Well-being in Belgium: Beyond Happiness and Income [1st ed.] 9783030585082, 9783030585099

Citation preview

Economic Studies in Inequality, Social Exclusion and Well-Being Series Editor: Jacques Silber

Bart Capéau · Laurens Cherchye · Koen Decancq · André Decoster · Bram De Rock · François Maniquet · Annemie Nys · Guillaume Périlleux · Eve Ramaekers · Zoé Rongé · Erik Schokkaert · Frederic Vermeulen

Well-being in Belgium Beyond Happiness and Income

Economic Studies in Inequality, Social Exclusion and Well-Being Series Editor Jacques Silber, Ramat Gan, Israel

Traditionally, economists have identified well-being with market command over goods rather than with the “state” of a person. The books in this series go precisely beyond the traditional concepts of consumption, income or wealth and offer a broad, inclusive view of inequality and well-being. Topics include, but are not limited to: Capabilities and Inequalities, Discrimination and Segregation in the Labour Market, Equality of Opportunities, Globalization and Inequality, Human Development and the Quality of Life, Income and Social Mobility, Inequality and Development, Inequality and Happiness, Inequality and Malnutrition, Income and Social Mobility, Inequality in Consumption and Time Use, Inequalities in Health and Education, Multidimensional Inequality and Poverty Measurement, Polarization, Poverty among Children and Elderly People, Social Policy and the Welfare State, and Wealth Distribution.

More information about this series at http://www.springer.com/series/7140

Bart Capéau
Laurens Cherchye
Koen Decancq
André Decoster
Bram De Rock
François Maniquet
Annemie Nys
Guillaume Périlleux
Eve Ramaekers
Zoé Rongé
Erik Schokkaert
Frederic Vermeulen

Well-being in Belgium Beyond Happiness and Income

123

Bart Capéau KU Leuven Leuven, Belgium

Laurens Cherchye KU Leuven Leuven, Belgium

Koen Decancq University of Antwerp Antwerp, Belgium

André Decoster KU Leuven Leuven, Belgium

Bram De Rock Université Libre de Bruxelles Brussels, Belgium

François Maniquet Université Catholique de Louvain Louvain-la-Neuve, Belgium

Annemie Nys University of Antwerp Antwerp, Belgium

Guillaume Périlleux Université Libre de Bruxelles Brussels, Belgium

Eve Ramaekers Namur, Belgium

Zoé Rongé KU Leuven Leuven, Belgium

Erik Schokkaert KU Leuven Leuven, Belgium

Frederic Vermeulen KU Leuven Leuven, Belgium

Translated by Blue Lines (www.bluelines.be).

First published as Wat heet dan gelukkig zijn? Geluk, welvaart en welzijn van de Belgen. © Garant-Uitgevers nv & the authors. ISSN 2364-107X ISSN 2364-1088 (electronic) Economic Studies in Inequality, Social Exclusion and Well-Being ISBN 978-3-030-58508-2 ISBN 978-3-030-58509-9 (eBook) https://doi.org/10.1007/978-3-030-58509-9 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Foreword

At the time of writing this foreword, a pandemic is wreaking havoc throughout the world. People lose their comfort, their usual social gatherings, they are afraid for their lives, and many lose their livelihoods too. Hard decisions have to be made, trading off lives against jobs, health security against financial security. In such a crisis, governments would need a compass telling them what is best for the population. Welfare economics, the branch of economics which studies societal well-being, is supposed to provide such a compass. In order to evaluate the well-being of a population, one needs a measure of the well-being of each member of society and a way to assess the distribution of well-being between different social groups. But it is fair to say that much work remains to be done to provide simple, reliable and accessible tools making such a measurement possible. One key difficulty is that people’s lives have many aspects and that people have a diversity of views about what is important for them. Very few surveys cover all these aspects at the same time. Typically, some data describe inequalities in income and wealth while other data describe the distribution of health, or how people spend their time, or what they think of their lodging conditions. This book makes a pioneering contribution toward addressing this issue. Its comprehensive survey of well-being in Belgium has covered essentially all aspects of people’s lives, and offers a unique snapshot of the situation of the population in the country, shedding light on many facets. While covering the whole distribution in the social hierarchy, it focuses on the least advantaged populations and this should be of tremendous use at a time when one needs to preserve the most vulnerable from the economic and social devastation brought by the current pandemic. Differences and inequalities between men and women are also thoroughly scrutinized, revealing important variations across social groups by level of income or education. In addition, the book explores how much people care about various aspects of their lives, and even how not only their objective conditions of life, but also their personality traits affect how satisfied they are with their situation, how happy they feel, as well as what matters to them. The authors offer useful warnings against easy shortcuts to the assessment of well-being. As they argue, well-being is not the same thing as subjective happiness. As the survey shows, a sizable fraction of people are happy in spite of dire

v

vi

Foreword

conditions of life. It is a beneficial human characteristic that people can adapt to difficult conditions, but, as the authors argue, this should not be a reason to redirect social support away from them. Economists and social scientists are still debating how to take account of people’s own perspective on their own situation. At one end in this debate, some argue for a purely objective approach which concentrates exclusively on hard facts about living conditions. At the other end, others argue that subjective happiness is a good proxy and is more respectful of people’s own views. This book defends a middle-ground position which I find very convincing. A purely objective measure gives us no idea of what is truly important to people. It cannot help in assessing the trade-off between health security and financial security that we are confronting in the current crisis. But a purely subjective statement of happiness depends too much on individual personality and adaptation to guide the allocation of public resources and social policies. Therefore the survey is used to elicit the importance of different aspects of life such as health and lodging conditions, providing valuable data on the trade-offs that people are willing to make between the many dimensions of their lives. Belgium can be considered one of the most advanced societies in the world, in terms of social equality. Nevertheless, this survey reveals important inequalities along several dimensions of disadvantage that fail to be noticed by the usual statistics of income. In particular, such a survey is uniquely able to show the correlation between different aspects of life. The multiple correlations between deprivations, gender inequality in the family, income, education, health and quality of lodging show a map of social stratification that was not previously available and will be most useful to anyone interested in fostering social cohesion and supporting the most vulnerable. My best hope for this book is that it will launch a series of similar surveys around the world. A decade after the Sen-Stiglitz-Fitoussi commission argued that measuring social progress required comprehensive surveys of this sort, the authors have made great strides toward delineating a methodology that implements this broad vision while avoiding two pitfalls to which some readers of the Sen-Stiglitz-Fitoussi report have succumbed. One pitfall consists in looking for shortcuts such as subjective happiness, and the other pitfall consists in casting different aspects of life into dashboards that provide no way to make trade-offs according to the population’s own priorities. The authors of this book do not dodge difficulties and do set their sights on the type of comprehensive assessment that everyone needs for improving our institutions and our social fabric. At a critical moment of history in which the importance of public services, the strong interdependence between people, and the need to protect the most vulnerable appear more vividly to many decision-makers, developing measurement tools to assess well-being, as this book does in a clear and accessible but uncompromising way, is doing all of us a great service. Princeton, USA

Marc Fleurbaey

Acknowledgements

This book is a collaboration between many different people and institutions. First and foremost, we would like to thank the Belgian Science Policy Office (BELSPO) for funding the MEQIN project that led to this book. We would also like to express our gratitude to Aziz Naji of BELSPO for monitoring the project and for his flexibility, which allowed us to concentrate fully on its content and objectives. We would also like to thank Francisco Santana Ferra (Université Catholique de Louvain) and Andras Avonts (KU Leuven) for their efficient administrative support. Many people also helped with the MEQIN survey on which the figures in this book are based. Karel Van den Bosch (Federal Planning Bureau and University of Antwerp) helped us to develop the method for drawing our representative sample. Koen Ponnet (University of Antwerp and Ghent University) and Carine Van de Voorde (Belgian Health Care Knowledge Centre and KU Leuven) contributed a great deal to developing the survey. Bea Cantillon (University of Antwerp) provided us with excellent advice that proved very useful. We would like to express our sincere thanks to all concerned for their assistance. This book was first published in Dutch. A French translation is also available. We would like to express our gratitude to Marion Collewet (Université Catholique de Louvain) for her careful checking of the French translation. This book also has a website that provides more detailed background information about the MEQIN survey. The website was expertly built and created by Veerle Hennebel (KU Leuven). We also owe her a debt of gratitude. Finally, we would like to thank Kantar Belgium for the practical and professional implementation of the survey. In particular, we would like to express our gratitude to the dozens of interviewers who collected the necessary data from over 3000 individuals in more than 2000 Belgian families. We would of course also like to thank the latter. Without their willingness to provide insights into their lives, we would never have been able to write this book.

vii

Contents

1

What Constitutes a Good Life? . Income . . . . . . . . . . . . . . . . . . . . Happiness . . . . . . . . . . . . . . . . . . Well-Being and Capabilities . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

1 2 3 5

2

Why This Book? . . . . . . . . . . . . Structure of the Book . . . . . . . . . Why a New Dataset? . . . . . . . . . Some Preliminary Methodological

.......... .......... .......... Comments . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

7 8 9 10

3

Is the Conventional Family Still the Cornerstone of Which Family Types Are Found in Belgium? . . . . . . . Are There Differences Between the Regions? . . . . . . . Age and Family Type . . . . . . . . . . . . . . . . . . . . . . . .

Society? . . . . . ............ ............ ............

13 13 15 16

. . . .

. . . .

21 22 24 25

Part I

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

Well-being Involves Many Different Aspects

4

How Unequally Are Our Incomes Distributed? . A Parade of Dwarves and Giants . . . . . . . . . . . . . A Spoke in the Wheel? . . . . . . . . . . . . . . . . . . . . Inequality Measured . . . . . . . . . . . . . . . . . . . . . .

5

Who Is Poor in Our Society? . . . . . . . . . . . . . . . . . . . . . . . . . Risk of Poverty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Material Deprivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . How Many Belgians Are Both on a Low Income and Materially Deprived? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.... .... ....

27 27 29

....

30

How Healthy Are We? . What Is Health? . . . . . . . General Health . . . . . . Functional Limitations . Chronic Diseases . . . . . Emotional Well-Being . Physical Well-Being . .

. . . . . . .

33 33 33 34 35 35 36

6

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . .

. . . . . . .

. . . .

. . . . . . .

. . . .

. . . . . . .

. . . . . . .

. . . .

. . . . . . .

. . . .

. . . . . . .

. . . .

. . . . . . .

. . . .

. . . . . . .

. . . .

. . . . . . .

. . . .

. . . . . . .

. . . .

. . . . . . .

. . . .

. . . .

. . . . . . .

. . . .

. . . . . . .

. . . . . . .

ix

x

Contents

Who Is Ill? . . . . . . . . . . . . . . Are Poorer People Sicker? . How Sick Are the Elderly? Are the Flemish Healthier?

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

36 37 38 40

7

What Makes Us Sick? . . . . . Lifestyle, Living Environment Lifestyle . . . . . . . . . . . . . . Living Environment . . . . . . Job Characteristics . . . . . . . Emotional Well-Being . . . . . . The Shadow of the Past . . . .

................... and Job Characteristics . ................... ................... ................... ................... ...................

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

43 43 45 45 46 47 47

8

Can People Afford Their Healthcare? . The Use of Healthcare . . . . . . . . . . . . . Financial Consequences of Illness . . . . . Postponement of Care . . . . . . . . . . . . . . People with Chronic Illnesses . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

49 49 53 57 60

9

Do We Find the Job of Our Dreams? . . . . . . . . . . . . Who Works and Who Doesn’t? . . . . . . . . . . . . . . . . . . Which Types of Jobs Are Carried Out? How Much Do People Work? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Job of My Dreams, an Unattainable Ideal? . . . . . .

........... ...........

63 63

........... ...........

67 68

...........

73

........... ........... ...........

73 76 77

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . . .

. . . .

. . . . .

. . . .

. . . . .

. . . .

. . . . .

. . . .

. . . . .

. . . .

. . . . .

. . . .

. . . . .

. . . .

. . . . .

. . . .

. . . . .

. . . .

. . . . .

. . . . .

10 What Do We Spend Our Money on? . . . . . . . . . . . . What Do the Consumption Patterns of Belgian Families Look Like? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Expenditure Patterns per Income Quartile . . . . . . . . . . . Expenditure Patterns by Level of Education . . . . . . . . .

. . . .

79 79 80 84

12 How Do We Spend Our Time? . . . . . . . . . . . . . . . . . . . . . . . . . . . . Time Use of Men and Women by Family Type . . . . . . . . . . . . . . . . . Time Use of Men and Women by Level of Education . . . . . . . . . . . .

85 86 88

11 Do We Live Comfortably and in a Pleasant Environment? The Quality of Our Housing . . . . . . . . . . . . . . . . . . . . . . . . . Tenants and Owners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Elderly People and Young People . . . . . . . . . . . . . . . . . . . . .

Part II

. . . .

. . . .

. . . .

. . . .

. . . .

An Insight into Families

13 Who Forms a Couple with Whom? . . . . . How Did Partners First Meet? . . . . . . . . . . . Like Seeks Like? . . . . . . . . . . . . . . . . . . . . Is Like-Seeks-Like Behaviour Age-Related? .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

93 93 95 97

Contents

xi

99 99

14 How Do Partners Within Couples Spend Their Time? . . . . . . . Time Use of Partners Across All Types of Couples . . . . . . . . . . . Time Use of Working Couples with and Without Children Living at Home . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Time Use of Partners by Level of Education . . . . . . . . . . . . . . . .

. . . 100 . . . 101

15 What Do Partners Expenditure Within Expenditure Within Expenditure Within

. . . .

in Couples Spend Their Money on? . . . Couples with and Without Children . . . . . Couples According to Total Expenditure . Couples According to Level of Education

16 Who Wears the Trousers? . . . . . . . . . . . . . . . . . . . . . . . Distribution of Time and Money Between Partners . . . . . . . Distribution of Time and Money in Relation to Relative Hourly Wage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Distribution of Time and Money in Relation to Full Family Income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marriage Market . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Part III

. . . .

. . . .

. . . .

... ...

. . . .

. . . .

107 108 110 112

. . . . . . . . 115 . . . . . . . . 115 . . . . . . . . 116 . . . . . . . . 118 . . . . . . . . 119

Who Deserves Special Attention? . . . . . . . . . . 123 . . . . . . . . . . 123

17 Is Life Harder for Single-Parent Families? . . . . . . . . . Who Are the Parents in Single-Parent Families? . . . . . . . Do Parents in Single-Parent Families Have Lower Levels of Well-Being? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Which Dimensions of Well-Being Make the Difference? . Are There Differences When It Comes to the Children? .

. . . . . . . . . . 124 . . . . . . . . . . 125 . . . . . . . . . . 127

18 Which Children Grow up in Poverty? . . . . . . Poverty Is Hard . . . . . . . . . . . . . . . . . . . . . . . . Health and Housing Quality . . . . . . . . . . . . . . . How Much Do Parents Invest in Their Children? As Long as They’re Happy … . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

129 130 131 133 134

19 A Nice Retirement? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 How Many Older People Are There? . . . . . . . . . . . . . . . . . . . . . . . . . 137 The Situation of the Elderly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 Part IV

Towards a Measure of Individual Well-being

20 Who Suffers from Cumulative Deprivation? . . . . . . . . . . . . . . . . . . 145 How Much Cumulative Deprivation Is There in Belgium? . . . . . . . . . 145 Who Suffers from Cumulative Deprivation? . . . . . . . . . . . . . . . . . . . . 147

xii

Contents

21 How Happy Are We? . . . . . . . . . . . . . . . . . . . . Happiness and Life Satisfaction . . . . . . . . . . . . . . What Makes Belgians Satisfied with Their Lives? . (Dis)satisfaction with a Partial Aspect of Life . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

151 151 153 156

22 As Long as We’re Happy …? . . . . Personality and Life Satisfaction . . . Level of Education and Expectations Poor but Happy . . . . . . . . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

159 161 161 164

23 How Do We Measure Well-Being? . . . . . . . . . . . . . . . An Example of the Alternative Approach . . . . . . . . . . . . Different Opinions on the Good Life: Willingness to Pay Equivalent Income . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

167 168 169 172

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

24 Who Has the Lowest Levels of Well-Being? . . . . . . . . . . . . . . . . . . 175 How Do We Measure Willingness to Pay? . . . . . . . . . . . . . . . . . . . . . 176 Equivalent Incomes for Health and Housing . . . . . . . . . . . . . . . . . . . . 179 25 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Just Looking at Averages Can Be Misleading . . . . . Income Is Not a Good Measure of Well-being . . . . Happiness Is Not a Good Measure of Well-being . . Well-being Is Best Measured in a Multidimensional

. . . .

. . . .

. . . .

. . . .

. . . . Way .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

183 183 184 184 185

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187

About the Authors

Bart Capéau (Geraardsbergen, 1963) holds a master’s degree in philosophy and has a Ph.D. in economics. He is a researcher at the Department of Economics of KU Leuven. Previously, he was affiliated with HIVA—Research Institute for Work and Society (KU Leuven) and the European Centre for Advanced Research in Economics and Statistics (ECARES) of the Université libre de Bruxelles. Laurens Cherchye (Ypres, 1974) is a full professor at KU Leuven. He is also an honorary senior research associate at University College London and an international research fellow at the Institute for Fiscal Studies (London). His research interests lie within family economics with a focus on consumption, labour supply and the distribution of time and money within families. He has published over 70 articles in international journals including Econometrica, American Economic Review, Review of Economic Studies and the Journal of Political Economy. Among other things, his research was awarded the EURO Management Science Strategic Innovation Prize in 2004, the Vereniging Voor Economie (Economics Association) prize in 2008, the Vlaamse Wetenschappelijke Stichting (Flemish Scientific Foundation) prize in 2014, a Consolidator Grant from the European Research Council in 2014, the Pioneer Prize for Humanities from KU Leuven in 2015, together with Bram De Rock and Frederic Vermeulen, and the Francqui Prize in 2019, together with Bram De Rock and Frederic Vermeulen. Koen Decancq (Wilrijk, 1980) is a senior lecturer at the Faculty of Social Sciences of the University of Antwerp. He is an associate fellow at the London School of Economics and the Centre for Operations Research and Econometrics at UCLouvain. He is particularly interested in how inequality, poverty and well-being can be measured within a multidimensional context. His work in this area has been published in the Journal of Public Economics, Theoretical Economics and the Journal of Economic Theory, among others. As the country team leader, he is responsible for the collection of SHARE data among the 50+ age group in Flanders. He has appeared as an expert on various advisory committees for the development of multidimensional indicators of well-being and policy evaluation in the areas of well-being and care.

xiii

xiv

About the Authors

André Decoster (Avelgem, 1958) is a full professor at the Department of Economics of KU Leuven, where he lectures on introductory economics, public finances and the measurement of welfare, inequality and poverty. His research mainly focuses on the evaluation of tax systems, social security and government action, in general, using micro-econometric simulation models. His work in these areas has been published in the Journal of Public Economics, European Economic Review, Review of Income and Wealth and Journal of Policy Analysis and Management, among others. In 2015, he received KU Leuven’s Society Award for calculating the feasibility of election manifestos based on these models. Since 1997, he has been coordinating the editions of the introductory book Economie. Een inleiding. (Economics. An introduction). Bram De Rock (Bruges, 1977) is a full professor at the European Centre for Advanced Research in Economics and Statistics (ECARES) at the Université libre de Bruxelles and the Department of Economics at KU Leuven. He is also an honorary senior research associate at University College London and an international research fellow at the Institute for Fiscal Studies (London). His research interests lie within family economics with a focus on consumption, labour supply and the distribution of time and money within families. He has published over 50 articles in international journals including Econometrica, American Economic Review, Review of Economic Studies and the Journal of Political Economy. Among other things, his research has been awarded a European Research Council grant in 2010, the KU Leuven Research Council prize in 2012 and the Pioneer Prize for Humanities from KU Leuven in 2015, which he received together with Laurens Cherchye and Frederic Vermeulen. In 2019, he was awarded the Francqui Prize, together with Laurens Cherchye and Frederic Vermeulen. François Maniquet (Namur, 1965) is a full professor at the Department of Economics and Centre for Operations Research and Econometrics at UCLouvain in Louvain-la-Neuve. He is also an honorary researcher at the Fund for Scientific Research (FRS-FNRS). His research interests lie within welfare economics and public economics, with a specific focus on how welfare and well-being can be measured and the theory of optimal taxation. His work has been published in Econometrica, Review of Economic Studies, Journal of Economic Literature and the Journal of Economic Theory, among others. He received the Social Choice and Welfare Prize in 2004 and the Francqui Prize in 2010. Annemie Nys (Schoten, 1986) holds a master’s degree in applied economics from the University of Antwerp and a master’s degree in statistics for social sciences from KU Leuven. She conducts research at the Herman Deleeck Centre for Social Policy and is a Ph.D. student at the Department of Economics of the University of Antwerp, focusing on how well-being can be measured and compared.

