seymour Martin (1990), Continental Divide: The Values and Institutions of the United States and Canada (New York: Routledge). Jobs Gap: Underemployment or Economic Livingstone, David W. (1998), The Education-​ Democracy (Boulder: Westview Press). Lowe, Graham S. (2000), The Quality of Work: A People-​Centered Agenda (Ontario: Oxford University Press). Lucifora, Claudio, and Meurs, Dominique (2006), ‘The Public Sector Pay Gap in France, Great Britain and Italy’, Review of Income and Wealth, 52/​1, 43–​59. McKay, Lindsey, Mathieu, Sophie and Doucet, Andrea (2016), ‘Parental-​Leave Rich and Parental-​Leave Poor: Inequality in Canadian Labour Market Based Leave Policies’, Journal of Industrial Relations, 59/​4, 543–​562. Mehdi, Tahsin, and Murphy, Brian (2021), ‘Finances of the Nation: Net Income Tax Rates and the Changing Progressivity of the Cash Tax/​Transfer System’, Canadian Tax Journal/​Revue fiscale canadienne 69/​2, 575–​593. Morissette, Rene, and Johnson, Anick (2005), ‘Are Good Jobs Disappearing in Canada?’ Economic Policy Review, 11/​1, 1–​52. O’Donnell, Ian and Skuterud, Mikal (2021), ‘The Transformation of Canada’s Temporary Foreign Worker Program’, Working Paper Series No.39, University of Waterloo, Canadian Labour Economics Forum (CLEF), Waterloo. OECD (2015), Indicators of Immigrant Integration 2015 (Paris: OECD Statistics). OECD (2022a), Net Pension Replacement Rates (Indicator) (Paris: OECD Statistics). OECD (2022b), Gender Wage Gap (Indicators) (Paris: OECD Statistics). OECD (2022c), Temporary Employment (Indicators) (Paris: OECD Statistics). OECD (2022d), Work-​Life Balance (Paris: OECD Better Life Index). http://​www. oecdbetterlifeindex.org/​topics/​work-​life-​balance Pocock, Barbara, and Clarke, Jane (2005), ‘Time, Money and Job Spillover: How Parents’ Jobs Affect Young People’, Journal of Industrial Relations, 47/​1, 62–​76.

362    arne l. kalleberg, sylvia fuller and ashley pullman Polanco, Geraldina (2016), ‘Consent Behind the Counter: Aspiring Citizens and Labour Control Under Precarious (Im)migration Schemes’, Third World Quarterly, 37/​8, 1332–​1350. Presser, Harriet. B. (2005), Working in a 24/​7 Economy: Challenges for American Families (New York: Russell Sage Foundation). Ray, Rebecca, Gornick, Janet C. and Schmitt, John (2008), Parental Leave Policies in 21 Countries: Assessing Generosity and Gender Equality (Revised June 2009) (Washington, DC: Economic Policy Institute). Saez, Emmanuel, and Veall, Michael R. (2005), ‘The Evolution of High Incomes in Northern America: Lessons from the Canadian Evidence’, American Economic Review, 95/​3, 831–​849. Statistics Canada (2015), Employment Insurance Coverage Survey, 2014 (The Daily catalogue no. 11-​001-​X) (Ottawa: Statistics Canada). http://​www.stat​can.gc.ca/​daily-​quotid​ien/​151​123/​ dq1511​23b-​eng.pdf. Statistics Canada (2017), CANSIM Table 282-​0223, ‘Labour Force Survey Estimates (LFS), employees by union status, sex, age group and education level’, North American Industry Classification System (NAICS), Canada, annual. Statistics Canada (2020), ‘Aspects of Quality of Employment in Canada, February and March 2020)’. The Daily. March 22, 2021. https://​www150.statcan.gc.ca/​n1/​daily-​quotidien/​210322/​ dq210322a-​eng.htm Stecy-​Hildebrandt, Natasha, and Fuller, Sylvia (2019), ‘ ‘Bad’ jobs in a ‘Good’ Sector: Examining the Employment Outcomes of Temporary Work in the Canadian Public Sector’, Work, Employment and Society, 33/​4, 560–​579. Strazdins, Lyndall, Clements, Mark S., Korda, Rosemary J., Broom, Dorothy H., and D’Souza, Rennie M. (2006), ‘Unsociable Work? Nonstandard Work Schedules, Family Relationships, and Children’s Well‐Being’, Journal of Marriage and Family, 68/​2, 394–​410. Tang, J. (2017), ‘Industrial Structure Change and the Widening Canada–​US Labor Productivity Gap in the Post-​2000 Period’, Industrial and Corporate Change, 26/​2, 259–​278. United States Bureau of Labour Statistics (2017), ‘Union Members Summary’, Economic News Release USDL-​17-​0107. Uppal, Sharanjit, and LaRochelle-​Côté, Sebastien (2014), Overqualification Among Recent University Graduates in Canada (pub. no. 75-​006-​X) (Ottawa: Statistics Canada). Vaisey, Stephen (2006), ‘Education and its Discontents: Overqualification in America, 1972-​ 2002’, Social Forces, 85/​2, 835–​864. Vosko, Leah (2010), Managing the Margins: Gender, Citizenship, and the International Regulation of Precarious Employment (Oxford: Oxford University Press). Western, Bruce, and Pettit, Becky (2010), ‘Incarceration & Social Inequality’, Daedalus 139/​ 3, 8–​19. Wiatrowski, William J. (2012), ‘The Last Private Industry Plans: A Visual Essay’, Monthly Labor Review, 135/​12, 3–​18. Wilson, George, Roscigno, Vincent J., and Huffman, Matt (2015), ‘Racial Income Inequality and Public Sector Privatization’, Social Problems 62/​2, 163–​185. Zong, Jie, and Batalova, Jeanne (2016), College-​Educated Immigrants in the United States. (Washington, DC: Migration Policy Institute). http://​www.migr​atio​npol​icy.org/​arti​cle/​coll​ ege-​educa​ted-​imm​igra​nts-​uni​ted-​sta​tes.