About the Authors

xv

Guillaume Périlleux (Brussels, 1992) holds a master’s degree in economics and a master’s in advanced research from the Université libre de Bruxelles. He is a Ph.D. student at the Solvay Brussels School of Economics and Management (ULB) within the domain of family economics, focusing on how families make decisions about consumption and incurring debts. Eve Ramaekers (Namur, 1977) has a Ph.D. in economics. She was a visiting lecturer at the Centre for Operations Research and Econometrics at UCLouvain and held positions at the University of Namur and the Ecole Polytechnique near Paris. Focusing on welfare economics and game theory, her research results have been published in Social Choice and Welfare and Games and Economic Behavior. Zoé Rongé (Liège, 1991) obtained a master’s degree in economics (specialising in the economic analysis of government policy) from the University of Liège and a master’s degree in advanced economic research from KU Leuven. She is a Ph.D. student in the Department of Economics at KU Leuven within the domain of family economics, focusing on the distribution of time and money within families in relation to the subjective well-being of family members. Erik Schokkaert (Puurs, 1954) is an economist and psychologist. As a full professor, he is affiliated with the Department of Economics at KU Leuven and chairs the steering committee of Metaforum, the interdisciplinary Think Tank of KU Leuven. In his research, he tries to develop a conceptual framework for the evaluation of policy measures (mainly in the fields of health care and social security) in the light of different views on justice and individual well-being. He was one of the managing editors of Economics and Philosophy. As an expert, he has served on various (Belgian and Dutch) advisory committees that evaluate policy measures in the domain of health care and sits on the Academic Council for Pension Reform. Frederic Vermeulen (Kortrijk, 1974) is a full professor at the Department of Economics of KU Leuven. He is also a research fellow at the Institute for Fiscal Studies (London) and the Institute of Labor Economics (Bonn). He was previously affiliated with the University of Tilburg. He was a managing editor of the Economic Journal and was previously an editorial board member of the Review of Economic Studies. His research focuses on family economics, more specifically consumption, labour supply, the distribution of time and money within families and matching in the marriage market. His work has been published in Econometrica, the American Economic Review, the Review of Economic Studies, the Journal of Political Economy and the Review of Economics & Statistics, among others. Together with Laurens Cherchye and Bram De Rock, he received the Pioneer Prize for Humanities from KU Leuven in 2015 and the Vereniging Voor Economie (Economics Association) prize in 2008. In 2019, he was awarded the Francqui Prize, together with Laurens Cherchye and Bram De Rock.

1

What Constitutes a Good Life?

In this book, we try to give a picture of the individual well-being of Belgians. We describe which aspects of life determine this well-being and how these aspects are distributed in Belgium. Particular attention is paid to the situation of Belgians who achieve the lowest levels of well-being, as well as the situation of specific groups such as children, the elderly and single-parent families. First of all, we must of course define precisely what we mean by the broad and somewhat vague concept of “well-being” and why we wish to identify it in the first place. A policymaker who wishes to use society’s scarce resources to increase the well-being of its members, or to prioritise those with the lowest levels of well-being, clearly needs an operational benchmark for well-being. But when exactly would we say that someone has a high level of well-being or a “good life”? In this introductory chapter, we examine three different responses to this question. The first, standard, response focuses on income. The second and more fashionable response looks at happiness as a benchmark for well-being. Finally, we examine a rather experimental response based on “capabilities”: an individual’s capacity to achieve important aspects of life if they wish. However, we will start with an important preliminary remark. In this book, we are interested in the well-being of individuals within a society. This contrasts with the usual practice of looking at the well-being of an entire society or family. The well-being of an entire society often refers to (a form of) average well-being, taking into account the average of all people in the society. However, the average can paint a very misleading picture, because well-being within a society may not be equally distributed at all. In a society with a high average level of well-being, for example, there may still be people who are dying of hunger. It is also possible for average life expectancy to increase even though the life expectancy of specific social groups is decreasing. A current example that stands out is the deteriorating health of poor white people in the United States, as described by Case and Deaton (2015). Many studies on distribution and poverty focus on the well-being of families rather than individuals. This is usually a pragmatic choice, because not enough detailed information is available to accurately measure the well-being of individual © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_1

1

2

1

What Constitutes a Good Life?

members of a family. However, what applies to society also applies to the family. Although many goods might be shared within the family (such as heating and the TV, for example), there may still be considerable differences in the quantity and nature of the goods that the various family members have, as well as in how they use their time. As a result, partners in a couple may not necessarily achieve the same level of well-being. One of the main contributions of this book is precisely the fact that we have collected information about the individual members of Belgian families and can therefore chart their well-being individually.

Income It is still very common to study well-being and poverty by focusing on a monetary measure such as income or wealth. Income inequality has once again received more attention in recent years, especially since the worldwide success of Thomas Piketty’s book (2013). In this extensive work, Piketty combines a great deal of empirical material which shows that income inequality is increasing in most Western countries and that the concentration of wealth is growing rapidly among the richest members of society in particular. Income still remains the most widely used criterion for determining who lives in poverty: according to the official European poverty line, families with less than 60% of the median income of the society are at risk of poverty.1 There are good reasons to assume that income does indeed make a significant contribution to people’s well-being. After all, an income allows them to buy goods and services. This is required in order to satisfy basic needs such as food, drinks and a roof over one’s head. Someone who does not have these things is regarded as poor. However, a “good life” is likely to require more. Most people also want to be able to afford other things: occasionally going on holiday, eating in a restaurant from time to time or buying a season ticket for their favourite football club. Most of these items can be bought if you have money. Income is therefore important as it provides a good indication of people’s consumption opportunities. Different people will naturally choose to spend their income differently. In itself, this freedom to decide what to buy also contributes to individual well-being. However, a high income does not guarantee a good life. After all, well-being involves more than material consumption alone. This idea, too, has recently received a great deal of international attention.2 Many important things cannot simply be bought. For example, people want to be healthy, free of pain and able to 1

The median income is the level of income which divides the population into two equal groups, half having an income below the median and half having an income above that amount. 2 The Stiglitz–Sen–Fitoussi report (2009) played an important role as a catalyst here. International institutions such as the United Nations and the Organisation for Economic Co-operation and Development (OECD) are paying increasing attention to the non-material aspects of well-being and have constructed multidimensional benchmarks such as the Human Development Index and Better Life Index (see https://hdr.undp.org/ and https://www.oecdbetterlifeindex.org).

Income

3

perform physical activities. They need friends and wish to spend their time in a meaningful way, by having a job that they find satisfying or through other meaningful and enjoyable activities. There are many different aspects to a good life. Some of these aspects relate to income: people on higher incomes are often healthier, find it easier to maintain social relationships and have better jobs. However, the correlation is certainly not perfect and cuts both ways. It is precisely because these various important aspects of life are linked that inequality in well-being increases when we include these non-monetary aspects in the analysis. If poorer people are also sicker, become more isolated and no longer feel integrated in society, the inequality in terms of well-being is greater than the income inequality alone. In this book, we will therefore examine various aspects of individual well-being in more detail, together with the extent of the correlation between these aspects.

Happiness The increasing awareness that non-material dimensions are important for individual well-being has led to greater attention being paid to the measurement of happiness in recent years. It is not easy to describe exactly what we mean by happiness. Throughout history, the quest for happiness often had a metaphysical or religious interpretation. For example, happiness might mean going to heaven after one’s death. Happiness can also be interpreted in a rather more secular way, such as “feeling good” or “being satisfied with life”. These latter two interpretations are the ones we will use in this book. This is partly because more and more empirical research is being carried out with the intention of measuring these concepts. In the meantime, thousands of people in different countries have answered questions such as: “Overall, how happy would you say you are?” and/or “Overall, how satisfied are you with your life today?” People answer on a scale from 0 to 10, for example (where 0 stands for extremely unhappy or extremely dissatisfied and 10 stands for extremely happy or extremely satisfied). The answers of these thousands of people turn out to be surprisingly robust. The extensive literature on the determinants of happiness has produced some very striking findings. While both happiness and life satisfaction are linked to income, the correlation is much more tenuous than is often assumed. To better understand the interplay between income, happiness or life satisfaction, it is important to distinguish between evolutions over time and comparisons between different people at a given time. Over time, the impact of income growth tends to be rather limited. In this context, reference is made to the “Easterlin Paradox”: in 1974, Richard Easterlin was the first person to point out that the strong economic growth in the United States after the Second World War did not lead to a sharp increase in the average level of happiness. All the same, within a particular society, richer people are generally happier and more satisfied than poorer people, and people in the rich West are generally more satisfied with their lives than people in the poor

4

1

What Constitutes a Good Life?

South. At first sight, this apparent contradiction between the rather tenuous link connecting income and happiness over time on the one hand and on the other hand the stronger link between income and happiness at a given point may appear surprising, but it is easy to explain. Our happiness depends not only on our objective situation, but also on our aspirations and expectations. In turn, these aspirations and expectations depend on our past experience and a comparison of our situation with that of other people around us. We adjust our expectations as our objective situation changes. As we get richer, after a while, we will regard our increased income as normal, and it will therefore bring us less satisfaction. If we become richer but observe an even greater increase in income among the people around us, we will tend to fixate on the deterioration of our relative position rather than the increase in our own income. As yet, there is no consensus within scientific literature on the precise explanation of the evolution of happiness over time and the relative importance of income in this development.3 However, there is a consensus that happiness or life satisfaction is partly determined by many other factors. Health is essential. Empirical research also reveals that the quality of social relationships, job quality, the environment in which people live and feelings of safety have a major impact. In Chap. 21, we will paint a global picture of the happiness and life satisfaction of Belgians. However, happiness is not a satisfactory criterion for comparing the well-being of different people. Let us examine the situation of two people, Alain and Bert. Alain has a high income and can therefore afford a high level of consumption. He also has a good job in the financial sector, and he is healthy. Bert is struggling a little with his health and has a job that doesn’t bring him a great deal of satisfaction, but it does provide an income that is high enough to make ends meet fairly well. Bert would like to trade places with Alain, whereas Alain is not attracted to Bert’s life situation. All the same, Alain is (or feels) less satisfied than Bert. He comes from a very wealthy family, and as a child, he became used to an even higher standard of consumption than he can currently afford. Most of his friends are more successful than he is. Bert, by contrast, comes from a poor family and is proud to have got this far. His parents never managed to get a steady job and had a hard time making ends meet. Such situations raise a pertinent but difficult question for policymakers: which of these two people should be given priority? If one accepts that happy poor people mainly tend to remain poor and therefore deserve to be helped and that rich people who are unhappy (because they expected even more) do not need to be the top priority of the policy, one will not wish to take subjective life satisfaction as the only measure of individual well-being. There are also other problems with happiness as an indicator of well-being. For example, the level of happiness also appears to depend on personality characteristics: by and large, extroverts feel happier than more introverted people even if they are objectively in similar circumstances. We Fleurbaey and Blanchet (2013, p. 163 et seq.) discuss the literature under the sceptical title: “The Easterlin paradox: have we been wrong for 70,000 years?”.

3

Happiness

5

will return to this point in Chap. 22. But is there an alternative to happiness as an indicator for well-being?

Well-Being and Capabilities Let us summarise what we have stated so far. Income is important for individual well-being, but a one-sided emphasis on material aspects does not adequately reflect the importance of non-material aspects of life such as health, job satisfaction and good social relationships. Happiness and life satisfaction are undoubtedly important too, but cannot be regarded as the only measures of individual well-being either, because they are “too subjective” and depend on personal expectations and aspirations. This would lead to unacceptable consequences if we wanted to compare the situation of different people in order to establish priorities for the (re-)distribution policy. If we accept these two insights, we arrive at a conclusion that was already reached by Amartya Sen in his ground-breaking and influential work on functionings and capabilities.4 A good description of living standards or individual well-being requires us to look beyond income and happiness and focus on the various different dimensions of life: what people (can) do and who people are (or can be). These are what are referred to as “functionings”. Sen goes further, however, as he mainly wishes to emphasise people’s freedoms to choose their own project of life. In this context, he refers to capabilities: a person who is hungry because he wishes to fast based on his religious convictions should not be treated in the same way as a person who is hungry because he is too poor to buy food. Freedom is certainly extremely important for individual well-being. However, if we only focus on people’s capabilities or options, we can become very hard on those who make the “wrong” choices, because they are not sufficiently informed or not sufficiently aware of the longer-term consequences of their decisions. The ability to make good choices is also strongly linked to the level of education of the people concerned. In addition, it is not easy to measure capabilities properly: in reality, we can only observe the result of the choice. For all these reasons, we will not attempt to operationalise an approach based on capabilities in this book. Instead, we will focus on the outcomes that people achieve, the functionings. However, we will always examine the figures in light of the conviction that all people, including the weakest members of society, must be given the opportunity to realise their own life project. We accept Sen’s premise that many aspects of life, including non-material aspects, must be taken into account when measuring individual well-being and that income and happiness fall short as a measure of individual well-being. These different aspects will be covered in the various chapters of the second part of this 4

For example, see Sen (1985, 2009). This approach has now also become very popular outside the academic world among policymakers and action groups.

6

1

What Constitutes a Good Life?

book. However, if we were to describe and analyse these different dimensions of life separately, we would not yet be in a position to compare the individual well-being of different people. In the last two chapters of this book, we therefore look for a new measure of well-being which allows us to evaluate people’s situation in a global manner without attaching too much importance to differences in expectations and aspirations. Here, we wish to take into account the fact that not all people have the same life project: some people attach more importance to health than others, and the size of some people’s income takes precedence over the quality of their job, while for others, the reverse is true.

2

Why This Book?

The results described in this book were obtained by means of a large-scale survey across the whole of Belgium. This survey was funded by the Belgian Science Policy Office (BELSPO) and was entitled MEQIN, an acronym that stands for “Measuring Equivalent Incomes”. When you read this book, it will become clear why we chose this strange name for the study. In the MEQIN project, we wanted to gather information not only about the different dimensions of life that are important for individual well-being, but also about the importance that Belgians attach to these dimensions. As we already stated at the end of the previous chapter, we wish to take into account the fact that people can have different views on what is important in their own lives. As a result, our questionnaire was quite comprehensive and included several difficult and sometimes unconventional questions. We therefore opted for the face-to-face interview method, in which a professional interviewer visits the respondents at home.1 We collected data on 3404 adult respondents in 2098 families.2 An additional questionnaire, which we called the “drop-off”, provided us with additional information about 618 children.3 We will provide more methodological details at the end of this chapter. Firstly, we will briefly describe the structure of this book and the original contribution of our questionnaire.

The fieldwork was carried out by Kantar Belgium in the period from February to July 2016. We are aware that the social scientific literature distinguishes between the concepts of family and household. However, this is much less the case in everyday language. We have chosen to follow common parlance and use the term “family” everywhere in this book. 3 The distinction between adults and children for this survey was based on age. Adults are all family members who had reached the age of 18 at the time of the interview. 1 2

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_2

7

8

2

Why This Book?

Structure of the Book Chapter 3 provides, by way of introduction, more detailed information about the structure of families in Belgium. Subsequently, the book is divided into four parts. In the first part, we describe the living conditions of Belgians for each dimension. Although the ultimate emphasis lies on the non-material dimensions of life, in Chaps. 4 and 5, we start with a traditional analysis that focuses on material welfare: Chap. 4 describes income distribution in Belgium and Chap. 5 delves deeper into poverty. These first chapters are a useful starting point as they link to the more traditional approach that is still most popular in policy circles. We then examine some very important non-material dimensions of life: health (Chaps. 6–8) and work situation (Chap. 9). Consumption and spending patterns are discussed in Chap. 10. Chapter 11 covers the quality of housing. Housing can be regarded as part of material welfare, so this could also have been covered in the previous chapters. However, we chose not to do so for three reasons: Firstly, housing in itself is a very important dimension of life, especially for the poorest groups in our society. Secondly, housing quality depends less on the current income as it is a durable good. For many people, buying a house is a decision that will be made only once in their lives. Thirdly, we interpret housing quality in the broadest sense: in Chap. 11, we also take into account the quality of the living environment and the extent to which it contributes to people’s sense of safety. These are important non-monetary dimensions of well-being. Finally, Chap. 12 discusses in greater depth how Belgians spend their time. Where the data is available at an individual level (e.g. for health and work situation), it is also analysed at this same level in the first part. For incomes, we follow the traditional analysis at a family level. However, the limitations of this traditional analysis are clear. It is not easy to make accurate corrections for the size and composition of the family and to take into account the unequal distribution of income and expenditure within the family and the differences in views between various family members. This latter idea is further elaborated on in Part II of this book. Here, we discuss how expenditure is distributed within traditional families (couples with or without children) and who has most leisure (Chaps. 14 and 15). This distribution naturally also reflects the bargaining power of various family members, in particular who has the greatest impact on family decision-making. In Chap. 16, we examine some of the factors that determine the bargaining power of partners in couples. The bargaining power of partners is also partly determined by the way in which the couples are formed, which is discussed in Chap. 13. In itself, the way in which couples are formed also has relevant implications for the distribution of wealth in society. For example, if highly educated people with a high personal income mainly tend to live with a highly educated and rich partner, relationship formation in itself becomes a source of inequality.

Structure of the Book

9

In Part III, we focus on the situation of specific vulnerable groups: single-parent families (Chap. 17), children (Chap. 18) and the elderly (Chap. 19). In these chapters, we bring together information about various life dimensions of the people in these vulnerable groups, although without combining them into a single measure of individual well-being. We conduct this latter exercise in Part IV. In Chap. 20, we first describe how the correlation between various dimensions leads to the phenomenon of cumulative deprivation. The people with the lowest incomes often have the worst health and the least attractive housing too. The deprivation in one dimension is, as it were, accumulated and aggravated by the deprivation in other dimensions. As a result, examining these three dimensions of life separately does not allow us to paint a complete picture of the actual differences in well-being. In Chap. 21, we discuss how the happiness or life satisfaction of Belgians is determined by various aspects covered in the previous chapters. Chapter 22 delves deeper into the discussion we touched upon in Chap. 1, on the importance of aspirations and expectations. Here, we confirm the insight that looking at happiness is not necessarily the best way of comparing people’s individual well-being and of setting priorities. In Chap. 23, we look for an alternative and propose equivalent income as a measure of individual well-being. This also immediately explains the name we chose for the survey. Equivalent income is income corrected for the degree to which people score better or worse for other dimensions of well-being. The magnitude of this correction takes into account the importance that people themselves attach to these other dimensions. Chapter 24 discusses the results for this concept. We will demonstrate the importance of choosing the well-being measure for identifying the poorest members of society and argue that a one-sided focus on income or subjective happiness is highly contentious from a normative point of view.

Why a New Dataset? There is already a great deal of important and high-quality research in Belgium that provides us with insights into the distribution of aspects of individual well-being, as well as various datasets with specialised information about one or more dimensions of life. These include the Statistics on Income and Living Conditions (SILC) survey which compiles information on the living conditions of citizens in a coherent manner for all the countries in the European Union and for Iceland, Norway, Switzerland and Turkey, as well as the Survey Sociaal Culturele Verschuivingen [Survey on Sociocultural Shifts] (SCV) which has been describing the social environment in Flanders for the past 20 years. Other surveys focus more on sub-groups such as the 50-plus age group in the Survey of Health, Ageing and Retirement in Europe (SHARE), or mainly collect detailed information about specific dimensions of life such as the Gezondheidsenquête [Health Survey] of the Scientific Institute for Public Health (WIV) on health, the Budgetenquête [Budget Survey] of the Federal Public Service Economy on spending and consumption, the

10

2

Why This Book?

time-use surveys by the research group Tempus Omnia Revelat (TOR) at the Free University of Brussels (VUB) and the Nationaal Geluksonderzoek [National Happiness Study] at Ghent University. Although this list is already quite impressive, it is also incomplete and given purely by way of illustration. As a result, with regard to the description of the individual dimensions of life, the MEQIN dataset could certainly not be called original. We can add little or nothing to what is already known in each of these areas. Our only contribution in this respect is that we collect information about the main dimensions of life for the same people in a single survey and present it together. We must always bear in mind that the breadth of our survey inevitably comes at the expense of depth, i.e. the level of detail with which the information is collected for each dimension of life. We will come back to this later on in the book. We therefore regard the MEQIN dataset as a complementary tool within the already rich landscape of the existing surveys. However MEQIN does make an original contribution in two other areas. Firstly, we collected detailed information on all adult members of the family, not just with regard to their income and labour market situation. This allows us to analyse multiple dimensions of welfare at an individual level. For example, we also asked about the distribution of time and expenditure within the family. This has never been done before in Belgium, and our data therefore allows us to shed new light on the distribution of well-being and power within the families. Secondly, we gathered information on the relative importance of various dimensions for the people in our sample. To this end, we expressly asked how much income (or expenditure) they would be willing to sacrifice for better health or a more attractive job, for example. Based on this willingness to pay, we determine an original measure of individual well-being: equivalent income. We believe that identifying the most deprived people in society on the basis of these equivalent incomes makes an interesting contribution to the social debate in Belgium.

Some Preliminary Methodological Comments In this book, we wish to present our descriptive results in the most accessible possible way. Nevertheless, we will briefly discuss a few methodological points here. Readers who are not interested in this subject can simply skip this section. However, readers who are interested in the scientific foundation of our writings are likely to be eager for more after reading this section. These readers can find more information in the (deliberately scarce) references at the end of this book or on the MEQIN Website: https://sites.google.com/view/meqin/. For reasons explained in the previous paragraph, our sample aimed to collect as much information about entire families as possible. This is the only way to obtain information on the distribution within these families. The random sample was therefore selected on the basis of the family types in the Belgian National Register, geographically clustered. A reference person was identified within each family.