chapter 16

the great recession and job quality trends in Europe christine erhel, mathilde guergoat-​larivière, janine leschke and andrew watt

Introduction Job quality has been intensively analysed in recent years in the context of national and international debates and policy developments. In addition to attempts to define and measure job quality, research has focused on developing international comparisons, particularly across the EU and OECD countries, in order to capture comparative trends in job quality. This research reveals important heterogeneity across countries and suggests that there has been an overall upward trend in job quality over the 2000s, though with some polarization across occupations and social groups. However, the period of the Great Recession in 2007–​2008 appears very specific and puts these trends into question. Taking a multidimensional, socioeconomic approach to job quality, this chapter further develops the empirics with a focus on the consequences on job quality of the economic crisis that followed the global financial crash. In comparison to previous analyses, its originality is also methodological: in addition to the aggregate analysis, we develop an individual perspective on job quality dynamics based on panel data. The aim of the chapter is twofold: first, to identify trends in job quality in the EU during the first period of the crisis; second, to explore the links between these trends and cyclical as well as institutional factors. Such analysis is performed at both the aggregate and the individual level to test the hypothesis that these cross-​country differences are of significance in seeking to explain job quality trends. The chapter begins by reviewing the literature on job quality measurement in a multidimensional perspective and summarizes existing research for job quality developments for the recession period. The empirical analysis in the following section is based on a

364    christine erhel et al synthetic job quality index—​the European Job Quality Index (JQI)—​covering six dimensions, and outlines changes in job quality between 2005 and 2010 based on this index. The third main section of this chapter analyses the trends in job quality at the individual level and in longitudinal perspective.

Job Quality in Times of Crisis: A Literature Review Job quality features in both the European Employment Strategy (ESS) from 2011 and the ILO Decent Work Agenda from 1999. The OECD has also developed its own approach to job quality in recent years (OECD 2016). Significantly, all three frameworks converge on the necessity of a multidimensional approach.