Some Preliminary Methodological Comments

11

Some questions, such as questions about the quality of the housing, were only asked to this reference person. However, we tried to ask all the adult members of the families most of the other questions. In some cases, they were also allowed to supplement the information provided by the reference person. In 77% of the 2098 families surveyed, we managed to get all the adult members of the family to complete the questionnaire. In order to obtain information about family members below the age of 18, an additional questionnaire (the “drop-off”) with questions about each individual child was left at the time of the survey visit. These questionnaires could be completed and returned later. 371 families provided us with information about 618 children in this way. As we mainly wanted to collect information about individuals and families in precarious living conditions, we deliberately overrepresented single-parent families and families with at least one member over the age of 60 in our sample. However, as the results in this book are based on reweighted figures, they still provide a good estimate of the living conditions and well-being of the entire Belgian population. Nonetheless, the relevant population is not the same for all chapters in this book. As a result, sub-groups were sometimes selected from the sample. For example, many of the job quality results in Chap. 9 are derived from the results of the respondents who were in work at the time of the survey, and Chap. 13 on the formation of couples only looks at the cohabiting (heterosexual) couples in the sample. In addition, some questions were not answered and these respondents could therefore not be included in the analysis. In the case of the health data, the data for 143 people could not be used as they did not give us permission to do so.4 Each chapter of this book states which sub-group of the sample the analysis is based on, without providing all the detailed quantitative information. Please consult the Website https://sites.google.com/view/meqin/ for this. Our sample of 3404 people is large enough to make reliable statements about the population as a whole. By this, we mean all people living in Belgium. As soon as we focus on sub-groups, however, the sample can become quite small. For example, we “only” have 333 respondents in Brussels (2000 in Flanders, 1071 in Wallonia) and 436 respondents who were not born in Belgium. Although we can certainly provide useful results for these groups, they need to be interpreted with greater caution. Even more striking is the fact that our sample only includes 43 people who live in a residential care home. It would therefore be very bold to make sweeping statements about this group based on our results. We have summarised the results in this book in clear figures and tables. Detailed statistical information is often not provided. Some of the results described are based on regression analysis. If this is the case, we explain it in the text. However, the detailed statistical results are not included.

4

Following the Belgian Privacy Commission’s decision, we asked the participants in our questionnaire for their permission to use this information once the health questions section of the verbal questionnaire had been processed.

3

Is the Conventional Family Still the Cornerstone of Society?

Like many Western countries, Belgium has undergone major demographic changes in recent decades. Although many people still regard a two-parent family with children as the typical family, this family type only represents a relatively small proportion of the current demographic landscape and therefore of our sample. In Chap. 1, we already discussed the fact that the type of family in which a person lives affects his or her level of well-being. Before going into the various dimensions of well-being in more detail in the following chapters, it is therefore a good idea to examine in greater depth the composition of families.

Which Family Types Are Found in Belgium? Belgium has about 11.2 million inhabitants, divided between roughly 4.8 million families. This comes to an average of 2.3 people in a Belgian family. There is naturally a wide variety of family types. This is illustrated in Fig. 3.1, which shows the distribution of families by family type in Belgium based on the MEQIN dataset. The figures show that around 1.7 million families consist of only one person. With a share of 35%, single people without children are by far the largest group. However, this group of single people is very diverse. It includes people who are not in a relationship or do not live with anyone, as well as people who have been widowed and do not have children living at home. It also includes people who have never lived with someone and individuals who are separated. Figure 3.1 also shows that just over 1 million (or about 21%) families consist of a married couple without children (living at home) and that there are about 300,000 (or around 6%) families in Belgium which consist of unmarried individuals cohabiting without children (living at home). Once again, the married and cohabiting couples without children are very diverse. This group includes both young couples who do not yet have children and older couples whose children have already left home. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_3

13

14

3 Is the Conventional Family Still the Cornerstone of Society?

1800000

35%

1600000

1400000

1200000 21%

1000000

19%

800000

600000 9%

400000

6%

6% 4%

200000

0 Single people Married Married Unmarried Unmarried Single-parent Other family couple without couple with couple without couple with family type children children children children

Fig. 3.1 Distribution of families between different family types

As we already mentioned, many people regard a couple with children as the standard family type. Figure 3.1 shows, however, that only a quarter of families satisfy this criterion. There are indeed roughly 900,000 (or 19%) married couples with children living at home and 280,000 (or just under 6%) cohabiting couples with children living at home. There are also major variations within these groups. Most of these families are indeed couples with their own children. However, there are also newly formed families in which the children do not all come from the same parents. Our country also has around 420,000 (or just under 9%) single parents. These parents may be divorced with full custody of their children or partly co-parenting, or they may deliberately be single parents. Single-parent families are often regarded as a group facing a higher than average risk of poverty. We therefore examine their level of well-being in more detail in Chap. 17. Finally, Belgium has just under 200,000 (or 4%) families that do not correspond to the categories described so far. For example, these families might consist of a couple living with one of the partner’s parents or two relatives living together. However, other compositions are also possible.

Which Family Types Are Found in Belgium?

15

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0 Single people

Married couple without children

Married couple with children

Unmarried Unmarried Single-parent Other family couple couple without family type with children children

Fig. 3.2 Average number of people per family type

Figure 3.2 provides more insight into the size of families. It goes without saying that the single-family type consists of only one person, and married or unmarried couples without children (living at home) consist of two people. More relevant information was obtained for the other family types. For example, a married couple with children living at home consists of an average of four people, suggesting that such couples have an average of two children living at home. On average, unmarried couples with children living at home have a somewhat smaller family size, in this case 3.8 people. Unmarried couples therefore have fewer children than married couples, on average. Single parents have an average of 1.5 children in their care.

Are There Differences Between the Regions? The family composition varies greatly from region to region. Figure 3.3 immediately shows that the percentage of single people in the Brussels-Capital Region is much higher than in the other two regions. In Brussels, almost one in two families

16

3 Is the Conventional Family Still the Cornerstone of Society?

60%

50%

40%

30%

20%

10%

0% Single people

Married couple without children

Married couple with children

Brussels-Capital Region

Unmarried couple without children

Unmarried Single-parent Other family family type couple with children

Flemish Region

Walloon Region

Fig. 3.3 Regional distribution of families between different family types

consists of a single person. In Flanders and Wallonia, the figures are 31% and 37%, respectively. The specific results for Brussels are partly explained by the fact that this is an urban population. Cities such as Antwerp or Liège also have a relatively high proportion of single people. Major variations are also found across the different regions with regard to the other family types. In Flanders, for example, around 25% of families consist of a married couple without children (living at home), while this type of family only accounts for 10% in Brussels. There are relatively more single people and single-parent families in Wallonia than in Flanders. This certainly also has implications for the distribution of individual well-being in each region.

Age and Family Type In Fig. 3.4, we examine the distribution of adults of different ages between the various family types. This figure shows that only 9% of people aged 18–30 live alone, while 38% of people aged over 70 live alone. In general, we note that older people are more likely to live alone or in a married couple without children. When

Age and Family Type

17

60% 50% 40% 30% 20% 10% 0%

Below 30 Between 50 and 60

Between 30 and 40 Between 60 and 70

Between 40 and 50 Over 70

Fig. 3.4 Distribution of adults of different ages between the various family types

we compare the distribution of adults between married couples and unmarried couples, we find that older people are less likely to live together without being married than young people. People in their thirties are more likely to form an unmarried couple with children than all the other age groups. From this chapter, we can conclude that the family landscape in Belgium is very diverse. It could therefore be misleading to mainly focus on one family type consisting of a couple with children. Not only are there many different family types, their relative importance also varies greatly across regions and age groups. We must therefore take this information into account when interpreting the data in the following chapters.

Part I

Well-being Involves Many Different Aspects

We will begin our analysis with an overview of the individual results for various important aspects of well-being. Firstly, we examine the traditional material measures of income inequality and poverty. We then describe the health of Belgians, including the pressure on their budget exerted by healthcare costs, the quality of their employment situation and the quality of their housing. Finally, we look at how they allocate their budget across different types of expenditure and how they spend their available time.

4

How Unequally Are Our Incomes Distributed?

Material welfare can be viewed from various different perspectives. Firstly, we can focus on disposable incomes. Disposable income is defined as the monthly net income from the work of all the family members together with benefits, transfers and pensions, as well as income from capital and investments. The greater the disposable income, the more material welfare can be acquired. We can also look at expenditure, i.e. the total amount of money spent each month on goods and services such as food, housing, clothing and transport. This expenditure represents the quantity of goods and services consumed: the higher the level of consumption, the greater the material welfare. Income and expenditure are related, but need not necessarily be equal. If a family spends less than its income, the family members can save at the end of the month and increase their wealth as a result. However, if they spend more than their income, they will need to borrow money to make ends meet at the end of the month or start “dissaving”, i.e. consume part of their assets. These assets, i.e. the value of all the possessions that people have accumulated over the years, can also be regarded as an indicator of material welfare. However, the MEQIN data does not allow us to make reliable statements about the assets of Belgians. We will discuss the composition of this expenditure further in Chap. 10. In this chapter, we focus on the distribution of disposable incomes. This is also the most common approach to studying economic inequality and poverty. Poverty will be examined in more detail in the next chapter. In this chapter, we take a closer look at income inequality. It does not make much sense to examine the personal income of each Belgian individually. Indeed, generally speaking, members of a family combine their income or at least part of it. Of course, this does not necessarily mean that the family’s welfare is equally distributed among all the family members. In most cases, however, the information required to make meaningful statements about this distribution within the family is lacking. As we already emphasised in Chap. 2, one of the original contributions of the MEQIN survey is that we did collect data on this aspect. We will return to this in detail in Part II of this book. In this chapter, we © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_4

21

22

4

How Unequally Are Our Incomes Distributed?

deliberately stick to the more traditional approach of looking at inequality between families, without distinguishing between various family members. Here too, of course, it is important to correct for the size of the family. It is easier to get by on a specific income with a small family than a large family, and the previous chapter has already revealed the extent of the diversity of family types in Belgium. We therefore use an equivalence scale to correct the disposable family incomes for family size. We use the standard modified OECD equivalence scale, also used in many other studies and in official documents. This equivalence scale assumes that 1 euro in the wallet of a single person produces as much material welfare as 1.5 euros in the wallet of a couple. Indeed, couples do not have to buy everything in duplicate and can share some goods such as a bathroom, central heating or the television set while still acquiring as much material welfare as a single person. More generally, the OECD equivalence scale is calculated by assuming that for every euro a family needs for the first adult, it needs an additional 0.5 euro for each member of the family aged 14 years or above and 0.3 euro for each child in the family below the age of 14. The corrected income obtained using this method is known as the equivalised income. Some caution should be exercised when using sample data to analyse income distribution. After all, we know that it is difficult for researchers to reach very poor Belgians, homeless people and refugees. In addition, rich citizens often prefer not to participate in research that charts their incomes in detail. In addition, generally speaking, it is very difficult to determine the exact income of the self-employed. These problems arise in every survey that aims to chart incomes and therefore also in our MEQIN dataset. We will start this chapter by presenting the distribution of incomes in Belgium. For this, we use what is called the “parade of dwarves and giants”. We then focus on income inequality by means of what is known as the Lorenz curve.

A Parade of Dwarves and Giants In the 1970s, the Dutch economist Jan Pen came up with an appealing way of illustrating income distribution. He represented it in the form of a parade in which the entire population parades in one hour. In the parade, people are ranked according to their income: the poor pass by first, then the middle class and finally the rich. The height of the participants in the parade is determined by their income. People with an average income are of average height (1.74 m). People on lower incomes are shrunk in proportion to their incomes, so poor people are no bigger than dwarves. People on higher incomes are magnified and can therefore take on huge proportions. Figure 4.1 shows this parade for Belgium. The horizontal axis of the figure shows the time. For each minute of the parade, we represent the average height of the passers-by with the height of a grey bar. On the left-hand vertical axis, we show the height in the parade (in metres), while the right-hand axis displays the

A Parade of Dwarves and Giants

23

8 8000 7500

7

7000

Height (in metres)

6000 5500

5

5000 4500

4

4000 3500

3

3000 2500

2

Incomes (euros per month)

6500

6

2000 1500

1

1000 500

0

0 0

6

12

18

24

30

36

42

48

54

60

Time (in minutes)

Fig. 4.1 Parade of dwarves and giants

corresponding equivalised income (in euros per month). The parade clearly consists of dwarves and giants. The first marchers are no more than short dwarves: the average height of the marchers in the first minute is less than half a metre. After 12 min, the marchers are about one metre tall, which corresponds to an income of around 1000 euros per month. Halfway, after 30 min, the average marcher has an income of 1620 euros per month. We have to wait a little longer until the marchers reach average height. Only after 36 minutes do we see a participant in the parade with an average height of 1.74 m. After this, the height of the participants continues to increase steadily. By the end of the parade, there are even a few giants walking along. The average height of the giants in the last minute of the parade is almost 7 metres. If all Belgians had an equivalised income equal to the average equivalised income from the dataset (about 1800 euros, represented by the horizontal black line), everyone would be the same height in the parade. The differences in height between the parade participants reflect the income inequality. As indicated earlier, these are the disposable incomes after the government has redistributed through taxes and social security. Without this redistribution, the inequality in the parade would be even greater.

24

4

How Unequally Are Our Incomes Distributed?

Fig. 4.2 Share of total income of ten equally large groups

4.3%

5.3%

22.1%

6.8% 7.2% 8.3%

13.8%

9.5% 12.0% 10.6%

A Spoke in the Wheel? The parade provides an intuitive insight into the order of magnitude of inequality in Belgium. However, we need a more accurate measuring instrument if we are to measure inequality accurately. We will start our analysis of Belgian inequality on the basis of Fig. 4.2. To create this figure, we divide the participants in the parade of dwarves and giants into ten equally large groups (“deciles”). The dwarves who marched during the first six minutes of the parade form the first group, the participants who marched between the sixth and twelfth minutes form the second group, and so on. As we ranked the parade participants according to their income, each group is richer than the previous one. The last group contains the giants and is the richest group. The pie chart in Fig. 4.2 shows the share of total income for each group. The paler the colour of the piece of the pie, the later the group marched in the parade and therefore the richer the participants are. If incomes were evenly distributed in Belgium, the share of each group would be exactly the same size and Fig. 4.2 would resemble a normal bicycle wheel, with spokes at equal distances. However, we can see that this is not the case: some spokes are much closer together and others are much further apart. The share of the poorest group is about 4%, whereas the share of the richest group is over 20%, even though each group contains exactly the same number of people. If we add up the shares of the first five groups, we find that the poorest half of the Belgian population only has slightly less than a third of the total disposable income.

Inequality Measured

25

Inequality Measured We will use the information from the pie chart in Fig. 4.2 to measure inequality accurately, presenting it in a slightly different way. We use what is termed the “Lorenz curve”, introduced in Fig. 4.3. To create this figure, we use the same ten ranked groups as in the previous figure. For each group, the Lorenz curve shows the aggregate share of all the previous groups in the total income. We start with the poorest group’s share of the total income, then we look at the combined share of the two poorest groups and so on. At the very end, we look at the share of all the groups together. If Belgium had an equal income distribution, each group would have a share of exactly the same size and the aggregate share in the total would increase uniformly, as shown by the oblique, solid grey line in Fig. 4.3. The further the Lorenz curve lies from this oblique line of equality, the more unequal the income distribution. Once we have drawn the Lorenz curve, we can calculate the magnitude of the inequality by measuring the area between the oblique line of equality and the Lorenz curve. The more equal the distribution, the closer the Lorenz curve is to the oblique line and the smaller this area will be. As the Lorenz curve moves further away from the oblique line, the area becomes larger. The area between the oblique line of equality and the Lorenz curve determines the most famous of all inequality 1 0.9

Aggregate share of the total income

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Aggregate share of the total population

Fig. 4.3 Lorenz curve

0.9

1

26 Table 4.1 Inequality measured

4

How Unequally Are Our Incomes Distributed? Income

Gini coefficient m60/m1 m60/m31 m31/m1

26.8 14.7 4.2 3.5

measures: the Gini coefficient. With perfectly equal distribution, the Gini coefficient would be 0, whereas it reaches a value of 100 in case of maximum inequality. Although many other measures of inequality naturally exist, the Gini coefficient has been the faithful companion of inequality analysis for over 100 years. We show the Gini coefficient for the MEQIN data at the top of Table 4.1. The value of the coefficient is 26.8. The parade of giants and dwarves demonstrates something that many people intuitively feel, namely that incomes in Belgium are unequally distributed. Nonetheless, international comparative research within the OECD shows us that Belgian income inequality is still fairly limited, especially if we compare ourselves with countries such as the UK or the USA, where Gini coefficients rise to 35 and even 40.1 To gain even more insights into this finding, we will also present some other inequality measures. These measures are derived directly from the parade of dwarves and giants and reflect the ratio between the average heights of the participants at two moments in the parade. For example, the measure m60/m1 shows how much bigger the giants in the last (60th) minute are compared to the dwarves in the first minute. We can see from the table that the giants are almost 15 times the size of the dwarves in the first minute. If we look at the measure m60/m31, which compares the height of the giants in the last minute with the height of “Average Joe” halfway through the parade, we note that the ratio is about 4 while the measure m31/m1 reaches 3.5. This means there is a little more inequality at the end of the parade than at the beginning.

For example, see the OECD report “In It Together: Why Less Inequality Benefits All”, published in 2015.

1

5

Who Is Poor in Our Society?

Many people believe that everyone in society should be able to achieve a minimum level of material welfare. People whose material welfare is below this minimum level are regarded as poor. Two standard measures are used to measure poverty in our society: the risk of poverty and material deprivation. The risk of poverty measures income poverty, i.e. the percentage of the population living in a family whose disposable income is below the poverty line. Material deprivation offers a complementary perspective on poverty and measures the percentage of the population with no access to certain goods or services that most people consider necessary to achieve a minimum level of material welfare. In this chapter, we examine how many poor people there are according to these two standards, together with their socio-demographic characteristics. We also look at the overlap between income poverty and material deprivation. However, in this chapter we will continue to focus on the material aspects of well-being. Some non-material aspects will be discussed in the following chapters.

Risk of Poverty To measure the risk of poverty, we compare the disposable income of each family with the poverty line. As in the previous chapter, once again we use the modified OECD equivalence scale to take family composition into account when measuring disposable income. The poverty line in Belgium is set at 60% of the median disposable income. The median disposable income is the amount of disposable income 50% of Belgians have less than and the other 50% more than. In the previous chapter’s parade, this would be the income of the people who pass by right in the middle of the parade, after 30 min. Our research shows that the median disposable income in 2016 was 1620 euros per month, making the poverty line 972 euros per month. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_5

27

28

5

Who Is Poor in Our Society?

Table 5.1 Percentage of people living below the poverty line in previous SILC surveys and MEQIN 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 SILC 14.7 15.2 14.7 14.6 14.6 15.3 15.3 15.1 15.5 14.9 15.5 MEQIN 14.5

According to our study, 14.5% of the Belgian population lived below the poverty line in 2016. This figure is comparable with the figures from the SILC survey. This survey is conducted annually and forms the main source of information for EU poverty policy. Table 5.1 shows the percentage of people living below the poverty line in previous SILC surveys. The table shows that there has been a surprisingly constant rate of about 15% poor people in Belgium for over 10 years. To further refine the analysis, we will divide the Belgian population into four groups. Firstly, we split the people living below the poverty line into two groups. By poor, from now on we are referring to people who live on a disposable income which is between 60 and 50% of the median income. The very poor are the people who live on a disposable income of less than 50% of the median income. We refer to the people who live on an income of between 60 and 70% of the median income as vulnerable. We regard all other people as non-poor. According to this classification, in Belgium there were 6.6% very poor people, 8.0% poor people (the sum of the two yields the 14.5% we mentioned previously) and 7.7% vulnerable people. The remaining 77.8% are non-poor. Table 5.2 shows some socio-demographic characteristics of the people in these four groups. Here and in the rest of this chapter, we will limit ourselves to the adults in our sample for whom we have income data. The first striking finding relates to the gender of the poor. While women make up half the adult population of Belgium, they account for remarkably more than half of the poor and vulnerable adults. Moreover, people who are not in a relationship and migrants appear to be over-represented in the group of very poor Belgians.1 Another notable characteristic of the poor population in Belgium is their level of education. While the low-skilled account for one-third of the total adult population, they represent between 47.3 and 61.7% of the very poor and poor population (we regard people as low-skilled if they do not at least have a certificate of secondary education). We refer to a person who is not working and is actively looking for work as unemployed. The unemployed are clearly over-represented among the poor and very poor. Finally, unlike a few years ago, pensioners are no longer over-represented in the very poor population. By pensioners, we mean people who are retired, including people who have taken early retirement but not those on an early leavers scheme. Although they account for 28.4% of the total population, they only represent 13.6% of the very poor group. At the same time, we note that In this book, the term “migrants” refers to people who were not born in Belgium.

1

Risk of Poverty

29

Table 5.2 Socio-demographic characteristics of the adults in the four groups classified by level of income poverty Female In a relationship Migrant Low-skilled Unemployed Retired

Belgium

Very poor

Poor

Vulnerable

Non-poor

50.4 69.1 13.2 33.3 6.5 28.4

54.8 62.2 39.7 47.3 36.7 13.6

52.7 67.9 32.3 61.7 24.9 24.5

57.4 54.5 15.9 48.3 7.3 35.4

49.3 71.0 9.5 28.5 2.7 29.2

pensioners are over-represented in the vulnerable group (35.4%). In Chap. 19, we will discuss the well-being of elderly Belgians in greater depth.