The Literature on Job Quality: Towards a Multidimensional Approach The focus on job quality has raised important questions about how it is defined and can be measured, not least to be able to monitor its trends. In labour economics, job quality was traditionally understood as being captured by wage levels; a Eurofound (2008) study, for example, defines job quality on the basis of hourly wages. In sociological or industrial relations studies it was frequently focused primarily on working conditions; Braverman (1974), for example, emphasised worker skills and their control of work. Recent developments in economics and socioeconomic approaches open up additional dimensions. Developments in human capital theory recognize the heterogeneity of jobs and workers, and thus call for a differentiation in assessing job quality according to the skills involved in particular jobs or the skill matching between workers and jobs. At the macro level, market failures can lead to underinvestment in human capital, implying that investment and participation in education and training activities could be regarded as an indicator of employment quality. In the ‘economics of happiness’ (Layard 2005), the approach to job quality is enriched by the consideration of the workers’ assessment of their job satisfaction and well-​being derived from survey data. Such surveys make it possible to determine the ‘subjective’ dimensions of job quality by asking people what is important to them (for further discussion of subjectivity in the measurement of job quality, see Knox and Wright in the chapter titled ‘Understanding Job Quality Using Qualitative Research’ in this volume). Using this approach, some studies consider that job quality can be measured using job satisfaction indicators. An important finding of such studies also is that the absolute wage level is less important than predicted by narrow economistic approaches to job quality, suggesting that an adequate living standard, wage equity and upward wage mobility could also be taken as indicators of

the great recession and job quality trends in Europe    365 employment quality, as Stuart et al. (2016) illustrate in their research on what makes decent work for low-​wage workers. Recent thinking about job quality would also include the impact of employment on other spheres of life (e.g. Pocock and Skinner 2012). Indeed, the possibility of reconciliation of work and family life appears to be a very important dimension of job quality according to workers’ responses to the European Social Survey. This approach is consistent with policy-​oriented approaches that include the ‘transitional labour market’ perspective (Schmid and Gazier 2002) which stresses the importance of out-​of-​work quality dimensions, including the right to training, occupational redeployment or retraining, engagement in family life and adaptation of one’s working hours throughout the life cycle. Current socioeconomic approaches thus build on multidimensional definitions of job quality. Existing frameworks take account of skills, work effort and intensification, worker’s discretion, wages, risk and job insecurity as well as workers’ well-​being (Green 2006; Gallie 2007; Muñoz de Bustillo et al. 2011). In a policy-​oriented perspective, multivariate frameworks have also been developed to complement the quantitative indicators of the European Employment Strategy (EES), including a critical revision of the main dimensions of the Laeken indicators developed in 2001 to measure job quality trends in Europe (Davoine et al. 2008; Leschke and Watt 2008; Leschke and Watt 2014). These revisions include common dimensions: working conditions, wages, working time and reconciliation, employment security, social dialogue and collective representation, and lifelong learning.

Intensity of the Crisis and Job Quality Developments The heterogeneity of national contexts in terms of institutional settings as well as the labour market impact of the crisis calls for a detailed and multidimensional approach to job quality in order to assess the effect of the crisis on the various dimensions of this concept. It is indeed likely that country-​specific structural features, such as the distribution of employment across sectors, labour market policies and labour market regulation, as well as cyclical factors will lead to different outcomes of the crisis on job quality in terms of dimensions such as wages, job security, job satisfaction, working time, working conditions. A key question is whether a systematic relationship exists across countries between quantitative labour market developments—​driven substantially by the depth and duration of the crisis—​and developments in terms of job quality. The crisis and the associated rise in unemployment might be hypothesised to have two contradictory effects on measured job quality. On the one hand, workers come under increased threat of losing their jobs, which reduces their bargaining power, thereby exerting downward pressure on job quality. On the other hand, recession leads to a disproportionate destruction of low-​productivity, flexible and marginal jobs that often combine a number of negative features including short-​term contracts, short working hours and low wages, etc. The disproportionate destruction of poor-​quality jobs would have compositional effects that push up average job quality. Moreover, job losses were

366    christine erhel et al not distributed equally across sectors, leading to impacts on aggregate job quality attributable to sector-​compositional effects. Given these contradictory trends, as well as differences in the intensity of the crisis and the consequences for the labour market in different countries, the dynamics of job quality over this period and the development of intra-​country differences remain difficult to pin down. Relatively few empirical studies on job quality cover the recession period: Hurley and Storrie (2011), Hurley et al. (2013), and Green et al. (2012) provide some results, all using the 2010 European Working Conditions Survey (EWCS) as well as other data sources such as the European Labour Force Survey (LFS). The OECD (2016) developed its own Job Quality Database, also offering some insights on the development of job quality during the crisis. Piasna (2017) provides an update of the ETUI JQI discussed in detail below to 2015. The results show a decline in non-​wage job quality and a deepening polarisation in some job quality dimensions across Europe as well as sluggish real wage growth in the aftermath of the crisis. Hurley et al. (2013) confirm the complexity of the links between economic cycle and job quality. On the basis of wage quintile analysis, the overall pattern of employment trends in the EU is characterized by polarization during the downturn with low and high-​wage jobs on the rise and medium-​wage jobs declining. On the other hand, a multidimensional, non-​pecuniary definition of job quality including four main dimensions provides a different picture, namely an overall upgrading trend within the jobs structure in the EU. Medium and low-​quality jobs are decreasing as a proportion of the total, whereas employment change in the upper quintile of jobs’ distribution is positive. The contribution of the Great Recession to job quality trends is therefore not straightforward. OECD results confirm that the crisis may have had complex effects on job quality: using their three dimensions (earnings quality, labour market security, quality of the working environment), they show that due to the destruction of low-​paid jobs, the first dimension has improved, while labour market security has decreased and variations in the quality of the working environment have been more heterogeneous across countries. In the following sections, we develop an empirical analysis at both aggregate and individual level with a focus on job quality trends across Europe during the Great Recession.

Cross-​C ountry Developments in Job Quality between 2005 and 2010 according to the Job Quality Index This section draws on the synthetic ETUI Job Quality Index (JQI) which is based on six dimensions, drawing on a number of data sources (see Appendix 1). On the basis of this index, we discuss the changes in job quality between 2005 and 2010, not least in the

the great recession and job quality trends in Europe    367 light of the economic crisis that swept through Europe from 2008 (and as early as 2007 in some countries). It is important to emphasize, however, that the results on the changes in job quality in its various dimensions cannot be interpreted as simply reflecting the impact of the crisis. They also reflect developments before the crisis (2005–​2008) which differ between countries and that were, in a number of cases, characterized by an economic boom and sharply tightening labour markets. Moreover, the findings also reflect longer-​term secular trends such as tertiarization.1 The synthetic job quality index includes six dimensions (for details on the construction of the index and its components, see Appendix 15A1). These dimensions are: • Wages (nominal compensation per employee in PPS, in-​work poverty) • Non-​standard forms of employment (involuntary temporary employment, involuntary part-​time employment); this sub-​index is inverted, that is a higher level of involuntary non-​standard jobs contributes negatively to the job quality index • Working time and work–​life balance (long and atypical hours—​both inverted, voluntary part-​time work, satisfaction with working time) • Working conditions and job security (work intensity, work autonomy, physical work factors, perception of job security) • Skills and career development (training, prospects for career advancement) • Collective interest representation (collective bargaining coverage, trade union density).

Overall Changes in the JQI and Differences by Gender This sub-​section examines overall changes in job quality in EU member states, also disaggregating the analysis by gender. Looking at the average developments over time, a marginal overall decline in job quality between 2005 and 2010 is observed (Figure 16.1). The disaggregation by gender reveals only very small differences with regard to the change between the 2005 and 2010 JQI and its six dimensions. Moreover, the direction of change is the same for men and women in all cases. Improvements are visible with regard to working conditions and working-​time and work–​life balance. However, there are marked deteriorations, evident in terms of wages2 and involuntary non-​standard employment. Slight declines are also visible in skills and career development and in collective interest representation.

1  One of the main data sources used for constructing the JQI, the EWCS, is only available at a five-​year interval. 2  It is important to note that the wage data in the 2005 JQI were from 2007; this choice was made at the time to ensure that the most up-​to-​date data available were used. In terms of wages, therefore, the point made on various occasions in this chapter, that the period considered includes both ‘boom’ and ‘bust’ phases, does not apply. We see a clear effect of the recession.

368    christine erhel et al 0.7 0.6 0.5 0.4 0.3 0.2 0.1

JQ I

te re st

to ta l

re p

(a dj .)

r.

v. lle ct ive Co

sa nd

in

ca re

er

jo b + Sk ill

ns tio di

kin

g

co n

de

c. se

LB W e+

g kin W or W or

In

vo l

.n

on

-s ta nd

ar d

tim

(in

W ag e

ve rse

)

s

0

2005

2010

Figure 16.1  Changes in sub-​indices of Job Quality Index between 2005 and 2010, EU27 Data Sources: Ameco, LFS, National Accounts, EWCS, EU-​SILC, ICTWSS database, own calculations