Material Deprivation In Chap. 1, we already mentioned why disposable income is often used to measure poverty. After all, disposable income enables people to buy the goods and services that are considered necessary to function in society: food, clothing, housing, transport, cultural goods and so on. A more direct way of measuring poverty is to examine the extent to which people have access to certain necessary goods or services. In poverty research, it is customary to focus on the following nine goods and services when doing so: (1) the ability to pay rent or a mortgage, (2) the ability to keep the home adequately warm in winter, (3) the ability to pay an unexpected expense of 1000 euros, (4) the ability to buy meat, fish or a vegetarian equivalent for the whole family at least every two days, (5) the ability to go on a week holiday with the whole family at least once a year, (6) being in possession of a colour TV, (7) of a washing machine, (8) of a car and (9) of a telephone (landline or mobile). In order to facilitate the comparison with the previous paragraph, once again we divide the Belgian population into four groups. We say that a person is severely materially deprived if he or she cannot pay for at least four of these goods or services. If a person is unable to pay for three of these goods and services, we regard him or her as materially deprived, and we refer to this person as vulnerable if he or she cannot pay for two of these goods or services. The Belgian population can then be broken down as follows: 77.4% of Belgians are not in a vulnerable position or materially deprived, 10.7% are vulnerable, 5.7% are materially deprived and 6.2% are severely materially deprived. For these four different population groups, Table 5.3 shows the percentage of people who are unable to purchase each of these nine goods or services. We note, for example, that 29.2% of the total population are unable to pay an unexpected expense of 1000 euros, while no less than 98.1% of people who are severely materially deprived do not have this capacity. We also see that material deprivation

30

5

Who Is Poor in Our Society?

Table 5.3 Percentage of people who are unable to pay for necessary goods and services in the four groups, broken down by level of material deprivation Belgium Severely deprived 1 2 3 4 5 6 7 8 9

Rent or mortgage Heating Expense of 1000 euros Meat Holiday Television Washing machine Car Telephone

13.8 3.0 29.2 2.4 20.4 0.6 3.0 6.0 1.4

66.8 36.0 98.1 31.8 97.1 4.0 38.2 53.9 16.4

Deprived Vulnerable Not deprived 69.9 3.3 98.7 3.8 79.3 4.4 6.8 31.9 1.9

30.2 4.4 86.7 1.0 69.6 0.4 0.0 5.4 2.3

3.2 0.1 10.5 0.1 3.1 0.1 0.3 0.4 0.0

mainly takes the form of difficulty in paying the rent (69.9% of deprived people and 66.8% of severely deprived people) and the inability to buy a car (31.9% of deprived people and 53.9% of severely deprived people). We also note that buying meat, fish or a vegetarian alternative at least every two days is a problem for 31.8% of severely deprived people, whereas this problem only affects 2.4% of the Belgian population as a whole. As in the previous section, we can now also study the socio-demographic characteristics of the adults in our sample who are materially deprived. The same socio-demographic characteristics that stood out with regard to the risk of poverty are also reflected in material deprivation. In general, the conclusions are similar to those for the analysis of income poverty. For this reason, we will not go into this in more detail here.

How Many Belgians Are Both on a Low Income and Materially Deprived? As part of the Europe 2020 strategy, the European Union aims to reduce the total European population at significant risk of poverty or social exclusion by 20 million people. Here, it is assumed that people are at significant risk of poverty or social exclusion when faced with at least one of the following three problems: (1) income poor, (2) materially deprived, or (3) a very low work intensity (these are the people living in a family in which the people of working age have worked less than 20% of a full-time equivalent over the past year). In the first sections of this chapter, we examined income poverty and material deprivation separately. We will now look at the extent to which these two measures identify the same people, i.e. how many people are both low income and materially deprived at the same time. We will examine employment more closely in Chap. 9.

How Many Belgians Are Both on a Low Income …

31

Table 5.4 Income poverty in terms of income (rows) and material deprivation (columns) Severely deprived Very poor Poor Vulnerable Non-poor Total

1.9 2.1 1.3 0.9 6.2

Deprived 1.2 1.3 0.7 2.6 5.7

Vulnerable 0.8 2.1 1.4 6.5 10.7

Not deprived 2.7 2.5 4.3 67.8 77.4

Total 6.6 8.0 7.7 77.8 100.0

In Table 5.4, we resume our breakdown of the total population into four groups based on income poverty (rows) and material deprivation (columns). For example, we can see that 67.8% of the population are not poor or deprived according to either of these two perspectives. However, it is also clear that 1.9% of the population are both very income poor and severely deprived. At the same time, they are among the 6.6% very poor according to disposable income and the 6.2% in a state of severe deprivation. Table 5.4 shows that while there is an overlap between the two poverty measures, this overlap is far from perfect. There are several reasons why someone could be income poor but not materially deprived, or vice versa. Someone who is temporarily income poor but has been able to save or buy essential goods in the past is not in the same situation as someone who has the same income but without the savings or essential goods. In addition, although two people may have the same level of income, one of them may have many more needs. He or she might be ill, for example, incurring high healthcare costs as a result. This could lead to material deprivation for the sick person, even if their disposable income is above the poverty line. It is important to study poverty risk and material deprivation in order to focus social policy on the least privileged in our society. Nonetheless, we can identify three key reasons why it can be dangerous to look, in the public debate, exclusively at the poverty measures in this chapter. The first reason is that it is very difficult to determine when a family with a certain income level and family composition achieves the same material level of well-being as another family with a different income level and family composition. In this and the previous chapter, we used the modified OECD equivalence scale to correct for differences in family composition. This correction is used in the official poverty statistics for Belgium and Europe. However, the scientific basis for the use of a specific equivalence scale is rather weak and the results also greatly depend on the method chosen. For example, if we used the Oxford equivalence scale,2 which allocates a smaller benefit to larger families, the proportion of Belgians living below the poverty line in 2016 would be 15.6% instead of 14.5%. The composition of the According to the Oxford equivalence scale, the first adult counts for 1, all other members of the family aged 14 years or above count for 0.7 and all children below the age of 14 count for 0.5.

2

32

5

Who Is Poor in Our Society?

population living below the poverty line will naturally vary too, with fewer single people and more large families, also causing child poverty to increase. This seemingly purely methodological change in equivalence scale would therefore bring about a change in social policy, with greater attention being paid to larger families. The point we wish to make is not that this policy choice is undesirable or bad, but merely that its justification has a rather flimsy basis. It would be better for the policy to directly address children’s actual living conditions rather than making it dependent on the somewhat obscure methodological choice of a specific equivalence scale. We will examine children’s living conditions in more depth in Chap. 18. The second reason why poverty measures—as studied in this chapter—are not ideal for guiding and evaluating social policy is because they are based on the assumption that material welfare is equally distributed within the family. By definition, they assume that all family members will have the same level of material welfare. This is naturally an arguable assumption. In order to find out how well-being is actually distributed within the families, we need to go further and look inside the black box of the family, as it were. Our data makes this possible, as will become evident in Part II of this book. The third reason for going beyond the poverty measures stated in this chapter is that non-material dimensions such as employment and health remain under the radar. After all, the loss of a job will have a much greater impact on well-being than simply reducing the disposable income. A job is also a source of self-esteem and satisfaction and offers a social network. We will discuss jobs and employment in Chap. 9. Not only is health essential for many people’s well-being, it also has a significant impact on other dimensions such as the ability to earn an income. We will study health in the following chapters of this book. In the last part of this book, we then look for a measure of well-being that brings together the various different dimensions in a coherent way.

6

How Healthy Are We?

Health is certainly a major concern for many people. “As long as you’ve got your health…” is a common expression in many regions of Belgium. In this chapter, we first describe what we mean by “healthy” or “sick”. We then investigate who is healthier in Belgium and who is sicker. The results of our survey largely confirm what has already been found in other studies.1

What Is Health? At first glance, it might seem easy to determine whether someone is sick or healthy. However, this is a misconception. Health involves many different facets that cannot easily be brought under a single denominator. Our questionnaire included around twenty questions on health. In this chapter, we describe how the answers to these questions can be summarised in five dimensions of health. This immediately provides an idea of how healthy (or sick) Belgians are.

General Health The first series of questions concerns general health. A few examples: How healthy would you say you are in general? Are you as healthy as other people? Do you get sick more often than other people? Do you think that your health will deteriorate

1

The main source of health data is the Gezondheidsenquête [Health Survey] which has been organised on a regular basis since 1997 (see https://his.wiv-isp.be/nl/SitePages/Introductiepagina. aspx). The 2013 Health Survey is the closest reference point for comparisons with our results.

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_6

33

34

6 How Healthy Are We?

over the next few years? We added up the answers to these questions and incorporated them into a single index: people with a score of 100 gave the “healthiest” answers to all the questions and a score of 0 was awarded if the “sickest” answers were given. The results are summarised in the first row of Table 6.1. Many Belgians have health problems: almost 60% have an index value of less than or equal to 70. The answers to the specific questions provide additional information.2 Almost one in five Belgians feel that their health has deteriorated since the previous year; more than 9% feel that they get sick more often than other people and just over a quarter think it would be incorrect to say that they were in excellent health. Of course, this does not mean that all these people are seriously ill. Instead, the questions about general health provide an overview of the subjective perception of the people concerned. Relatively limited complaints can already have a negative effect on subjective feelings, especially if people compare themselves with others. As an objective measure of health, these questions must therefore be put into perspective.3 However, the subjective feeling of health is undoubtedly of great importance for the perception of individual well-being.

Functional Limitations The questions about possible functional limitations that people experience in their daily lives are far less subjective: Are they capable of significant physical exertion? Can they walk up the stairs without a problem? Can they walk a few hundred metres? Can they wash and dress themselves, or do they need help with this? The second row of Table 6.1 shows that most Belgians do not have much trouble with these activities. All the same, there are quite a few people with limitations. Only slightly less than 50% of Belgians feel that they have no trouble whatsoever with significant physical exertion and between 20 and 25% have trouble with moderate exertion (such as climbing stairs or lifting groceries); 32% of Belgians say they have some difficulty bending, kneeling or squatting. Approximately one-fifth of Belgians experienced difficulties in carrying out their work during the four weeks prior to the survey as a result of (physical or emotional) health problems, or were limited in the type of work they could carry out. In about 30% (not necessarily different people), health problems caused difficulties with social interaction. Of course, many of these problems are related to age. We will return to this aspect later in this chapter.

2

This detailed information is not included in the table. The same applies to the separate results for the remaining health dimensions. 3 There is a comprehensive literature on the advantages and disadvantages of subjective health evaluations. In general, subjective measures—although far from perfect—are considered to have a predictive value for more objective conditions, even the chance of dying (e.g. see Idler and Benyamini 1997).

What Is Health?

35

Table 6.1 Health of Belgians (% of adult population) Score General health Functional limitations Chronic diseases Emotional well-being Physical well-being

0–10 (%)

11–20 (%)

21–30 (%)

31–40 (%)

41–50 (%)

51–60 (%)

61–70 (%)

71–80 (%)

81–90 (%)

91–100 (%)

1.6

2.2

4.2

6.4

9.7

14.7

19.3

21.6

13.9

6.4

2.8

1.7

2.0

2.0

3.7

2.8

4.7

8.5

15.3

56.4

12.2

14.8

10.5

11.1

51.4

0.4

1.2

2.2

4.7

5.0

7.3

14.8

28.0

29.3

7.2

4.9

6.7

2.9

3.6

3.0

2.4

2.9

3.7

15.0

54.8

Chronic Diseases We asked the respondents directly if they were suffering from a protracted illness, chronic condition or disability: 36% responded in the positive. However, the identification of the chronic diseases they are suffering from is very unreliable: for example, only 9% are aware of (or report) excessively high blood pressure and just over 4% report diabetes. We know from many other sources that these are both substantial underestimates and that a significantly higher proportion of people suffer from these conditions. We also asked if people had been limited in their activities during the last 6 months or more due to a health problem: 14% reported that they had been “severely limited” and 26% reported that they had been “slightly limited”. The global scores for “chronic diseases” are summarised in the third row of Table 6.1. This row looks different because the score for this dimension can only assume five different values (0, 25, 50, 75 and 100). People who did not report a chronic problem and were not restricted in their activities during the previous six months due to a health problem were awarded the value 100. People who did report a chronic problem and were “severely limited” during the previous six months received a score of 0. This concerns no less than 12.2% of Belgian adults.

Emotional Well-Being When examining the state of health, we must certainly not forget emotional well-being. As summarised in the fourth row of Table 6.1, emotional problems are widespread among the Belgian population: 35% have an index value below or equal to 70, and the average score (out of 100) is only 72. Several specific factors play a role here: 12% of respondents rarely or never felt joyful during the past four weeks, 17% rarely or never felt full of energy, 12% always or mostly felt anxious and 9%

36

6 How Healthy Are We?

always, mostly or often felt depressed. No less than 7% of Belgians always, mostly or often felt very depressed during the past four weeks. These results are not surprising, but they are still striking. There is obviously a very direct relationship with the individual perception of well-being here. Researchers who equate individual well-being with feelings of happiness regard this emotional well-being as by far the most important dimension (e.g. Layard 2005). According to this approach, any policy which aims to improve well-being in society should pay particular attention to mental healthcare. We have already argued in the first chapter that this is not our approach, as we think there are other important life dimensions besides emotional well-being. However, significant importance is also attached to mental suffering in our broader approach to well-being. Recently, some sources have claimed that emotional well-being in society has fallen sharply in recent years and that this is due to increasing inequality and stress in the work situation. We cannot make statements about the evolution over time based on our data. In the next chapter, however, we will examine whether there is a correlation between emotional well-being and objective environmental and economic factors.

Physical Well-Being Our fifth dimension brings together the answers to questions about physical well-being and pain. The overall score can be found in the fifth row of Table 6.1. Not surprisingly, the distribution is very similar to that for the second dimension (functional limitations). Once again, the generally rather positive picture must not make us forget that some people do indeed have to live with pain: 10% of people say they are suffering from severe or very severe pain and 9% (for the most part the same people) were impeded from their normal activities (both away from home and at home) by pain during the previous four weeks. Only 36% say they have no pain at all.

Who Is Ill? The general distributions that we discussed in the previous paragraph may not reveal all that much. Although they demonstrate that health problems are widespread in society, the question as to who is sicker or healthier is more interesting. Can we describe which groups in our society are particularly affected by health problems?

Who Is Ill?

37

Are Poorer People Sicker? A first question is whether poorer people actually do have higher sickness rates. Figure 6.1 shows the average values of the health dimensions for four equally large income groups (“quartiles”). The first income group contains the 25% poorest people (i.e. the passers-by from the first 15 min in the parade of dwarves and giants described in Chap. 4).4 The subsequent income groups each contain the next 25% of people according to their income. There is a clear positive link between health and income, and this link is stronger for the more objective dimensions. People on higher incomes are healthier and live longer. This is perhaps the clearest example of what we refer to as “cumulative deprivation”: for the weaker groups in society, the less favourable results for different dimensions of life accumulate. We cover the phenomenon of cumulative deprivation extensively in Chap. 20. We should make a few comments regarding the result in Fig. 6.1. Firstly, the figures reflect two different effects. Not only does poverty make people sick, sickness also makes people poor. Sick people often suffer an immediate loss of income, for example if the illness results in them having to give up their jobs. This will be discussed further in Chap. 8. Even if they have a job, less healthy people will often be less productive and thus earn lower wages. Secondly, there is also a strong link between health differences and the level of education. This is illustrated in Fig. 6.2. Below, we refer to people as low-skilled if they do not at least have a certificate of secondary education. People are regarded as highly educated as soon as they have a degree of higher education. If a person is neither low-skilled nor highly educated, we refer to him or her as having a medium level of education. On average, more highly educated people are healthier. Income and education levels are strongly linked in our society, and it is not always easy to determine their relative importance for health. From a policy viewpoint, however, this is a relevant question. If income in itself has a strong impact, income support can improve the health of poor people. If health differences primarily reflect differences in education and knowledge, however, it is this area that should be addressed. Of course, these two comments do not detract from the importance of cumulative deprivation. They merely illustrate that this is a complex phenomenon in which various factors play a role. Equally, it cannot be inferred from our findings that inequality makes people sick: this assertion implies that inequality would also have a negative impact on the health of the rich and highly educated.5 We are unable to make any statements about this on the basis of our data.

4

Here, we have used the same income concept as in Chaps. 4 and 5: the net disposable income corrected for family size by means of an equivalence scale. 5 Although this idea has recently become popular through the work of Wilkinson and Pickett (2010), it is certainly not always confirmed by in-depth scientific research (O’Donnell et al. 2015).

38

6 How Healthy Are We? 87 78

91

81

84 72

67 69 65 61 63

General health

61

82 84

77

65

69

FuncƟonal limitaƟons Chronic diseases QuarƟle 1

QuarƟle 2

QuarƟle 3

67

75 72 74 72

72

76

77

EmoƟonal well-being Physical well-being QuarƟle 4

General

Fig. 6.1 Health differences in accordance with four income groups

86

89 84

83

76 66

69

75 69

65

69

78

77

71

61

60

General health

69

72 74 72

FuncƟonal limitaƟons Low

Chronic diseases Medium

High

EmoƟonal well-being Physical well-being General

Fig. 6.2 Health differences by level of education

How Sick Are the Elderly? As we get older, our health deteriorates. This general truth is also confirmed in Fig. 6.3. The result is not surprising. Indeed, if we had found the opposite, it is likely that no one would believe our results. Incidentally, the age-related objective decline in health is probably even greater than can be deduced from the answers to

Who Is Ill?

39 100 90 80 General health 70

Funconal limitaons

60

Emoonal well-being

Chronic diseases

Physical well-being 50 40 18-29 30-39 40-49 50-59 60-69 70-79

80+

Age category

Fig. 6.3 Health by age group

our questionnaire. After all, as has already been said, these are subjective assessments of health by the people concerned. When answering our questions, people are likely to take into account the fact that they are getting older. An 80-year-old would consider it normal to be unable to run a mile, whereas this would be worrying for someone aged 20. Figure 6.3 somewhat confirms this hypothesis. The effect of age is strongest for the more objective dimensions (limitations in daily life and chronic diseases). Moreover, we should not forget that we work with sample data provided by actual living respondents. An obvious condition for taking part in our survey is the ability to answer the questionnaire. We clearly cannot survey very sick (or deceased) people. The older the group we are looking at, the more selective this condition becomes. Figure 6.3 provides a few other interesting insights. Firstly, for all health dimensions (except emotional well-being) there seems to be a decline until around the age of 50, when people reach a kind of plateau. Between the ages of fifty and sixty-five, their health status remains virtually the same. It starts to decline again from about 70 years of age. The implications of these results are discussed further in Chap. 19, when we examine the situation of the elderly in more detail. Secondly, it is interesting to look at the results for emotional well-being. After all, the age effect is much less apparent here. People are more or less subject to depression at different stages of their lives, but certainly do not necessarily have more emotional problems as they get older. On the contrary, between the ages of 60 and 70 they seem to feel quite well. The “young old age” does not seem to be a bad time in life for many people.

40

6 How Healthy Are We? 87 85 68

85

82

76

64 66

General health

FuncƟonal limitaƟons

70

65

69

77 76

75 67

Chronic diseases EmoƟonal well- Physical wellbeing being

Brussels-Capital Region

Flemish Region

Walloon Region

Fig. 6.4 Health differences by region

Are the Flemish Healthier? Economically speaking, on average the Flemish outperform the Walloons. Given the strong link between economic situation and health, we might also expect them to be healthier. However, as shown in Fig. 6.4, this is not really reflected in our figures. In general, there are few differences except for emotional well-being: on average, this is significantly lower in Wallonia than in Flanders. At first glance, the better results for Brussels may be surprising. They can be partly explained by the fact that, on average, people in Brussels are much younger than the inhabitants of the other regions. If we broaden our horizons, we can also investigate whether there is a link between health differences and country of birth. The results are summarised in Fig. 6.5. There are few differences between people born in Belgium and those born in southern European countries. People who were born in Morocco or Turkey are less healthy than people born in Belgium. However, it is the good health of those born in Eastern Europe that stands out most. Incidentally, these people also do remarkably well on the labour market and we return to this in Chap. 9.6 Although

This result is also confirmed in other research; for example, see the figures in the report by MYRIA, Migration in figures and rights 2017 (https://www.myria.be/files/MIGRA2017_NL_AS. pdf).

6

Who Is Ill?

41 93 8482

8888

90 84 76

72 6564

85

71

68 63

68

65

General health

79

64

78 68

77

76 72

69

FuncƟonal limitaƟons Chronic diseases

6869

68

7979

83 77

77

7272

EmoƟonal well-being Physical well-being

Belgium

Industrialised countries

Southern Europe

Turkey and Morocco

Rest of the world

General

Eastern Europe

Fig. 6.5 Health differences by country of birth

there could be various different explanations for this result, one of them may be that we are seeing the effect of selection once again. As it is mainly Eastern Europeans in search of (better) jobs who come to live in Belgium, it is not surprising that this subgroup of the population enjoys better average health.7

In Fig. 6.5, the category “industrialised countries” represents countries that belong to the OECD if they are not already included in one of the other categories.

7

7

What Makes Us Sick?