The decline in the overall JQI is slightly more pronounced among women, reflecting more pronounced declines or less pronounced improvements across all the five sub-​indices that are gendered. It is important to note, however, that the outcomes by gender differ markedly between sub-​indices in both years. Men do considerably better in terms of wages and they are less affected by involuntary part-​time and fixed-​term work, whereas women do better on the sub-​index of working time and work–​life balance—​also due to constrained choices related to family obligations in particular—​as well as working conditions, which reflects, among other things, sectoral segregation. Figure 16.2 shows developments in overall job quality by country. The following countries recorded a more than marginal improvement in overall job quality (ranked by the absolute size of the improvement): Poland, Czech Republic, Belgium and Denmark. By contrast, Ireland and France saw marked declines in job quality. There is no clear pattern of either convergence or divergence among countries. This lack of pattern is illustrated by, among other things, the fact that both the lowest and highest performing countries in 2005 in terms of overall job quality—​Poland and Denmark—​experienced substantial increases. Disaggregating changes in the JQI by gender does not in most cases reveal major differences across countries. Notably, those countries with substantial improvements or deteriorations tended to register a balanced development between men and women. The most pronounced improvements and deteriorations according to the six dimensions are summarized in Table 16.1.

Table 16.1 Most pronounced improvements and deteriorations in Job Quality Index (total and sub-​indices)—​2005–​2010

Wages

Working-​ Working Involuntary non-​standard time and conditions and job Skills and career WLB security development (inverted)

Improvement IE

PL, BE, LT

RO, PL, LV, CZ, PT, PL, DE, FI SK, HU, BG

LU, PL, CY, EE, BE

Deterioration RO, DE

IE, IT, UK

FR

FR, SE

IE, LU, FR, SE

Collective interest representation

JQI total PL, CZ, BE, DK

SK, PT, EE

IE, FR, UK, SE

Note: based on our own calculations, only results for improvements and deteriorations of more than +​/​-​0.1 compared with the respective EU27 average are presented. In the case of the overall JQI +​/​-​0.04 was used. Countries are displayed in the order of the magnitude of improvement and, respectively, deterioration. Country abbreviations are displayed in Appendix 1.

370    christine erhel et al 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Poland Romania Greece Lithuania Bulgaria Hungary Slovakia Latvia Spain Czech Republic Estonia Portugal Italy Germany EU27 Cyprus EU15 Slovenia France Malta Austria Ireland Belgium Luxembourg United Kingdom Finland Netherlands Sweden Denmark

0

2005

2010

Figure 16.2  Developments in overall Job Quality Index by country, 2005–​2010 Data Sources: Ameco, LFS, National Accounts, EWCS, EU-​SILC, ICTWSS database, own calculations

Few countries show marked improvements or deteriorations in more than one sub-​index. The exceptions are Poland, France, Sweden, Belgium and Ireland. Poland, the country with the largest total improvement, is a clear-​cut example, with marked improvements in all dimensions except wages and collective interest representation—​ albeit from a very low starting point. The other three countries that show marked improvements in overall job quality have very different profiles. Belgium improved markedly on both involuntary non-​standard employment and skills and career development. The Czech Republic shows marked improvements on working conditions and job security. Ireland, the country with the largest negative development in overall job quality, is an interesting case in that it combines deteriorations of large magnitude in two fields—​ involuntary non-​standard employment and working conditions—​with improvements in terms of the sub-​index wages.3 France, the country with the second largest decline in overall job quality, shows marked deteriorations in three fields: working time and work–​life balance, working conditions and job security, and skills and career development. The UK shows an overall negative trend that is particularly due to negative developments with regard to involuntary non-​standard employment. Sweden displays noticeable declines in two fields—​working conditions and job security, and skills and career developments.

3  In the case of Ireland, the improvement in ‘wages’ largely reflects the fact that while in Europe as a whole the share of working poor—​which is 30 per cent of the sub-​index—​rose, in Ireland it fell. This may also be a compositional effect (job losses at the bottom, income losses at the top of the wage distribution).