We can now go one step further and examine what influences the health differences described in the previous chapter. This question often forms the subject of a loaded debate, as it is closely linked to the issue of the degree of individual responsibility for health problems. Biological factors naturally play an important role: it is obvious that ageing is accompanied by deteriorating health, for example. However, biological factors are not the main explanation for the health differences between different socio-economic groups. This chapter therefore discusses the impact of lifestyle, living environment and job characteristics on health. We then focus specifically on the health dimension of “emotional well-being”. Finally, we place our results in a broader time perspective as it seems that health differences are partly determined by various characteristics of the parents.

Lifestyle, Living Environment and Job Characteristics Table 7.1 presents the results of a simple statistical analysis in which the results for the five health dimensions from the previous chapter are linked with a whole range of factors by means of a regression analysis. In addition to age, level of education and region, the effects of differences in lifestyle, living environment and job characteristics are also included in the analysis. Although the table only contains the results for these last three groups of factors, one must remember that the other characteristics were also taken into consideration. We will illustrate this interpretation for the first row of the table. Take two people of the same age who live in the same region, have the same income and level of education and come from the same type of family, but now suppose that one of

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_7

43

44

7

What Makes Us Sick?

Table 7.1 Health determinants

Lifestyle BMI Smoking Living environment Housing quality Inadequate heating Polluted environment Good social environment Job characteristics Physically demanding Dangerous work High work pressure Job satisfaction

General health

Functional limitations

−0.49 −2.34

−0.71 /

0.12 −4.07

Chronic diseases

Emotional well-being

Physical well-being

−0.75 /

/ −3.07

− 0.58

/ /

0.13 /

0.12 −4.45

0.16 /

−0.59

/

/

−0.78

/

0.09

/

/

0.09

/

/

/

/

/

/

/ /

/ /

/ /

/ −0.53

/ /

1.93

0.90

2.53

2.53

2.06

the two people has a higher value for the body mass index1 (BMI). This index reflects nutritional and exercise behaviour. A person is regarded as overweight from a value of 25 and obese from a value of 30 upwards. The table shows that on average, the person with a higher BMI achieves a lower score for the dimensions of “general health”, “functional limitations”, “chronic diseases” and “physical well-being” (all these effects have a minus sign in the table). However, the BMI does not appear to have a clearly defined effect on the dimension of “emotional well-being”. As we are comparing two people with the same level of education, the effects of the BMI in Table 7.1 cannot be attributed to the fact that people with a higher BMI often also have lower levels of education and perhaps also poorer health as a result. The table shows the direct effect of the BMI after controlling for these other characteristics. We will now discuss the results of the various factors listed in Table 7.1 in turn.

1

A person’s BMI is equal to their weight in kilograms divided by the square of their height in metres.

Lifestyle, Living Environment and Job Characteristics

45

Lifestyle The effects of the BMI on health are clear. According to our sample, almost half of the Belgians are overweight and the average BMI is 25.6. This confirms the well-known fact that obesity constitutes a significant challenge for public health in Belgium. In Table 7.1, an increase in BMI has the strongest effect on the increase in functional limitations and the level of chronic diseases. These results are intuitive. The results for smokers are slightly less intuitive. It goes without saying that smoking has a very negative effect on general health. However, based on medical knowledge of the strong link between smoking and cancer, we might also have expected to find a negative impact on the level of chronic diseases. Although there may be various explanations as to why this is not the case in Table 7.1, we should not forget that cancer patients are under-represented in our sample as they are often less willing to answer a long questionnaire about their individual well-being. Only a small number of respondents (2.7%) reported that they were suffering from a form of cancer. At first sight, the effect on the health dimension of “emotional well-being” could also be somewhat surprising: Why should smokers have a stronger tendency to become depressed? The most obvious interpretation is that the effect actually works in the opposite direction: it is probably not that smoking makes people particularly depressed, but rather that depressed people are more likely to smoke. This interpretation also illustrates the fact that many statistical results in this table (and throughout the book) must be interpreted with caution. It is very difficult to establish causal relationships based on data relating to a single moment in time. As already mentioned, the influence of lifestyle on health raises a difficult question about individual responsibility. If people get sick because they smoke, eat too much and don’t get enough exercise, should society have to pay for this? We will not go into more detail about this here. Nevertheless, we would like to make two comments. Firstly, it will become clear that differences in health are not only linked to lifestyle but also to other factors over which people have no influence (or much less of an influence). Secondly, the explanation of lifestyle differences in itself is not self-evident either. Genetic factors probably also play a role here (and surely people cannot be held responsible for this?). Lifestyle decisions are certainly influenced by a whole range of factors, with education, knowledge and the social environment playing a significant role. It is therefore best to avoid the moralistic and stigmatising allocation of individual responsibility.

Living Environment A second group of factors that can affect health is connected with people's living environment. The quality of housing is an important element here. We will return to this in detail in Chap. 11. On average, people living in good housing score better for all health dimensions with the exception of “functional limitations”.

46

7

What Makes Us Sick?

The strongest effects are found for the heating aspect. On average, people who are unable to heat their homes adequately have lower levels of “general health” and “emotional well-being”. This is yet another clear example of cumulative deprivation: housing quality, health and income are closely linked. People with a low score for one of these life dimensions often fail to achieve high scores for other dimensions too. Our survey also included questions about the quality of the living environment, in particular whether people suffered from pollution, smoke, dust, unpleasant odours or polluted water. Although a polluted environment can affect physical health in the longer term, this effect is not really reflected in our sample at a single point in time. We do find a slight negative effect on the perception of general health, but it is the consequences of a polluted environment for “emotional well-being” that are particularly striking (although perhaps not really surprising): people feel better in their skin when living in a non-polluted environment. The quality of the social network available to people is also important. People who trust their neighbours, can count on their help and who live in a neighbourhood where people generally get on well not only feel better emotionally but are also generally healthier. This finding ties in with an active recent line of research in various social sciences, which is uncovering more and more evidence of the major importance of good social interactions for individual well-being. Here, we therefore find that good relationships are not just important in themselves: they are also instrumentally important for other life dimensions such as health. The interactions between physical and mental health and the role of interpersonal relationships within this form a fascinating domain for further research. In any case, we already know that lifestyle differences are also strongly influenced by the standards in the social environment: people smoke more, exercise less and eat less healthily if other people in their environment also smoke, get little exercise and eat less healthy food.

Job Characteristics Finally, we also investigated whether the characteristics of the job have an impact on health. We had indeed expected that people with a physically stressful or dangerous job would also be in worse average health. However, the complete absence of significant effects for these job characteristics illustrates that this is not reflected in our data. There are many types of physically demanding work and some of these jobs (working as a surgeon, for example, or a gardener) can make people feel like they can enjoy themselves and need not necessarily have a negative effect on their health. This more subjective interpretation is confirmed (or at least not contradicted) by the very clear impact of job satisfaction: people who obtain more satisfaction from their work also achieve a better result for all health dimensions. In addition, higher work pressure and more stress lead to a decrease in emotional well-being. A bad job (as judged by the people themselves) does indeed seem to make them ill. This is yet

Lifestyle, Living Environment and Job Characteristics

47

another example of cumulative deprivation. The improvement of subjective working conditions could therefore be an element of the health policy.

Emotional Well-Being So far, we have focused on the row-by-row interpretation of Table 7.1. We can naturally also look at these figures per column. The results provide food for thought, especially for the fourth column: emotional well-being. As described in the previous chapter, we measured emotional well-being based on feelings of depression, anxiety, fatigue and lack of vitality. It is sometimes suggested that this is primarily an individual matter, strongly influenced by personality traits. Personality traits are indeed important for emotional well-being. In Chap.22, moreover, we will show that they also have a strong influence on general life satisfaction. However, the results in the fourth column of Table 7.1 make it clear that this is only part of the story. Emotional well-being is strongly influenced by a whole range of objective living conditions: the housing quality, a polluted or unpolluted living environment, the quality of the social network, a reasonable level of work pressure and the level of job satisfaction. We have already pointed out recent publications that link the evolution of depression to the global socio-economic evolution in our society. These evolutions over time are difficult to record in a scientifically sound manner because of the lack of properly comparable measurements. We certainly cannot say anything about this based on our own data. However, our results do show that social and economic factors have a major impact on people’s well-being. Mental healthcare must therefore focus not only on the mind, but also on the society in which this mind is shaped.

The Shadow of the Past Human health is therefore not determined solely by biological factors. The way in which our society is organised also has a clear impact on health. A health policy must extend beyond the boundaries of the medical sector, also focusing on education, housing and the labour market. A complex network of interactions exists between different characteristics of society and individual behaviour. It is therefore very difficult to define where social responsibility ends and individual responsibility begins. This latter idea is illustrated even more clearly when we take the intergenerational context into account. Figure 7.1 shows an initial attempt to do so. We examine the relationship between the level of education of each respondent’s father and their own level of health. Here, we used the same breakdown into low, medium

48

7 100 90 80 70 60 50 40 30 20 10 0

What Makes Us Sick?

89 90 81 63

75 76

68 70

General health

65

71 74 72

75

81 83

FuncƟonal Chronic diseases EmoƟonal well- Physical welllimitaƟons being being Low

Medium

High

Fig. 7.1 Health differences and the father’s level of education

and high levels of education for the father as for the respondents themselves in Fig. 6.2.2 Note the difference between this figure and Fig. 6.2 in the previous chapter. The latter figure showed the relationship between people’s health and their own level of education, while Fig. 7.1 reveals the relationship between their own health and their father’s level of education. The latter can be seen as an approximation of the economic and cultural environment in which they grew up (and for which they certainly cannot be held responsible). The results are particularly striking: for all physical health dimensions, the father’s level of education still plays a role in the health of the children once they are adults themselves. It is precisely the fact that the effect is less pronounced or even absent for emotional well-being which makes these findings even more striking. To put it in very concrete terms, children whose father had a low level of education are more likely to become chronically ill as adults, to suffer pain or to experience limitations in their normal everyday activities. This result is also in line with recent research which is uncovering more and more evidence of the lasting influence of the situation during childhood, or even the prenatal environment, on later life.3 Chapter 18 deals in more detail with the situation of the children in our sample. Results such as those in Fig. 7.1 suggest that the situation of these children to some extent reflects their later well-being as adults. As a result, a good poverty policy cannot start early enough.

2

The results for the mother’s level of education are very similar. The study by Almond (2006), for example, shows that children whose mother was pregnant during the Spanish influenza epidemic in 1918 had more health problems and did less well economically throughout their lives.

3

8

Can People Afford Their Healthcare?

The quality and accessibility of healthcare are primarily important because they (hopefully) contribute to better health, undoubtedly one of the most important dimensions of well-being. The organisation of healthcare and health insurance also affects well-being in other ways. When people are ill, they have to spend part of their income on healthcare expenditure which reduces their possibilities for consumption. The accessibility of good care for all also contributes to a sense of solidarity and harmony in society: people who do not receive the care they think they need often feel that they are being treated unfairly by the system. The importance of each individual’s dignity is a particularly sensitive issue when it comes to suffering, pain and death.

The Use of Healthcare We will first examine the use of healthcare as reflected in our data. Table 8.1 paints an indicative picture, also immediately illustrating the distribution of this use among the population. We show the percentage of people calling upon different forms of care, also specifying the extent of use for most of these forms of care: the number of days in a hospital, the number of consultations with a care provider and the number of home care visits, for example. These averages are not calculated for the total population, but only for the people who use this form of care. For example, 87.8% of people call upon general practitioners (GP) at least once a year and the

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_8

49

Global Male Female 18-39 40-59 60-69 70-79 80+ Poor health Medium low Medium high Good health Emotional, low Emotional, medium low Emotional, medium high Emotional, good

7.6 7.0 8.0 6.6 8.5 6.6 7.0 9.5 11.0 6.0

5.3

2.9 11.2

6.5

3.6

6.3

14.4

10.8 29.4

18.3

16.3

15.2

5.5 5.2 5.8 4.0 5.4 5.3 7.5 9.9 8.9 5.0

84.2 5.8

88.0 4.2

87.5 4.3

51.9

53.6

61.1

46.6 72.4

2.7

2.8

3.5

2.4 4.8

2.8

3.5 3.3 3.7 3.5 3.7 3.6 3.0 2.8 5.3 3.0

Specialist— number of consultations

70.1

71.3

69.0

71.5 64.5

71.3

68.2 65.7 70.5 69.9 74.0 69.4 55.1 43.1 65.3 66.6

Dentist (%)

22.4

22.8

24.5

19.3 39.7

23.0

26.4 21.1 31.8 25.3 27.5 28.3 26.5 23.0 41.1 27.6

Physiotherapist (%)

17.7

15.2

16.0

9.2 31.6

16.1

21.9 19.3 23.6 13.8 20.4 23.6 36.2 42.3 33.5 20.1

5.1

5.1

6.9

3.0 14.7

5.0

7.8 5.2 10.3 3.7 5.8 8.9 13.3 27.6 17.3 7.4

Physiotherapist— Home number of care consultations (%)

(continued)

16.2

14.7

16.5

7.9 23.2

12.1

19.6 19.6 19.5 7.6 11.2 14.5 26.9 36.8 23.2 20.4

Number of visits

8

82.0 4.2 92.7 7.7

54.4

58.5 50.3 66.3 48.8 58.5 65.4 71.0 65.7 78.5 61.6

GP— Specialist number of (%) consultations

84.9 3.7

87.8 84.8 90.8 83.1 86.2 92.7 94.6 96.9 95.7 81.0

Hospital— GP number of (%) days

19.2 16.9 21.4 15.2 16.7 21.4 29.3 30.2 35.6 19.9

Hospital (%)

Table 8.1 Use of healthcare (per year)

50 Can People Afford Their Healthcare?

Chronically ill Not chronically ill Education, low Education, medium Education, high Poor Low income Average income Rich

9.2

5.5

9.4

7.1

5.8

11.1 7.8 7.2

6.3

13.2

22.1

18.8

16.6

18.3 22.7 17.9

17.5

87.3 3.9

82.6 6.4 90.6 7.4 89.4 4.5

87.8 4.8

88.1 4.8

88.1 7.1

84.5 4.2

61.2

48.9 58.5 58.4

73.6

55.6

56.8

49.2

78.4

GP— Specialist number of (%) consultations

94.6 7.7

Hospital— GP number of (%) days

31.7

Hospital (%)

Table 8.1 (continued)

3.2

3.5 3.5 3.6

3.5

3.5

3.5

2.6

4.6

Specialist— number of consultations

77.5

49.6 62.2 71.8

78.9

72.3

52.9

70.6

65.6

Dentist (%)

29.1

21.4 27.5 27.7

30.8

25.0

24.3

21.4

38.2

Physiotherapist (%)

14.5

25.7 25.6 23.7

18.2

21.7

27.3

11.9

32.1

3.6

8.2 10.8 8.0

5.9

5.9

11.9

4.5

14.3

Physiotherapist— Home number of care consultations (%)

12.6

22.7 25.4 12.7

12.8

13.0

26.8

13.5

22.7

Number of visits

The Use of Healthcare 51

52

8

Can People Afford Their Healthcare?

average number of GP consultations by these people is 5.5. Among people in poor health, the results are 95.7% and 8.9, respectively.1 The first row shows the average use of healthcare by the entire Belgian population. These figures are in line with expectations.2 We note that almost 20% of people use hospital care, but it must be borne in mind that our questions did not distinguish between day hospital admissions and longer hospital stays. The breakdown of the results between the various groups is rather more interesting. Generally speaking, the use of healthcare increases with age and this is clearly the case for home care. There are two exceptions to this trend: the use of dental care and physiotherapists increases up to the age of 60–69, but then decreases. At the same time, the number of consultations shows that elderly people who use physiotherapy do so in a more intensive way, probably for more serious conditions. Women call upon healthcare more than men. This is partly due to the fact that on average, women live longer. However, other studies show that the effect persists even after correction for age. For example, the difference in the use of specialist care is partly explained by consultations with the gynaecologist. Age and health are, of course, closely linked, and it goes without saying that sick people call upon the care system more often. We show the results for “general health” (the first dimension described in Chap. 6) and “emotional well-being” (the fourth dimension). We find a positive link between illness and the use of healthcare for both dimensions, but the link is stronger for “general health” than for “emotional well-being”. People with chronic diseases also use the care system more often than other people: had we found the opposite, our results would not have been particularly credible. Two findings are worth highlighting. Firstly, the number of people who visit a GP is fairly evenly distributed among people with different levels of health (note how this differs from the results for specialist care), except for those with poor health, over 95% of whom ask a GP for advice at least once. The high percentage for all groups (over 80%) suggests that primary care is widely accessible in Belgium. Secondly, we again find different results for dental care here: people with a low score for “emotional well-being” and “general health” pay fewer visits to the dentist. Next, we examine the relationship between the use of care and socio-economic status. In Chap. 6, we saw that people with lower levels of education and those on lower incomes are generally in poorer health, so we could also expect them to make greater use of the care system. Indeed, people with a lower level of education (below secondary) visit hospitals more often and make greater use of home care. For general practitioners, the picture is slightly different: although the number of users is fairly evenly distributed across all levels of education, as soon as people call upon a GP the number of consultations is higher for people with lower levels of education, probably because they are indeed in poorer health. We find another 1

We do not include the number of consultations for dentists as it does not contain any relevant information: the average number of visits is two for all groups. 2 Our figures are in line with the findings of the 2013 Health Survey, but are somewhat higher overall. This brings them closer to the official data of the National Institute for Health and Disability Insurance (RIZIV).

The Use of Healthcare

53

different pattern for physiotherapists: although more highly educated people are more likely to visit a physiotherapist, the average number of visits is lower. More highly educated people are likely to consult the physiotherapist for less severe conditions, relatively speaking. The results for dental care (more highly educated people visit the dentist more regularly) and particularly for specialist care are striking: despite their better level of average health, more highly educated people visit a specialist more frequently. Similar results are also found in other research and for other European countries (e.g. see the work of Van Doorslaer and Masseria 2004). However, the most striking results are found when we look at income. Poor people, i.e. people living in families with an income below 60% of the median income (here, we use the official method of measuring poverty as described in Chap. 5), make less use of healthcare across the board. This finding even applies to visits to their GP. However, if they are admitted to hospital (less often than the other income groups) their stay is longer: this suggests that they must be “sicker” before being admitted to hospital, and in turn this may be due to their lower probability of consulting their GP. At first glance, these results are worrying. We will examine them in more detail later in this chapter.

Financial Consequences of Illness At this point, we will take a look at Belgians’ personal contributions to their care costs. In Belgium (and also in our sample), virtually everyone is insured under the compulsory health insurance system. Most people have no idea about the amount of social security contributions they pay for this and simply regard it as part of their taxes. We will not go into these contributions in more detail here. However, people also pay for their healthcare when they get sick. Even in a system of collective and solidarity-based health insurance that covers most of the expenses, these personal contributions can really add up. Here, we refer to three different types of personal contributions. First and foremost, these consist of the patient contributions, i.e. the patient’s own contributions as specified in the collective health insurance system. This is the portion of the official tariff that is not reimbursed by the health insurance. Then we have the supplements, which are the additional fees that can, under certain circumstances, be charged by providers on top of the official price. The hospital supplements for patients admitted to a single room are perhaps the most common form of these. Certain types of care (some drugs, for example) are not reimbursed at all and must therefore be funded entirely by the patient. Finally, chronically ill people in particular also incur additional expenses that are not directly related to healthcare but can have a considerable impact on their budget: special food will sometimes be needed, travel can become more difficult and in certain cases, their housing will need to be adapted. We will return to the specific situation of the chronically ill at the end of this chapter.

54

8

Can People Afford Their Healthcare?

The Belgian system has built-in safeguards. People on a low income pay lower patient contributions; they are entitled to increased reimbursements. The characteristics of these people are summarised in the first column of Table 8.2. It is mainly older people with lower levels of education and lower incomes who are entitled to these increased reimbursements. Because of the link between socio-economic status and health, these people are often less healthy. When the total level of patient contributions within a family reaches an (income-dependent) threshold, further contributions are reimbursed under the mandatory health insurance. This is known as the maximum billing system (abbreviated to the Dutch acronym MAF). The second column shows how many people have had their patient contributions reimbursed within this system during the previous year.3 Since an (income-dependent) threshold must be reached first, it is logical that the maximum billing system will primarily benefit sick people on a low income. This is confirmed in the second column of the table. Also note that, despite the income dependence of the thresholds, the MAF remains a universal system. The link between income and reimbursement through the MAF is therefore much weaker than the link between income and the entitlement to an increased reimbursement. In relative terms, we can see that the poorest and richest groups benefit least from this. For the poorest groups, this may be connected with the less frequent use of healthcare that we have already described. After all, expenditure must first have been incurred in order to benefit from the MAF. Supplementary charges are not covered by the MAF. If people do not wish to pay these costs themselves, they can take out supplementary hospitalisation insurance. This is voluntary insurance for which people pay a premium that depends on their age and risk profile but not their income. The results in the third column show that almost 80% of Belgians have this kind of hospitalisation insurance, but also confirm that the coverage of this insurance is very unevenly distributed among the population. Wealthier and more highly educated people are more likely to have it. This immediately leads to the somewhat paradoxical result that it is mainly relatively healthy people who have hospitalisation insurance. The fourth column shows the percentage of people who have received a reimbursement through their hospitalisation insurance: these percentages do not vary greatly with age or socio-economic status (except for the poor who make less use of the care system), but do vary greatly with health. The last two columns shed a global light on the subjective perception of financial costs associated with healthcare. The last column shows the percentage of people who claim they were unable to pay a health-related bill for financial reasons. The penultimate column shows the percentage of respondents who say that they find it difficult or impossible to budget for their personal contribution to health costs, i.e. the money they have to pay themselves. When interpreting the results, it is interesting to look at these two columns together. A more accurate (but cumbersome) wording would be as follows: “How many people over the age of 18 live in a family that received a reimbursement through the maximum billing system in 2015”.