the great recession and job quality trends in Europe    371

What Links Exist between Job Quality and the Intensity of the Crisis? Two hypotheses implying opposing impacts on job quality can be put forward to analyse the effect of the crisis on job quality: the ‘bargaining hypothesis’—​the rise in unemployment and the declining bargaining power of labour brought on by the crisis are expected to have a deleterious effect on the quality of existing jobs—​and the ‘compositional hypothesis’—​according to which the crisis leads to a disproportional destruction of low-​quality jobs and thus an improvement in average job quality. As they cancel each other out, it may be difficult to distinguish these effects using the aggregate data on which the JQI is based. The sub-​indices are expected to differ in their sensitivity to the overall crisis impacts but also regarding the relevance of the bargaining vs the composition hypotheses. Some of the indicators (for instance in collective interest representation) are institutional in nature and thus likely to be slow moving; crisis effects will not be anticipated in the short time-​span considered here. In other areas, lags are likely to be shorter (e.g. concerns about job security, working time, non-​standard employment). Some suggestive linkages between the JQI and unemployment dynamics emerge but there is no consistent overall picture (Figure 16.3). Most obviously, the country with the greatest job quality improvement according to our index, Poland, also enjoyed the strongest real economic growth and the sharpest fall in unemployment. This finding seems to indicate that quantitative and qualitative job improvements are indeed complementary and probably reflect a mixture of technological upgrading, sectoral change and an improvement in workers’ bargaining power. Conversely, Ireland, the country which suffered the sharpest fall in the JQI, was one of the countries worst hit by the crisis.

Figure 16.3  Change in the unemployment rate between 2005 and 2010 and change in the overall Job Quality Index, 2005–​2010 Data Sources: Ameco, LFS, National Accounts, EWCS, EU-​SILC, ICTWSS database, own calculations

372    christine erhel et al However, the pattern of quantitative labour market change and our overall measure of job quality is not consistent, as is illustrated by the other three countries that substantially improved their performance between 2005 and 2010. Belgium and the Czech Republic were negatively affected by the crisis to a below average extent, but Denmark, the best overall performer in 2005 and 2010, saw no economic growth and experienced a substantial rise in unemployment over the period. Therefore, as a first result of cross-​ country analysis between 2005 and 2010, the links between job quality and the intensity of the crisis appear weak. This finding could be explained by the observation period, which does not coincide precisely with the economic downturn, but it reveals also that there are other determinants of job quality trends, such as the structure of employment by sector and occupation (Amossé and Kalugina 2013) as well as institutional factors (Fernandez-​Macias 2012).

Changes to the Job Quality Dimensions over the Crisis Turning now to the six dimensions, there does appear to be some link between the size of the economic and labour market shock and job quality performance on the various aspects of job quality. Nevertheless, any such link is rather tenuous and, as hypothesized, is stronger for some dimensions. It seems plausible once again to invoke differential lags as a likely explanation. With regard to the wage dimension, little correlation emerges between quantitative labour market developments and this aspect of job quality. Countries with the greatest positive change include both Poland and Ireland, with entirely different economic and labour market developments. At the other end of the scale, there are the crisis-​hit countries such as Greece, but also Germany. As mentioned, the wages comparison is between the years 2007 and 2010. In the case of non-​standard employment, the relationship seems more clear-​cut. The worst instances of deterioration are to be found in the southern periphery countries and in Ireland and the UK. The greatest improvements were among countries with, relatively speaking, favourable economic and unemployment trends (Poland, Belgium, Finland, Luxembourg and Germany). Both involuntary temporary employment and involuntary part-​time employment contributed to the large decline in this sub-​index in Ireland, Italy and the UK. This finding suggests that, in these countries, employers enforced working-​ time reductions and/​or increased use of fixed-​term contracts that were seen as negative from the workers’ perspective. Spain, by contrast, displayed very large composition effects: fixed-​term workers, in particular, bore the brunt of the labour market adjustment, with the consequence that their share of total employment declined substantially. The countries with marked improvements in this sub-​index had more varied patterns. Poland, with a tightening labour market over much of the period, saw a fall in the share of part-​time employment, with more workers obtaining full-​time employment and, presumably, some conversion of part-​time to full-​time jobs. In addition, fewer of these part-​timers reported that they could not find a full-​time job.

the great recession and job quality trends in Europe    373 For working time and work–​life balance there is a secular and regionally specific trend, as all the substantial improvements occurred in the Central and Eastern European (CEE) countries—​with improvements across all the four sub-​indicators in virtually all cases. There has been a fall in the incidence of long and atypical working hours, while the two indicators that aim to capture work–​life balance show a consistent improvement. This finding may reflect the fact that companies in crisis have reduced overtime work and scrapped extra shifts. Declines were registered only by western European countries. The three countries with the largest declines are countries with a relatively favourable (Germany), average (France) and unfavourable (Greece) experience over the 2005–​2010 period; the crisis thus seems not to have had a consistent impact on this dimension. With regard to working conditions and job security, particularly the latter variable, which picks up ‘subjective job security’, is expected to be impacted directly by rising unemployment. Indeed, not only does the overall level of unemployment matter (not shown) but there is also a strong negative correlation between the changes in unemployment and the changes in the share of workers not worried about losing their job (Figure 16.4). Germany and Poland, the countries that experienced the most marked declines in unemployment in the period 2005 and 2010, are the only ones where the share of people not worried about job security increased noticeably. The opposite is true for high unemployment countries such as the Baltic countries, Spain and Ireland which, saw the share of workers not worried about losing their job decline by around 30 per cent and—​in the case of Lithuania—​close to 40 per cent. These findings illustrate a clear relationship between labour market performance and this subjective indicator of job quality. In terms of working conditions (work intensity, autonomy and physical work factors) there is some evidence that composition effects have served to raise measured average

Figure 16.4  Changes in unemployment rate and changes in share of people ‘not worried about losing job’, 2010–​2005 Data sources: LFS and EWCS, own calculations.

374    christine erhel et al job quality in this dimension, while the economic slack, whatever its other negative impacts, has in many cases reduced stress and pressure at work, at least for some workers. At the same time, the rather positive experiences in Germany and especially Poland and the negative trends in Ireland are in line with the bargaining power hypothesis in which good (bad) labour market developments exert upward (downward) pressure on job quality. Overall the findings are thus mixed. Moving to skills and career development we see that both the groups of countries experiencing and improvement and those suffering a decline in this aspect of job quality contain some member states badly and others relatively lightly affected by the crisis: Estonia and Poland, and Lithuania and Sweden respectively. The same conclusion can be drawn for collective interest representation, an institutional variable that is relatively slow moving; the findings are mixed. Overall this aggregate analysis of job quality reveals some limited correlation between the economic cycle and global trends in job quality, mostly concentrated on the ‘non-​ standard employment’ dimension and on subjective indicators. One limit of the JQI is that it captures only aggregate trends for 2005 and 2010, thus including (except for wages) also the pre-​crisis boom. In the following section, we focus on the 2007–​2009 period complementing the above analysis with longitudinal data.

Job Quality Changes at the Individual Level: Individual and Contextual Factors In this section we analyse the trends in job quality using a more limited set of indicators, which can be calculated at the individual level and in longitudinal perspective. Our aim is to disaggregate the analysis and to identify trends through a sample of individuals who were on the labour market in 2007. Our analysis is based on the European Union Statistics on Income and Living Conditions (EU-​SILC) panel that follows individuals over a four-​year period in twenty European countries4 and provides information about their labour market and income situation. Although the set of variables available is limited, it is the only comparative dataset that allows such an observation period. We use data for three years (2007 to 2009) so as to characterize individual trajectories between the beginning of the crisis and 2009. The dataset provides information about individuals’ labour market situation, including some variables that reflect the job quality indicators considered in the previous section: type of contract (temporary vs permanent, part-​time vs full-​time), 4  Given missing variables at the individual and country level this part of the analysis includes all EU28 countries except Bulgaria, Croatia, Germany, Ireland, Cyprus, Lithuania, Luxembourg, Malta, Romania (+​Norway).

the great recession and job quality trends in Europe    375 weekly working time and wages, and occupation (using the International Standard Classification of Occupations, ISCO). The latter can be used to proxy skills and career development. We chose not to use information on wages because it is not available in a suitable form for a large number of countries in the EU-​SILC panel5. In comparison with the multidimensional framework presented earlier, the main limitation here is the absence of information on working conditions, training or collective interest representation dimensions (sub-​indices 4, 5 and 6 of the Job Quality Index), and the fact that the variables provided in the longitudinal survey do not allow the researcher to distinguish between voluntary and involuntary non-​standard employment. Basic individual characteristics including gender, educational level, and the number and age of children can be identified. To summarize the trends in job quality at the individual level during the crisis, we compute an indicator of job quality deterioration (from 2007 to 2009) for all individuals in employment in 2007. It includes the following criteria: either a transition from permanent to temporary job or a transition from full-​time to part-​time work or a change in occupation that can be interpreted as downward mobility (from high-​skilled to low-​ skilled clerical or manual). We then analyse the factors influencing the probability of experiencing a decrease in job quality from 2007 to 2009 (for individuals employed in 2007). Given the importance of job destructions and transitions to non-​employment (whether inactivity or unemployment) during a recession, the analysis of individual transitions also takes into account transitions to non-​employment. Therefore the perspective taken in this section is wider and does not entirely focus on job quality trends. Job quality changes at the individual level are assessed by using multilevel multinomial logit models to account for both individual factors and country characteristics. This approach provides a complementary insight into changes to aggregate job quality during the crisis as it allows variations in job quality at the individual level to be taken into account. It thereby accounts for possible compensating effects over the population, especially if there have been divergent developments across sociodemographic groups. For instance, a stable overall trend in job quality in a given country may in fact conceal a two-​way trend of deterioration for some individuals and improvement for others.