3

Global Male Female 18–39 40–59 60–69 70–79 80+ Education, low Education, medium Education, high Poor Low income Average income Rich Poor health Medium low Medium high Good Emotional, low Emotional, medium low

8.9 8.3 9.4 4.7 7.7 12.1 16.1 15.2 11.4 9.0

6.4 5.6 14.2 7.0 4.2 19.0 8.3 5.4 4.0 16.6 7.6

18.4 17.2 19.5 13.3 17.4 20.3 25.1 33.0 27.2 17.7

10.4 36.1 27.3 12.3 8.3 32.0 17.6 13.0 12.9 33.6 15.6

Increased Maximum reimbursements (%) billing (%)

Table 8.2 Financial consequences of illness

87.1 42.1 73.7 85.4 88.1 75.3 82.6 82.8 80.4 71.2 78.4

78.7 77.8 79.4 76.2 79.6 88.0 76.0 63.8 67.0 81.4

Hospitalisation insurance (%)

27.5 14.0 27.3 29.7 25.9 45.5 22.7 20.7 18.3 34.0 27.3

25.9 23.1 28.6 25.8 25.0 26.6 28.4 26.4 24.8 25.2

Reimbursement by hospitalisation insurance (%)

7.3 40.3 26.1 9.3 3.6 35.5 15.0 9.3 7.8 37.7 15.2

17.0 14.4 19.5 15.4 17.0 16.1 18.5 25.1 25.9 17.9

Difficult or impossible to fit into the budget?

(continued)

3.7 17.9 11.4 2.3 1.3 12.7 7.6 3.8 4.7 15.8 6.3

7.2 7.0 7.4 9.1 8.9 4.6 1.4 4.6 9.5 8.2

Unable to pay?

Financial Consequences of Illness 55

Emotional, medium high Emotional, good Chronically ill Not chronically ill

5.6

5.9 16.0 5.1

13.7

11.8 30.3 12.3

Increased Maximum reimbursements (%) billing (%)

Table 8.2 (continued)

86.2 79.6 80.6

84.6

Hospitalisation insurance (%)

19.6 35.7 20.7

24.7

Reimbursement by hospitalisation insurance (%)

6.4 29.0 9.8

7.3

Difficult or impossible to fit into the budget?

1.9 10.5 5.3

4.6

Unable to pay?

56 8 Can People Afford Their Healthcare?

Financial Consequences of Illness

57

In both cases, there is a clear link with socio-economic status. Once again, the results for the poorest group are very striking: although they make less use of healthcare, 40% still say that healthcare expenditure is difficult or impossible to budget for and almost 18% say that they sometimes cannot pay the bills at all. There is also a very strong link with health. It is also striking that 7.3% of highly educated people, 3.6% of people in the highest income quartile and 7.8% of people in good health state that they find healthcare expenditure difficult or impossible to budget for. However, we must bear in mind that subjective feelings rather than objective budget data are involved. This insight also helps us to understand the results for the demographic characteristics: women and older people are more likely to say that the expenditure is difficult to budget for, but at the same time they report less often that they (have to) postpone the payment of bills. A somewhat daring interpretation (which is nevertheless consistent with the results) is that older people and women attach greater importance to healthcare and thus make their use of care (and the payment of the associated health bills) less dependent on the size of their budget: they continue to use and pay for healthcare, increasing the pressure on their budget as a result. An alternative explanation could be that women and the elderly are sicker and that the care is more necessary. Our findings raise a pertinent question: Are people obliged to postpone care for financial reasons?

Postponement of Care International and Belgian scientific literature on the accessibility of healthcare focuses strongly on whether people are obliged to postpone the use of care for financial reasons. However, the results of surveys for the same country and period often differ: for example, the Health Survey found that no fewer than 12.3% of people had to postpone care in 2008, while for the same year, the SILC survey found a percentage of less than 1%. It appears that the results of the various surveys are strongly influenced by the specific formulation of the questions (Schokkaert et al. 2017). Our results are in line with those of the Health Survey, partly because it is the one our questions tie in most closely with. Despite the wide variation in the answers, it is interesting to study them, as the distribution across the different groups of the population gives relatively stable results in the various surveys. Let us first look at the last two columns in Table 8.3. The penultimate column shows the percentage of people living in a family that had to postpone at least one form of healthcare for financial reasons during the previous year. We find that this affects 13.4% of Belgians. The last column shows the percentage of people who have had to postpone urgent care. This share is much lower, at 4.7%. From a social viewpoint, this latter result is perhaps the most relevant. We note that sick people in particular report that they have had to postpone care. This is not particularly surprising, as the others do not need any (or less) care. What is more interesting is the observation that the postponement of urgent care occurs

Global 7.2 Male 6.7 Female 7.7 18–39 9.8 40–59 6.7 60–69 5.6 70–79 5.4 80+ 5.9 Education, 7.6 low Education, 8.2 medium Education, 5.5 high Poor 16.1 Low income 8.5 Average 2.7 income Rich 5.3 Poor health 10.3 Medium 5.6 low

7.7 7.7 7.7 9.0 9.0 6.3 3.6 4.0 7.7 8.7 6.2 17.2 8.7 3.7 4.5 10.5 7.7

2.4

1.6

7.2 2.1 0.6

1.4 4.2 0.5

Postponement of dental care (%)

2.2 2.3 2.2 2.5 2.3 2.3 1.2 2.3 2.4

Postponement of Postponement GP or specialist of operation (%) (%)

Table 8.3 Postponement of care for financial reasons

4.1 7.8 3.2

1.9 9.1 4.3

13.5 6.1 3.9

4.6

6.7

5.8 5.2 6.3 6.0 7.9 3.8 2.1 3.4 5.7

1.6 4.6 1.4

3.5 3.8 1.1

1.8

2.4

2.3 2.0 2.4 2.3 3.0 2.2 0.4 0.9 2.1

7.0 19.0 12.1

28.1 16.9 7.0

9.7

15.5

13.4 12.6 14.1 15.7 14.6 11.2 8.0 9.3 14.2

Postponement Postponement of Any of glasses (%) mental healthcare postponement (%) (%)

1.0 10.0 4.9

12.0 7.1 1.7

1.3

5.8

4.7 3.7 5.6 4.8 6.3 4.5 1.4 2.1 6.7

(continued)

Postponement of urgent care (%)

8

9.8 5.4 2.2

4.0

5.4

5.1 4.5 5.7 6.3 4.1 5.0 5.5 4.7 5.6

Postponement of prescribed drugs (%)

58 Can People Afford Their Healthcare?

Medium 6.4 high Good 6.2 Emotional, 12.5 low Emotional, 6.6 medium low Emotional, 4.8 medium high Emotional, 4.5 good Chronically 8.1 ill Not 6.5 chronically ill

5.9 7.5 15.0 6.5 4.5

5.6 8.6 7.6

2.3 3.7

1.9

1.6

2.0

3.0

1.9

Postponement of dental care (%)

2.2

Postponement of Postponement GP or specialist of operation (%) (%)

Table 8.3 (continued)

4.4

6.2

3.4

3.5

5.0

4.7 8.3

4.6

Postponement of prescribed drugs (%)

4.4

7.6

2.8

3.0

4.9

4.9 11.5

4.0

1.4

3.8

1.1

0.3

2.0

1.3 5.8

2.1

11.9

16.1

8.1

7.7

12.8

11.0 25.1

11.7

Postponement Postponement of Any of glasses (%) mental healthcare postponement (%) (%)

3.1

7.6

0.9

1.5

3.7

2.3 12.6

2.0

Postponement of urgent care (%)

Postponement of Care 59

60

8

Can People Afford Their Healthcare?

somewhat more frequently in younger people: this confirms the assumption that we formulated above, namely that elderly people continue to use and pay for healthcare even if it proves difficult or impossible to budget for it. In particular, there is a very strong link with education and income. Of the adults living in a family with an income above the median, 7% live in a family in which some form of care has been postponed, but less than 2% of cases involve urgent care. For the poor, these percentages are 28.1 and 12%. This is fully in line with all previous results showing that poor people find it harder to fit healthcare spending into their budgets and are more often unable to pay healthcare bills. Despite the social protection measures already built into the system, ensuring the financial accessibility of care for the poor in our society remains an important concern. Further insights can be derived from the other columns of the table, which provide similar results for different forms of healthcare. Dental care is most commonly postponed, even by people on above-average incomes. Some people also postpone the consultation of a general practitioner or specialist to times when their budget is less tight. The results for the poor are worrying across the board. For example, we note that almost 10% have postponed the purchase of prescribed medicines and just over 7% have postponed surgery. This confirms the earlier result that it is particularly the poor who postpone urgent forms of healthcare for financial reasons or even put them on hold indefinitely (given their reduced use of the care system).

People with Chronic Illnesses The chronically ill deserve special attention in this respect. We define a “chronically ill person” as someone who reported suffering from a chronic condition in the MEQIN survey. In total, 36% of the population are affected. There are, of course, drawbacks to this kind of self-reporting and—as described in Chap. 6—chronically ill people are therefore a heterogeneous group who suffer from very different conditions. Nonetheless, it is still interesting to investigate how good the lives of these chronically ill people are compared to the rest of the population. The results for the use and accessibility of healthcare have already been included in the tables above. Chronically ill people make greater use of healthcare in general, although they visit the dentist less often. They are more likely to benefit from increased reimbursements and reimbursements under the maximum billing system. Although they are no more likely than other groups to have hospitalisation insurance, a higher percentage has received reimbursements through this insurance. Almost 30% of chronically ill people feel that their healthcare expenditure is difficult or impossible to budget for; just over 10% say that they sometimes cannot pay their health-related bills and 7.6% were obliged to postpone a form of urgent care for financial reasons during the past year.

People with Chronic Illnesses

61

Table 8.4 Situation of the chronically ill Chronically ill people How satisfied are you with your health? (0–10) Emotional well-being (0–100) How happy would you say you are? (0–10) How satisfied are you with your life today? (0–10) Average monthly expenditure on leisure activities Can't afford an unexpected expense of 1000 euros Below the income poverty line

Non-chronically ill people

5.96 64.24 7.15 6.99

7.97 76.05 7.82 7.63

25.25 euros

32.18 euros

31.7%

20.7%

12.1%

11.1%

However, chronic illness can also affect well-being in other ways. Some relevant indicators are included in Table 8.4. It is not surprising that the chronically ill have lower levels of emotional well-being and are less satisfied with their health. This effect also has a strong negative impact on their happiness and overall life satisfaction. As we will see in Part IV of this book, good health is indeed regarded by most people as one of the most important characteristics of a good life. A chronic illness also has economic consequences. On average, chronically ill people spend less on leisure activities. Almost a third of people with chronic illnesses state that they would be unable to pay an unexpected expense of 1000 euros with their own resources. It should therefore come as no surprise to learn that chronically ill people fall below the poverty line relatively often.

9

Do We Find the Job of Our Dreams?

Alongside health, the employment situation can be regarded as an essential non-material dimension of well-being. Employment can contribute to social integration and the impact of involuntary unemployment on overall life satisfaction often goes far beyond what can be expected based on income loss alone. However, the description of an employment situation does not just involve the simple distinction between working and not working. The quality of the job is also of great importance for people’s well-being. We will examine these different aspects in this chapter. Labour market statistics are, of course, widely available and also play a major role in the macroeconomic debate. Table 9.1 compares some common labour market participation statistics, as calculated on the basis of our data, with the official statistics of the International Labour Organization (ILO) which are available on the OECD website. The activity rate is the ratio of the labour force to the population of working age. In international statistics, the latter is defined as people aged from 15 to 64. The labour force consists of workers and jobseekers. The employment rate is the ratio of the number of workers to the population of working age, and the unemployment rate is the percentage of jobseekers among the labour force. Table 9.1 shows that the MEQIN data comes close to official statistics. However, the unemployment rate in our data is almost 2% points higher than that found in the ILO statistics (10.4 vs. 8.5%).

Who Works and Who Doesn’t? In the social debate, much attention is paid to the relationship between employment and age. After all, given the ageing population, efforts are being made to get more older people to work. Table 9.2 shows the percentage of workers in different age groups. We notice a marked drop in the relative number of workers once people pass the age of 50. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_9

63

64

9

Do We Find the Job of Our Dreams?

Table 9.1. Labour market statistics Activity rate (%) Employment rate (%) Unemployment rate (%)

MEQIN dataset (2016)

OECD (2014)

69.0 61.8 10.4

68.0 62.2 8.5

Table 9.2. Labour force participation by age Age

Percentage of workers (%)

18–29 30–39 40–49 50–59 60–64 65+

53.5 80.1 81.4 68.2 24.8 3.0

People between the ages of 50 and 65 certainly do not form a homogeneous group. Less than half (43%) of non-working people between the ages of 50 and 59 say they do not work for health reasons. Less than a fifth of the non-working people in this age group cite retirement, early retirement or being in an early leavers scheme as their reason for not working. This is in stark contrast to non-working people between the ages of 60 and 64, who are no longer actually looking for work. Nearly two-thirds of them say they are not working because they are retired, took early retirement or are on an early leavers scheme. Health reasons only explain why 12% of the people in this age group are not working. It is striking that 3% of people over the age of 65 say they are still engaged in some form of paid work. Incidentally, the picture is simple for non-workers over the age of 65: with only a few exceptions, they say they are not working because they are retired. Even the youngest group includes a lot of people who are not working, but of course this is primarily due to the fact that some of them are still studying: 57% of non-working people between the ages of 18 and 29 say they are not actively seeking work because they are studying. Age is not the only factor that relates to working. Table 9.3 shows the results for several other factors: gender, education, country of birth, region, position in the family and health. Men still work relatively more, and there are also regional differences. However, the effects of education level and health are much more pronounced.1 Many of these factors are naturally related. We therefore also estimated the individual contribution of each of these factors to the likelihood of someone working, while controlling for the other factors. For this purpose, we used the same statistical technique as in Chap. 7 when we examined the underlying determinants 1

Here, we have defined and classified health in the same way as in Chap. 8: on the basis of general health and in four equal groups.

Who Works and Who Doesn’t?

65

Table 9.3. Labour participation by socio-demographic characteristics Percentage of workers Adult population (18+) Gender Male Female Education Low Medium High Country of birth Belgium Southern Europe Eastern Europe Morocco/Turkey Industrialised countries Rest of the world Region Brussels-Capital Region Flanders Wallonia Position in the family Single Partner Parent in single-parent family Child Health Poor Medium low Medium high Good

Adult population below the age of 65

55.80% 46.20%

71.20% 61.50%

27.20% 54.60% 70.30%

45.40% 64.50% 84.10%

52.30% 25.90% 64.90% 31.20% 44.20% 45.30%

69.50% 37.70% 67.40% 35.30% 56.80% 48.10%

49.10% 53.20% 47.10%

60.20% 70.80% 60.30%

35.40% 56.40% 56.70% 40.90%

59.20% 72.20% 64.50% 40.90%

33.50% 48.20% 56.10% 65.10%

49.60% 68.60% 71.80% 73.10%

of health differences (Table 7.1). Even after taking other factors into account, the gap between men and women remains over 10% points. For each decade of age, the likelihood of people working decreases by 6% points. More highly educated people are almost 30% points more likely to work than people with lower levels of education. Poor health in itself reduces the chances of employment by 15% points. Remarkably, the chances of employment for people of Eastern European origin are almost as high as those for people of Belgian origin, even after correcting for the fact that this population group is younger than the Belgians. In Chap. 6, we have already seen that people of Eastern European origin enjoy better than average

66

9

Do We Find the Job of Our Dreams?

Table 9.4. Reason for not working and jobseeking behaviour All non-working 18–64 year olds (%) Reason for not working Studying Not worked before and not yet found a job Lost previous job Prefer income replacement Health (Early) retirement or early leavers scheme Too old Other Total

Percentage of jobseekers in the same subgroup (%)

16.20 3.8

11.50 76.30

14.20 5.9 21.90 21.20

79.20 41.00 5.0 0.0

0.9 15.80 100.00

0.0 21.40 22.90

health. Their higher chances of employment support the hypothesis that these are often relatively healthy and young people who have moved to Belgium in search of work. The percentage of working people who do not have an Eastern European or Belgian background, on the other hand, is considerably lower. Half of the jobseekers have been looking for work for more than a year and over two-thirds have not worked in the last 12 months. In Table 9.4, we examine the reasons the respondents give for not working. In the second column, we report the percentage of people effectively seeking employment for each of the reasons given. Since people over 65 are not looking for work and are not working because they are retired, we focus only on people between the ages of 18 and 64. Health and (early) retirement are the main reasons for not working. If people specify one of these reasons for not working, they are not looking for work either. Jobseekers are mainly found among those who say they are not working because they lost their previous job and have not yet found a new job, and those who have not yet worked and are still looking for their first job. All the same, it is striking that a quarter of the latter category says they are not actively looking for work. These are not students, as they are in a different category. However, of the people who say they are not going to work because they prefer their current income replacement to a job for which they may be eligible, just under half are still looking for work (41%). The unemployment trap—the knock-on effect of unemployment due to the (excessively) small difference between wages and the income that replaces them—therefore does not automatically imply resignation to their situation. Approximately half of the non-working people in the “other” category state that they are housewives. Apparently, the modern househusband remains a rare phenomenon. Chapter 12 deals with the specific division of the household tasks and childcare in more depth.

Which Types of Jobs Are Carried Out? How Much Do People Work?

67

Which Types of Jobs Are Carried Out? How Much Do People Work? Table 9.5 shows that almost four-fifths of the working population are wage earners, a good half are white-collar workers (including civil servants) and more than a quarter are blue-collar workers. According to our data, 14.2% are self-employed and 3.7% combine their self-employed status with employment. These figures are comparable with the corresponding statistics based on administrative figures reported in the second column.2 A quarter of the workforce works part-time. Almost a third (31%) of part-time employees work on a 4/5 basis, 29% work half-time and 11% on a 3/4 basis. Part-time work is more common among the elderly, women and people with medium and low levels of education, and less common among people who live in the Brussels-Capital Region. People are also more likely to work part-time as the number of young people in the family increases. Almost 90% of all employees have contracts for an indefinite period; almost over 7% have fixed-term contracts and 2% work exclusively on an interim basis. Almost 93% of workers have only one job, meaning that the others combine multiple jobs. 62% of the blue-collar workers are on hourly wages, whereas 90% of white-collar workers have monthly salaries. The number of working hours is specified in the contract for 90% of wage earners. According to 62% of these people, these hours correspond to the number of hours actually worked. People who say that the hours specified in the contract differ from their actual number of hours, work for an average of almost seven hours more than the hours stated in the contract. We also asked people about the net income they earned through their work and their net replacement income if applicable. The average net monthly salary of a white-collar worker in our MEQIN dataset is 2070 euros, almost 500 euros higher than that of a blue-collar worker (an average of 1598 euros). On average, self-employed people earn an additional 300 euros (an average of 2368 euros3). Although the average unemployment benefit is 1028 euros, this may be a somewhat too favourable estimate of the benefit paid by the government as it also includes maintenance payments received. An average pension for non-working people is 1481 euros. The difference in hourly wages between blue-collar workers and white-collar workers is 3.5 euros (10.8 vs. 14.3 euros).

These figures come from the Crossroads Bank for Social Security Data Warehouse on Labour Market and Social Protection and refer to the fourth quarter of 2014: https://dwh-live.bcss.fgov.be/ nl/dwh/dwh_page/content/websites/datawarehouse/menu/webtoepassing-globale-cijfers.html. 3 Here, we only took into account the respondents who stated a positive amount. Incidentally, when we try to allocate the income of the self-employed to individual family members for activities involving more than one family member, the average income drops considerably to 2117 euros and is only a little over 50 euros higher than that of employees. 2

68

9

Do We Find the Job of Our Dreams?

Table 9.5. Types of jobs Classification of workers Population aged 18–64 Blue-collar workers (%) White-collar workers (%) Self-employed (%) Combination of wage-earning and self-employed (%)

MEQIN (%)

CBSS DWH labour market fourth quarter of 2014 (%)

26.8 53.0 14.2 3.7

27.2 51.5 16.2 5.1

The Job of My Dreams, an Unattainable Ideal? We surveyed working people about the characteristics of their job. In particular, we asked respondents to indicate on a six-point scale to what extent they agreed with a series of statements about their job. Table 9.6 shows how many of the respondents agreed or fully agreed (score 5 or 6) with these various statements. Most people feel that they do worthwhile things in their job, that they are given the opportunity to show what they can do and that their job requires a lot of mental effort. To a large extent, people share the view that their work is a challenge and that they can decide for themselves how to do their job. Although more than a third feel they can decide for themselves how much and what work they do in a day, at the same time almost 43% feel that their work rhythm is high and that they are under time pressure. More than a third also feel that a lot of physical effort is required for their work. A relatively small group feel that they have to work in dangerous or unsafe conditions or that they get dirty. “Relative” is indeed relative here: it could also be concluded that it is worrying from a social perspective that 13% of workers have to work in dangerous or unsafe jobs. In order to gain a better overview of their own preferences, we also asked the people concerned what they would consider an ideal job. Table 9.7 summarises the results. The main—albeit perhaps unsurprising—finding is that people seek recognition of their abilities: almost everyone thinks that in an ideal job people should do things that are worthwhile, that they should be able to show what they are capable of and that the work should be challenging. Freedom and autonomy are also considered very important. For many respondents, being able to decide how to do their work and how much and what work they do in a day also appear to be important aspects of an ideal job. Quite surprisingly, 35% of people disagree or agree only to a limited extent with the statement that one should not be exposed to dangerous or unsafe conditions in an ideal job. Few people are attracted by a high level of physical effort at work. Little time pressure and a reasonable pace appear to prove important for fewer people than one might think on the basis of opinion

The Job of My Dreams, an Unattainable Ideal?

69

Table 9.6. Characteristics of current jobs Percentage of respondents who (completely) agree with the following statements about their job

Percentage

1. My job requires a lot of physical effort 2. My job requires a lot of mental effort 3. I have to work in dangerous or unsafe conditions 4. I have to work at a fast pace or under time pressure 5. I have to get dirty at work 6. My work allows me to show what I can do 7. I can decide for myself how to do my job 8. I can decide for myself how much work I do in a day 9. I can decide for myself what work I do in a day 10. My work is a challenge 11. I do things in my job that are worth doing

38.2 68.1 12.8 42.7 17.4 72.3 56.2 40.1 37.2 58.8 72.8

Table 9.7. Characteristics of ideal jobs Percentage of respondents who (completely) agree with the following statements about their ideal job

Percentage

1. My ideal job requires a lot of physical effort 2. My ideal job requires a lot of mental effort 3. In my ideal job, I am not exposed to dangerous or unsafe conditions 4. In my ideal job, there is little time pressure and the pace is not particularly high 5. In my ideal job, I do not have to get dirty 6. My ideal work would allow me to show what I can do 7. In my ideal job, I can decide for myself how to do my work 8. In my ideal job, I can decide for myself how much work I do in a day 9. In my ideal job, I can decide for myself what work I do in a day 10. My ideal work is challenging 11. In my ideal job, I do things that are worth doing

22.5 57.5 65.1 45.2 50.1 92.2 83.2 76.2 76.4 86.8 93.2

papers.4 Again, we must be careful with our interpretation of the results. After all, we could also say it is surprising that almost half the respondents (45%) would choose a job with little time pressure. In order to look beyond the averages and reach a more accurate assessment of the quality of the jobs in the opinion of the people concerned, we need to compare their actual situation with their personal preferences. Here, we distinguish between four dimensions: challenging work (characteristics 6, 10 and 11 of the list), physically demanding job (characteristics 1, 3 and 5), autonomy (characteristics 7, 8 4

There is an increasing number of publications on stress at work and burnout, such as the Swinnen publication (2012), for example.

70

9

Do We Find the Job of Our Dreams?

Table 9.8. Degree of correspondence between current and ideal job characteristics Dimension

Challenging work

Physical effort

Autonomy

Intellectual effort

Average score (0–100)

83.4

72.9

73.5

72.6

and 9) and intellectual effort (characteristics 2 and 4). For each of these dimensions, we then constructed a score ranging from 0 to 100 that reflects the extent to which a respondent’s current job characteristics match those of his or her desired job. Respondents whose current job characteristics are more in line with what they expect from an ideal job will receive a higher score. For example, people who do not wish to make excessive physical effort at work but have a job where this is required will have a relatively low score for the “physical effort” dimension. However, if they have a job that requires a lot of physical effort and said that they enjoy work that requires physical effort, they would actually achieve a high score for this dimension, as would someone of whom little physical effort is required and who also has little desire to work in this way.5 The results are summarised in Table 9.8. The challenging and worthwhile aspects of the current job largely correspond to what Belgian workers expect from this dimension. However, this applies to a lesser extent for the other three dimensions. Here too, one may wonder to what extent socio-economic factors and other dimensions of well-being relate to the degree to which a person succeeds in finding a job that corresponds to his or her own image of an ideal job. Part of the answer is given in Table 9.9. Men are better than women at finding a job that meets their expectations in terms of autonomy and the challenging nature of their work. It is striking that older people also score better in these two areas. In every respect, a higher level of education helps people to find a job that they consider interesting. Working people who were born in other industrialised countries are even more likely to find a job that matches their ideal than people born in Belgium. Southern and Eastern Europeans do less well than Belgians in this area, apart from the expected degree of intellectual effort for working Southern Europeans and the expected physical effort for working Eastern Europeans. People born in Morocco and Turkey score significantly worse in all respects: if they find work, it is not exactly the job of their dreams. Apart from the expected degree of intellectual effort required by the job, blue-collar workers are much less successful than white-collar workers and the self-employed at finding what they consider to be an ideal job. Not unexpectedly, self-employed people score particularly well in terms of autonomy and the 5

Technically, we deducted 20 points for each point of difference between the respondent’s response for the characteristic of the current job and the corresponding degree of desirability of this characteristic in an ideal job from a basic score of 100 points for a particular job characteristic. For each dimension, we then averaged the resulting scores of the associated characteristics. A score of 100 therefore means that the characteristics of the current job correspond perfectly to the respondent’s vision of an ideal job. The lower the score, the less this is the case.

The Job of My Dreams, an Unattainable Ideal?

71

Table 9.9. Correlation between individual characteristics and alignment between current and ideal job characteristics Dimension Gender Male Female Age 18–29 30–39 50–59 60+ Education Low Medium High Origin Belgium Industrialised countries Southern Europe Eastern Europe Morocco/Turkey Rest of the world Earnings First quartile Second quartile Third quartile Fourth quartile Job status Blue-collar worker White-collar worker Self-employed Wage-earning and self-employed Health Poor health Medium low Medium high Good

Challenging work

Physical effort

Autonomy

Intellectual effort

84.3 82.4

72.1 73.7

75.5 71.2

73.2 72.0

81.0 83.0 83.9 87.4

71.0 73.2 73.2 72.3

69.8 72.2 74.3 81.6

72.9 73.7 72.0 73.4

81.2 81.5 85.8

68.5 70.6 76.3

69.6 70.6 77.5

71.4 71.8 73.8

83.7 88.6 79.6 79.7 74.5 80.5

72.9 77.6 70.8 77.2 63.5 70.4

74.0 78.1 72.3 63.9 60.9 69.5

72.6 78.3 76.9 72.3 62.8 72.3

78.1 83.1 85.2 87.2

70.8 71.3 74.4 74.3

70.1 70.4 73.8 79.7

69.7 72.9 73.7 75.1

78.6 84.7 88.9 79.3

68.2 75.6 71.8 74.6

66.1 73.6 86.4 73.8

72.2 72.8 72.7 74.1

79.5 82.5 84.9 84.6

72.4 71.0 72.4 74.8

69.2 71.5 75.2 74.7

70.6 72.3 71.6 73.6

72

9

Do We Find the Job of Our Dreams?

challenging nature of their work. Good remuneration (we did not correct for differences in working hours here) is apparently also associated with a job that ties in more closely with what people expect from an ideal job. Here, of course, one could also wonder whether high earners are more likely to regard their job as ideal, or whether jobs that are better paid are also more likely to match people’s idea of a dream job. In any case, this is yet another example of what we have already described several times as cumulative deprivation: people who do a job that they do not regard as particularly appealing are also paid less for it. It seems that good health also helps people to find a job that matches their ideals. This also applies to the results for emotional well-being and chronic illness, which are not reported here.

What Do We Spend Our Money on?

10

In Chapter 1, we already emphasised that a family’s income—or the expenditure that a family can afford—is very important component of well-being. As explained in Chapters 4 and 5, income and expenditure do not fully coincide. In this chapter, we will discuss in more detail how Belgian families divide their expenditure between different categories of goods and services, as this offers an interesting view of their way of life. Consumption patterns vary widely between the different types of Belgian families. Our MEQIN dataset is not the only dataset in Belgium that contains information about household consumption patterns. For example, the Federal Public Service Economy has a long tradition of conducting budget surveys. The consumption data in the MEQIN dataset is much less detailed than the data in a standard budget survey. While budget surveys look into how Belgian families distribute their expenditure over many hundreds of goods and services, the MEQIN survey is limited only to some broader categories of non-durable goods and services. However, the MEQIN dataset is also more comprehensive than standard budget surveys as it contains information about which family members consume what in each of the surveyed categories. Chapter 15 deals extensively with this aspect of distribution within the family.

What Do the Consumption Patterns of Belgian Families Look Like? Families spend their income on durable and non-durable goods. Durable goods, for example, include cars or televisions and are characterised by a service life that sometimes lasts many years. Non-durable goods, such as food, personal care products or clothing, have a shorter service life. As a result, people spend money on

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_10

73

74

10 What Do We Spend Our Money on?

these goods on a regular basis. The same applies to some goods that, although durable, are associated with regular user costs, such as the Internet or telephone, and rent or mortgage payments for housing. As a result of this regularity, expenditure on non-durable goods and services often provides a more accurate picture of the level of welfare of families or individuals than the total expenditure in a given (short) period. After all, this is subject to major fluctuations as durable goods are purchased less frequently. Figure 10.1 shows how the total expenditure on non-durable goods and services for different family types is distributed between seven categories of goods and services. The height of the bars reflects the share of these categories in the total budget, and the figures above the bars show the average expenditure in monetary terms (in euros per month). The difference between these two approaches is essential. For example, the figure shows that single-person families have lower expenditure in most categories than the other family types, even if the share of this category within their spending pattern is fairly high. For example, this effect is very pronounced for the housing category: compared to other family types, single-person families have the lowest expenditure. At the same time, however, their expenditure on housing takes up the largest share of their budget. Of course, these two findings are not contradictory. They are easily explained by the fact that the overall level of expenditure of single-person families is lower. Although we will continue to focus on the shares of the various categories in this chapter, for a full interpretation, one should also bear the total amounts in mind. The figure clearly shows that current expenditure on housing (more specifically rent, energy and utilities and insurance) consumes most of the family budget.1 In Chap. 3, we discussed the classification of families into seven different family types. For each of the seven family types, the data shows that the broadly defined category of housing accounts for over 40% of the total expenditure on non-durable goods and services. Single-parent families and single-person families spend the most on housing, relatively speaking, with average budget shares of 49% and 56%, respectively. In themselves, these shares do not reveal much about the quality of the housing, although this quality is an important component of well-being. We will therefore discuss the quality of the housing in more detail in the next chapter. Food (including drinks) also takes up a significant chunk of the family budget for non-durable goods and services. Here, there is less variation across the seven family types. On average, they all spend around 18% of their total expenditure on non-durable goods and services on food and drink. There is more variation across the different family types when it comes to relative spending on recreation. In addition to recreation, this also includes expenditure on holidays, culture and hotels,

1

For homeowners, we took the imputed rental price into account. This is an estimate by the reference persons of the amount they think they would get if the property was rented out on the private rental market. By means of this choice, we aimed to provide an idea of housing consumption rather than the actual expenditure on people’s own homes.

What Do the Consumption Patterns of Belgian Families Look Like?

1218 1321 1098 1152 1019 1105

839

60%

75

50% 40%

176 90 19

154

110 258 255 250 304 164 274

516 286 408

61 112 168 149 165 107 180

44 116 146 141 128 73 133

10%

202 452 570

256 478 569 465 483 383 505

20%

614

30%

0% Food

Clothing

Housing

Transport

Recreaon and Child-related restaurants expenditure

Other

Single people

Married couple without children

Married couple with children

Unmarried couple without children

Unmarried couple with children

Single-parent family

Other family type

Fig. 10.1 Budget allocation across the various family types

restaurants and cafés. With an average budget share of 22%, cohabiting couples without children appear to spend the most on recreation in relative terms. Single-parent families and single-person families spend the least on this category, however, with an average budget share of 12%. The budget shares for transport (expenditure on petrol or diesel and public transport costs) and clothing (including shoes) are lower than the budget shares discussed above. They fluctuate at around 5% for the various family types, although there is some variation across the types. Families with children (people below the age of 18) appear to spend an average of 5% of their total expenditure on non-durable goods and services on child-related expenses. There is little variation between married and cohabiting couples with children or single-parent families. Finally, most family types spend an average of just under 10% of their budget on other non-durable goods and services. Among other things, this includes personal care items, (continuing) education and tobacco products. The differences revealed in Fig. 10.1 can be explained by various factors. Firstly, there are differences in needs and preferences. It should not be surprising that families without children do not incur child-related expenses, or that cohabiting couples without children spend a relatively high amount on recreation. In addition, however, income differences also play a role here. The share of necessary expenditure (such as food and housing) decreases as disposable income increases, while the share of expenditure on luxuries (such as recreation) increases. We will now examine the latter effect in more detail.

76

10 What Do We Spend Our Money on?

Expenditure Patterns per Income Quartile

1041 1327

803

60%

924

To gain a proper understanding of the relationship between spending patterns and income, it is important to take the size of the family into account once again. To this end, we apply the same correction as in Chaps. 4 and 5 and divide the total family income by the standard OECD equivalence scale to adjust the family income for family size. Whenever we mention income in this chapter, we are referring to this adjusted income. Figure 10.2 shows the average distribution of total expenditure on non-durable goods and services for each income quartile. In order to achieve this, we first ranked all the families in income order, from the family with the lowest income in the MEQIN dataset to the family with the highest income. The families were then divided into four equal groups (quartiles), in such a way that the first quartile contains the 25% of families with the lowest income (i.e. the dwarves passing by during the first fifteen minutes of the parade of giants and dwarves from Chap. 4) and the fourth quartile contains the 25% of families with the highest income. Figure 10.2 confirms the subdivision into necessary goods and expenditure on luxuries. It clearly shows that the average share of the budget spent on housing (including energy, utilities and insurance) decreases as the family income increases. For example, the 25% of families with the lowest income spend an average of 55%

50%

40%

114 37

27 46 44

138 197 303

396 661 228

139

84 107 152

71

45

10%

62 93 154

339 389 503

20%

296

30%

0% Food

Clothing Quarle 1

Housing Quarle 2

Transport

Recreaon and Child-related restaurants expenditure Quarle 3 Quarle 4

Fig. 10.2 Budget allocation according to income group

Other

Expenditure Patterns per Income Quartile

77

of their total expenditure on non-durable goods and services on housing (according to our broad definition), while this figure only comes to 44% for the 25% of families with the highest income. We can also see that families spend a relatively smaller proportion of their budget on food as their income increases. For example, families in the first income quartile spend an average of 20% of their total expenditure on non-durable goods on food, compared to an average of 17% for families in the fourth quartile. As we have already emphasised, these figures do not mean that the families with the highest incomes spend less money on housing or food each month; on the contrary, in fact. In monetary terms, richer families spend more on housing and food than poorer families. We note a reverse spending pattern when it comes to recreation and clothing. While the 25% of families with the lowest income spend an average of 8% and 3% of their budget on recreation and clothing, this comes to 20% and 5%, respectively, for the 25% of families with the highest income. There appears to be less variation according to income for the remaining three categories of goods (transport, child-related expenditure and other expenditure).

Expenditure Patterns by Level of Education Figure 10.3 shows the spending patterns of families classified according to the level of education of the reference person in the MEQIN dataset.2 The figure is based on the same breakdown into three levels of education as in previous chapters. If we compare Figs. 10.2 and 10.3, it is striking that both figures reveal very similar spending patterns. For example, families with a highly educated reference person appear to spend a relatively smaller amount on housing and food than families whose reference person has a low level of education. On average, however, the budget shares for recreation and clothing rise as the level of education of the reference person increases. Of course, this similar pattern can largely be explained by the fact that, on average, families with a more highly educated reference person

2

The classification of families based on the characteristics of the reference person should be interpreted with caution, as to some extent, the choice of reference person within the family is arbitrary. Nonetheless, some useful insights can still be gained from Fig. 10.3.

78

10 What Do We Spend Our Money on?

50%

1312

981

856

60%

40%

568

173

260

141 65

39

16

141

71

122

84

63

10%

104

201

332

473

340

20%

362

30%

0% Food

Clothing Low

Housing

Transport

RecreaƟon and Child-related restaurants expenditure

Medium

Other

High

Fig. 10.3 Budget allocation according to level of education

will also have a higher income. However, this average trend does not apply to all families. For example, a family whose reference person has a low level of education can have a family income which places it in the fourth quartile of the income distribution.

Do We Live Comfortably and in a Pleasant Environment?

11

Another important aspect of individual well-being is the quality of the housing and the living environment in which people spend their lives. In Chap. 7, we have already seen the importance of this aspect for the health of Belgians. Although we saw in the previous chapter that the recurring expenditure on housing forms a very significant proportion of the total family budget, this information is not really enough to say anything about the quality of the housing. Indeed, an uncomfortable home in an unpleasant environment can also lead to high current expenditure. In order to zoom into the quality of the housing, we also asked the reference person in each family some specific questions about the housing and living environment in the MEQIN survey. In this chapter, we look at the answers to these questions and then compare the answers for tenants and owners, as well as young people and the elderly.

The Quality of Our Housing We examine housing quality on the basis of five dimensions. Firstly, we look at the characteristics of the housing itself. This includes problems such as damp walls or a lack of space. We then look at the living environment: pollution, quality of cycle paths and availability of areas where children can play. A third dimension is the proximity of services such as a GP or post office. A fourth dimension concerns feeling unsafe. Finally, we also describe the quality of social relationships in the area. Table 11.1 provides information about the percentages of people for whom the dimension concerned does not form a problem. For example, the first row of the table means that 95% of the respondents disagree that there is a water leak in their home. At first glance, we can see that most of the problems relate to the quality of the living environment. Only 57% of people have no problem with the availability © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_11

79

80

11

Do We Live Comfortably and in a Pleasant Environment?

and quality of the cycle paths in the district. Here, we only look at the people who answered that the question is relevant to them. The second most problematic dimension is the lack of safe facilities where children can play.1 The quality of the housing itself seems to be the least problematic aspect. In addition, many people have no issues with social relationships in their neighbourhood. When it comes to the availability of services, the picture is more varied. Some services do not seem to be a problem; this applies in particular to schools and GPs. We already stated in Chap. 8 that GPs in Belgium are very accessible to the majority of the population. However, post offices appear to be less accessible.

Tenants and Owners Table 11.1 only shows average statistics, and it is interesting to look at the differences in housing quality for different types of families. We first look at the difference between owners and tenants. Belgium is a country of owners: 67% of families own their home, while 31% rent.2 We can reasonably assume that owners have chosen to be owners, while some tenants would prefer to be owners. By comparing the quality of the housing, we can at least partially understand what motivates people to choose a particular home when they become owners. Table 11.2 shows the differences in the percentages of owners and tenants who are satisfied with the various quality dimensions of their home. The first row, for example, means that 1.6% points more owners than tenants have no problem with a leaking roof. Most aspects of the housing itself pose fewer problems for owners than for tenants even if certain differences are negligible, as is the case of the availability of hot running water. The differences are greatest when it comes to the lack of outdoor space (such as a garden, courtyard or terrace): 16.5% points more owners than tenants feel that they have sufficient outdoor space. Further down in the table, we can see that the lack of green spaces near the housing is more often regarded as a problem by tenants than by owners (13.4% points more in the case of tenants). These two elements suggest that people attach importance to sufficient outdoor space when buying a home. This may relate to the exodus of owners from the city centre to a more rural environment or the outskirts of the city. The differences in the quality of the living environment are less pronounced. The lack of good footpaths is more of a problem for owners than for tenants, and the same applies to cycle paths. These differences may also relate to the fact that owners generally live in a more rural environment. 1

Again, we only take into account the answers given by people for whom the children’s play infrastructure is relevant. 2 The remaining 2% live free of charge. To keep things simple, we do not include them in the analysis.

Tenants and Owners

81

Table 11.1 Percentage of people for whom specific aspects of the housing are not problematic Quality of the housing itself Leak in the roof Damp on the walls or floors Woodwork damaged by rotting Unsuitable electrical installations Unsuitable plumbing installations/pipes No hot running water Too dark, not enough daylight Too little space inside Too little space outside Too much noise from neighbours or outside Living environment Pollution, litter or other environmental problems in the neighbourhood Vandalism, violence or crime in the neighbourhood Insufficient green spaces in the neighbourhood No footpaths, or problems relating to their quality No cycle paths, or problems relating to their quality Not enough parking spaces Not enough facilities where children can play safely Too much traffic in the immediate environment Availability of services nearby Grocer or supermarket General practitioner School (primary or secondary) Bank/cashpoint/self-banking Post office/Post Point Public transport Nursery, crèche Sense of safety Do you sometimes avoid certain places in the immediate vicinity because you don’t feel safe there? Do you sometimes not open the door when strangers call, because you don't feel safe? Do you sometimes not leave the house after dark? Neighbourly relations The neighbours help each other Sufficient protection of privacy Trust in the neighbours People in the neighbourhood usually get along well No conflicts between neighbours

95% 87% 96% 94% 95% 98% 93% 93% 91% 85% 86% 87% 88% 77% 57% 70% 58% 61% 89% 91% 93% 85% 78% 84% 84% 93% 84% 86% 86% 94% 92% 91% 80%

82

11

Do We Live Comfortably and in a Pleasant Environment?