The Methodology and the Use of Multilevel Models to Explain Job Quality Changes at the Individual Level over the Crisis Multilevel models offer an interesting framework enabling both individual and contextual determinants of an observed event to be taken into account (Snijders and Bosker 1999). They differ from more traditional techniques such as fixed effects regressions

5 

There are many missing values for wages.

376    christine erhel et al mainly because their goal is not to ‘cure’ regressions of contextual effects but to interpret and compare these effects. Multilevel models are used here since we assume that a change to job quality measured at the micro level is likely to be explained by both individual and institutional/​country characteristics (specific national institutions) and also by specific economic trends, especially in the crisis. As mentioned before, the dependent variable in our model can take three values. In all regressions presented below, the reference state is ‘employed in 2009 with no decrease in job quality compared to 2007’ (i.e. job quality can be the same or higher). In each model, there are two sets of equations: a first set representing the log odds of being out of employment relative to being in employment with no decrease in job quality compared to 2007, and a second set representing the log odds of being in employment with a lower job quality than in 2007 relative to being in employment with no decrease in job quality compared to 2007. The methodology proceeds in different steps that are described in detail in Appendix 2. First, we estimate an ‘empty’ or ‘unconditional’ model that only includes a random intercept. This empty model delivers a picture of countries’ relative situation in terms of change to job quality measured at the individual level between 2007 and 2009. In a second step, individual variables are introduced to estimate the influence of individual characteristics on job quality change. In a third step, several models are tested with contextual variables introduced to see how the national context and institutions affect the individual probability of experiencing a transition out of employment or a decrease in job quality. Contextual variables (listed in appendix) relate to the economic context (change in unemployment during the recession, economic structure by sectors) as well as to some labour market institutions which may impact workers’ transitions (labour market policy expenditures, employment protection legislation). Finally, we focus on the effect of contextual variables on some specific sociodemographic groups (women, youth, older workers and low-​educated workers) by making these contextual variables interact with individual variables.

Relative Position of Countries The estimation of multilevel models requires the availability of both contextual and individual data for all the countries that are introduced into the analysis. Therefore, our sample is restricted to twenty countries for which both EU-​SILC panel and institutional or sector data are available. It is composed of 57,149 individuals. As shown by the results of the empty model (first step of the multilevel analysis, see Table 16.2), on average in the 20 European countries considered here, people who were employed in 2007 are less likely to be out of employment in 2009 than to be employed with an equal or higher job quality. They are also less likely to be employed with a lower job quality in 2009 than to be employed with an equal or higher job quality. However, these probabilities vary across countries as shown by the country-​specific effects presented in Figures 16.5a and 16.5b. The two average log odds are represented

the great recession and job quality trends in Europe    377 Table 16.2 Results of the empty model Non-​employment vs no deterioration in job quality

Deterioration in job quality vs no deterioration in job quality

Intercept

-​2.661***

-​2.196***

Variance of the intercept

0.27951

0.25353

Source: EU-​SILC individual data (longitudinal database from 2007 to 2009); 57149 individuals aged 15 to 64 years; 20 countries Note: +​p < 0,10 ; * p < 0,05 ; ** p < 0,01 ; *** p < 0,001

Table 16.3 Role of individual characteristics on job quality change between 2007 and 2009 Not employed vs job quality staying the same or improving

Job quality deteriorating vs job quality staying the same or improving

Coefficient

p-​value

Coefficient

p-​value

Intercept

–​2.953342