Table 11.2 Differences (in percentage points) between owners and tenants for whom specific aspects of the housing are not problematic Quality of the housing itself Leak in the roof Damp on the walls or floor Woodwork damaged by rotting Unsuitable electrical installations Unsuitable plumbing installations/pipes No hot running water Too dark, not enough daylight Too little space inside Too little space outside Too much noise from neighbours or outside Living environment Pollution, litter or other environmental problems in the neighbourhood Vandalism, violence or crime in the neighbourhood Insufficient green spaces in the neighbourhood No footpaths, or problems relating to their quality No cycle paths, or problems relating to their quality Not enough parking spaces Not enough facilities where children can play safely Too much traffic in the immediate environment Availability of services nearby Grocer or supermarket General practitioner School (primary or secondary) Bank/cashpoint/self-banking Post office/Post Point Public transport Nursery, crèche Feeling unsafe Do you sometimes avoid certain places in the immediate vicinity because you don’t feel safe there? Do you sometimes not open the door when strangers call, because you don't feel safe? Do you sometimes not leave the house after dark? Neighbourly relations The neighbours help each otherss Sufficient protection of privacy Trust in the neighbours People in the neighbourhood usually get along well No conflicts between neighbours

1.6 8.5 2.9 6.2 4.8 1.0 5.6 7.9 16.5 9.8 1.0 6.5 13.4 −6.0 −2.5 7.4 7.1 6.8 1.4 0.7 −0.2 −0.2 −2.5 −9.0 −3.9 4.5 5.5 7.7 7.8 7.2 5.1 6.7 4.3

Tenants and Owners

83

With regard to the availability of services in the area, the perceptions of tenants and owners are fairly similar. A significant difference only becomes apparent with regard to the proximity of public transport, to the detriment of owners. When buying a house, people seem to attach fairly low priority to easy access to public transport. In general, owners find relationships with their neighbours less problematic than tenants. This may be due to the fact that more spacious accommodation with sufficient outdoor space is less likely to lead to conflict between neighbours than living in an apartment might do, for example. It is important to emphasise that this data reflects the perception of the respondents and does not necessarily correspond to the objective situation. If the decision to buy your own home is generally accompanied by a feeling of greater satisfaction, it is highly likely that the same objective situation will be perceived more positively by owners than by tenants. This comment also applies to the differences in perception with regard to safety. Owners find safety less problematic than tenants. This may stem from a combination of factors that do not all work along the same lines. For example, we know that owners tend to be older than tenants and that older people tend to feel less safe (see also the last section of this chapter). A lack of safety, on the other hand, is mainly an urban phenomenon, and tenants are more common in the city. To summarise, it seems that home ownership is accompanied by an improvement in the quality of the housing itself, particularly with regard to more outdoor space. However, both the availability and quality of footpaths and cycle paths and the proximity of certain services, especially public transport, are more problematic for owners. Although owners feel safer and have better relationships with their neighbours, a more in-depth study is needed especially for these more subjective components to better understand the mechanisms at work here. Given the importance of ownership for the quality of the housing, it is interesting to find out who the owners are. There is a strong correlation between home ownership and family income. Nonetheless, our sample shows that an increase in monthly family income of 1000 euros at the time of the survey only increases the likelihood of home ownership by 4.4% points. This may be because the decision to become an owner, and especially when to become an owner, depends more on income expectations over the entire further life cycle than on current income. Since we only have data for a single moment in time, we are unable to accurately determine evolutions throughout the further course of the respondents’ lives. However, we can gain an initial useful insight by examining the relationship between age and the status of owner. At a constant income and for a given family size, a family whose reference person is between 30 and 39 years of age is 8.7% points more likely to become an owner than a family whose reference person is below the age of 30; this difference becomes 25.7% points if the reference person is between 60 and 69 and 31.9% points for people between the ages of 70 and 79. We also note that at a constant income and for a given age of the reference person, a family with one more family member is 3.7% points more likely to own the family home.

84

11

Do We Live Comfortably and in a Pleasant Environment?

Elderly People and Young People We can now further investigate the extent to which the perceived quality of the housing evolves as life progresses. As housing quality differs for owners and tenants, and as owners are usually older, the purely age-related effect can only be found after correcting for the influence of ownership status. In order to establish this purely age-related effect, we used the same technique (regression analysis) which was already discussed in Chaps. 7 and 9. Somewhat surprisingly, this purely age-related effect actually appears to be very limited. Firstly, there are various dimensions for which the perception does not change during the course of life. For example, this applies to a number of aspects relating to the quality of the housing itself such as a lack of daylight, the living environment, the quality of the footpaths or the availability of certain services (e.g. general practitioners) in the vicinity. Secondly, certain aspects of housing quality are actually less problematic for older people than for young people: they have sufficiently spacious housing, plenty of parking spaces nearby3 and plenty of green spaces in the vicinity. Other dimensions, on the other hand, are considered more problematic by people over the age of 70 than by young people. This mainly involves the local presence of supermarkets and post offices. Older people also tend to feel much less safe. The percentage of people aged from 70 to 79 who are afraid to leave their homes in the evening is 15.2% points higher than for younger people. This seems consistent with the idea that housing quality mainly changes when one becomes a homeowner, but otherwise remains fairly stable. This provides the impression that people stay in the same housing for a long time and that the change in perception of the quality of this housing is mainly determined by the evolution of their needs: when the children leave home, the problem of a lack of space becomes less serious, while problems relating to a perceived lack of safety and the availability of services in the neighbourhood are felt more strongly by older people.

3

The percentage of respondents aged between 70 and 79 who consider the lack of parking spaces to be a problem is 10.5 percentage points lower than for respondents under the age of 30.

How Do We Spend Our Time?

12

We do not all live for the same length of time, and the major socio-economic differences in life expectancy face us with a major social challenge. Of course, we cannot say much about this based on the MEQIN data. However, the time available to us each day is one of the few things in life that are completely equally distributed. Whether rich or poor, male or female, highly educated or otherwise, we all have just 24 h a day at our disposal. The strict limitation of the time available per day to 24 h ensures that time use plays an important role in the “production” of welfare and well-being. For example, people can earn an income by using their time in the form of market labour. Or they could spend their time developing their skills, for example, by learning a new language. It also takes time to maintain friendships or practise a hobby. The same applies to caring for a baby or taking children to the music academy or sports club. Clearly, the limited time available each day means that people have to make choices about exactly how they spend their time. In this chapter, we will take a closer look at how Belgian adults divide their time between different activities such as market labour, household tasks, childcare or leisure. We will do this for the various family types. In Chap. 14, we go one step further and focus specifically on time distribution between partners in couples. Our MEQIN dataset is not the only dataset in Belgium that contains information about people’s use of time. For example, the TOR research group at the Free University of Brussels (VUB) has a long tradition of collecting and analysing Flemish time-use data, and the Federal Public Service Economy also conducts time-use surveys. The biggest difference between these specialised surveys and the MEQIN survey is the survey method. Whereas the studies mentioned above focus on the detailed reporting of time use on a weekday and a weekend day by means of a diary, MEQIN measures the average amount of time that people devote to a number of broad and exhaustive categories in a typical week. Although the data we collected is less detailed, it is no less interesting as we can examine it in conjunction with the information on other dimensions of life such as income and expenditure. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_12

85

86

12

How Do We Spend Our Time?

Time Use of Men and Women by Family Type Figures 12.1 and 12.2 show the average time use of men and women in the different types of families that we looked at in Chap. 3. Here, we examine their use of time during a typical week. Everyone has 168 h (i.e. 7 times 24 h) at their disposal. Figure 12.1 shows that the time spent on market labour, including commuting, takes the biggest chunk out of the available time for many men. For example, married and cohabiting men with children living at home spend an average of one quarter of their time, 41.5 and 43.5 h per week, respectively, on market labour. Cohabiting men without children appear to spend an average of 38 h per week on market labour. These figures are substantially higher than the number of hours of market labour performed by single fathers (an average of 28 h per week), single men without children (an average of 21 h per week) or married men without children living at home (an average of 14 h per week). We must be careful when interpreting these figures, as the various family types are not homogeneous. The group of single men without children, for example, consists of both young people in their twenties who live alone and work full-time and retired single men who no longer work. The perhaps surprising differences between the group of married men without children living at home and cohabiting men with children living at home can also be explained by the different characteristics of the two types of families. For example, a substantial proportion of married men without children living at home are no longer working, whereas cohabiting men with children living at home are often younger. This is an obvious explanation for the higher proportion of market labour in the time budget of cohabiting men with children living at home. 70

Hours per week

60 50 40 30 20 10 0 Market labour (including commuƟng)

Childcare

Household tasks Personal care

Helping others Leisure acƟviƟes

Single people

Married without children

Married with children

CohabiƟng without children

CohabiƟng with children

Single father

Son over 18

Fig. 12.1 Time use of men by position within the family

Time Use of Men and Women by Family Type

87

70

Hours per week

60 50 40 30 20 10 0 Market labour (including commuƟng)

Childcare

Household tasks Personal care

Helping others Leisure acƟviƟes

Single people

Married without children

Married with children

CohabiƟng without children

CohabiƟng with children

Single mother

Daughter over 18

Fig. 12.2 Time use of women by position within the family

In addition to market labour, the time spent on leisure activities also occupies a relatively important place in men’s time budgets. According to Fig. 12.1, married men without children living at home spend an average of almost 31 h a week on leisure activities. Single men and cohabiting men without children have an average of 28 to 29 h of leisure time per week. The least leisure appears to be available for married or cohabiting men with children living at home. They spend an average of 18 and 16 h a week on leisure activities. This is less than single fathers, who have an average of 21 h of leisure time per week. We note relatively small differences in the time spent by men on personal care (between 9 and 13 h per week on average). However, there are relatively large differences in the time spent on childcare (caring for people in the family who are below the age of 18). For example, cohabiting men with children living at home spend an average of 11 h a week on their offspring. This is almost twice as much as married men with children living at home (6.5 h a week) and single fathers (5.5 h a week). The relatively small number of hours per week spent on childcare by single fathers can be partly explained by co-parenting and visiting arrangements with the mothers of their children. There are also relatively major differences when it comes to household tasks. For example, cohabiting men with children living at home appear to spend an average of 7 h a week on household tasks, while this amounts to 13.5 h a week for single fathers. We will now examine the time-use patterns of women in the various family types. Figure 12.2 shows that market labour (including time spent commuting) accounts for a smaller percentage of women’s time budgets. Cohabiting women with children living at home spend most of their time on market labour (an average of 31.5 h per week), followed by cohabiting women without children living at

88

12

How Do We Spend Our Time?

home (28.5 h per week) and, at a slightly greater distance, by married women with children living at home (26.5 h per week). With an average of 23.5 h of market labour, single mothers come close to the figures for married women with children living at home. If we compare Figs. 12.1 and 12.2, we can see that on average men spend more time on market labour than women. This applies to all family types. These findings are consistent with the employment results described in Chap. 9. Women also spend a relatively large amount of time on leisure activities. For example, single, cohabiting or married women without children spend between 26 and 28 h a week on average on relaxation and leisure. This is much more than the time spent on leisure activities by cohabiting or married women with children living at home, who report an average of 15 and 16 h a week. It is interesting to note that, within each family type, on average men have more hours of leisure time per week than women. This is despite the fact that on average, men spend more time on market labour. Let us therefore take a closer look at the time allocated by women to other time-use categories. Figure 12.2 shows that household tasks and personal care take up a significant proportion of women’s time. While the time spent on personal care does not vary greatly across the different family types (between 11 and 14 h per week), there is slightly more variation when it comes to the time spent on household tasks. The average of 24 h spent on these tasks per week by married women with no children living at home is particularly striking. This outlier can be partly explained by the fact that on average, this group of families consists of older women who are no longer working (or have never worked). Compared to Fig. 12.1, it is striking that the average man spends less time on household tasks than the average woman. Finally, Fig. 12.2 reveals major differences in the time spent on childcare. As in the case of men, cohabiting women with children spend the most time on their offspring, with an average of 18 h per week. Married women with children living at home spend an average of 10 h a week on childcare, while for single mothers, it is around 8 h. In all cases, this is more than their male counterparts. We will further examine these striking differences between men and women in Chap. 14.

Time Use of Men and Women by Level of Education Time-use patterns do not relate just to family type. This is clearly illustrated by Fig. 12.3, which shows the average time use of men according to their level of education. Highly educated men spend more than twice as much time on market labour (including commuting) as low-skilled men, on average 35 h versus 16 h per week. These differences are only partly reflected in a smaller number of hours spent on leisure activities, and hardly at all in a smaller number of hours spent on household tasks, childcare and personal care. Overall, it appears that highly educated men have almost 20 h less time per week for sleep or rest (and other activities which we do not discuss) than men with low levels of education.

Time Use of Men and Women by Level of Education

89

70

Hours per week

60 50 40 30 20 10 0 Market labour (including commuƟng)

Childcare

Household tasks Personal care

Low

Medium

Helping others Leisure acƟviƟes

High

Fig. 12.3 Time use of men by level of education

Figure 12.4 shows the time use of women categorised by level of education. We can see relatively similar patterns to those in Fig. 12.3 for men. However, the absolute figures differ. Whereas highly educated women spend an average of 29 h a week on market labour (including commuting), this figure is only 7 h a week for low-skilled women. As in the case of men, this difference is not fully offset by the fact that women with lower levels of education spend more time on household tasks and leisure activities. Furthermore, women with lower levels of education also spend less time on childcare than highly educated women (an average of 2 h per week versus 6.5 h per week).

70

Hours per week

60 50 40 30 20 10 0 Market labour (including commuƟng)

Childcare

Household tasks Personal care

Low

Medium

Fig. 12.4 Time use of women by level of education

High

Helping others Leisure acƟviƟes

90

12

How Do We Spend Our Time?

The women in Fig. 12.4 naturally differ in several dimensions and therefore not just in terms of educational level. For example, the group of low-skilled women includes a substantial group of women who no longer have children living at home, which is reflected in the lower average number of hours spent on childcare. In general, the differences between women for each level of education (Fig. 12.4) are greater than the differences between men for each level of education (Fig. 12.3). One explanation for this is that the level of women’s education is strongly linked to their age, which is less the case for men.

Part II

An Insight into Families

In Part I, we followed the most common approach in which income and expenditure are analysed at a family level, while correcting this income for family size by means of an “equivalence scale”. However, we paid no attention to the distribution of income or expenditure within the family. We also described the average time use of both men and women, but without making an express comparison of the relative time use of men and women within the same family. In this part, we go one step further and try to gain an overview of the distribution and power relations within the families. With regard to the distribution of both time and expenditure, the MEQIN data makes this analysis possible in Belgium for the first time. Before we go into the distribution within families, however, we will first examine how these families are formed.

Who Forms a Couple with Whom?

13

In Chap. 3, we saw that a wide variety of family types exist in our country. Although single person households make up the largest group of Belgian families, many families consist of two partners who may or may not be married. It goes without saying that before two partners come together to form a family, they must first meet. In this chapter, we look at how Belgian couples are formed. We will start by addressing the question of how the partners first met. These days, people look for a suitable partner in completely different ways than was the norm a few decades ago. Whereas 20-somethings in the 1970s could never have foreseen the existence of the internet, in 2018, they take full advantage of social media and online dating apps such as Tinder to find a partner. We then examine to what extent the partners are alike and whether there is an assortative mating. Are highly educated women more likely to form a couple with highly educated men or men with a lower level of education?1 Finally, we examine whether these kinds of assortative mating occur more or less frequently among younger or older couples.

How Did Partners First Meet? Figure 13.1 shows how partners in a married or cohabiting couple first met. As the dating landscape has changed so much in recent years, this is shown separately for different age groups.2 Here, we only look at their current relationship. For example,

1

In this chapter, and the rest of this part of the book, we only look at heterosexual couples consisting of a woman and a man. This is because of the limited number of gay couples in our sample. 2 The age category is based on the age of the reference person in the couple in question. This does not preclude the reference person’s partner being in a different age category. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_13

93

94

13

Who Forms a Couple with Whom?

40% 35% 30% 25% 20% 15% 10% 5% 0%

Below 30

Between 30 and 40

Between 40 and 50

Between 50 and 60

Between 60 and 70

Over 70

Fig. 13.1 How did partners first meet?

a 75-year-old man who remarried after a divorce may have met his new partner on the Internet, although this was not yet possible in 1960. We first look at the dating behaviour of married or cohabiting people below the age of 30. According to Fig. 13.1, most of the couples in this group (around 29%) first met through friends. Many of the young couples (26%) met for the first time during a social activity (such as a group trip). Around 14% of couples up to the age of 30 first looked each other in the eye in a bar or café and 6% met online. Finally, about 7% of the couples met for the first time in secondary school. Studying together at college or university or working together appear to be less common ways for the youngest group of couples to meet. The way in which partners between the ages of 30 and 40 met does not vary greatly from that of individuals below the age of 30. For this group, too, meeting through friends is the most common method (around 25%). Respectively, 19% and 16% of these couples met in a bar or café or during a social activity. 8% of couples between the ages of 30 and 40 sowed the seeds of their relationship at college or university, at work or online. Partners between the ages of 40 and 50 mainly appear to have met in a bar or café. This applies to just over one in five of these couples. Social activities also constitute a common method of meeting people (18%). Many partners in the 40–50 age group met for the first time at work (15%) or through friends (14%). Couples with partners between the ages of 50 and 60 and 60 and 70 show relatively similar dating patterns. For example, the most common meeting place for this group of individuals is a bar or café. This is where 24% and 35% of couples, respectively, met for the first time. Around one-fifth of these two groups met through a social activity and approximately 10% met through work. For these

How Did Partners First Meet?

95

couples, the Internet hardly plays a role at all: less than 1% of the individuals in the age group between 50 and 70 met online. This is in stark contrast to the online dating behaviour of people below the age of 30. A quarter of the couples in the oldest group were formed after an initial meeting during a social activity and another quarter first met in a bar or café. The third-most common way of meeting for the first time appeared to be through friends (13%). It is clear that the way in which partners meet has changed greatly over time. Figure 13.1 illustrates in a striking way that meetings in a bar or café are much more important for older couples than for younger couples, and that the reverse applies to contacts via the internet or friends.

Like Seeks Like? We will now examine the extent to which partners in a couple are alike. If it were purely coincidental who forms a couple with whom, we could expect no correlation between the individual characteristics of the partners.3 In this case, it would be just as likely for a highly educated person to live with someone with a low level of education as it would for two highly educated people to form a couple. However, if there is a system behind who forms a couple with whom, we would expect a correlation between the characteristics of the individuals in couples. In principle, this could be both positive and negative. If there is a positive correlation, we are dealing with assortative mating. Figure 13.2 shows a series of correlations between the individual characteristics of partners. Here, we examine whether there is a correlation between the characteristics of partners within couples and whether this correlation is statistically significant. A statistically significant correlation indicates an identified pattern that is probably not based on pure coincidence, as would be the case if partners were to form a pair completely at random. Each bar in the figure represents the correlation coefficient between a certain characteristic of the partners in a couple. The greater the correlation coefficient, the stronger the correlation between this individual characteristic of the partners. The correlation coefficient is always a number between -1 (perfectly negative correlation) and 1 (perfectly positive correlation). Figure 13.2 shows that all the investigated correlations are positive. This means that only positive correlations were found between the individual characteristics of partners in a couple. Apart from the rather minor correlations for extraversion and emotional stability, the correlations are statistically significant which indicates that they are not based on chance. The strongest correlation is found for age. Indeed, it is not surprising that younger people are mainly with younger people and older people are mainly with older people. There are, of course, exceptions to this rule. For example, a more 3

For the sake of simplicity, we assume here that the relative sizes of the socio-demographic groups to which partners may belong are the same.

96

13

Who Forms a Couple with Whom?

BMI Age Weight Height Health Religious beliefs Openness Neuro cism/Emo onal stability Conscien ousness Agreeableness Extraversion Hourly wage Educa on 0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Correla on coefficient

Fig. 13.2 Correlation between characteristics of partners

detailed analysis shows that the average age difference between the man and the woman in a couple is about 2.5 years. However, this minor average age difference conceals relatively major variations. For example, the data includes couples in which the man is over ten years older than the woman and couples in which the woman is over ten years older than the man. In general, however, the strong positive correlation implies the existence of like-seeks-like behaviour with regard to the age of the partners. There is also a strong positive correlation between the levels of education of partners in a couple. Highly educated women are more likely to live with highly educated men, and women with low levels of education more often form a couple with men with a similar level of education. In this context, there is also a positive correlation between the productivity of working partners (measured in terms of their hourly wage). Men with a higher hourly wage are mainly in a relationship with women with a higher hourly wage, and men with a lower hourly wage are more likely to be with a woman with a lower hourly wage. The impact on the distribution of income between families is evident. The more we observe assortative mating in terms of productivity, the greater the inequality between families. The connections between the socio-economic characteristics of the partners may perhaps not be particularly surprising. However, there is also a strong positive correlation between their physical characteristics. For example, we find a positive correlation for height and weight and also for BMI, which we looked at earlier in Chap. 6. Incidentally, there is also a positive correlation for the self-reported health status of both partners. This probably has less to do with assortative mating, instead relating to the fact that partners in a couple live together and therefore share many conditions such as lifestyle, living environment and housing and thus develop a

Like Seeks Like?

97

similar health profile. We have already discussed the impact of such factors on health levels in Chap. 7. Furthermore, people who are not religious or less strongly religious tend to partner with people with similar beliefs, and more religious people generally tend to form a couple with people with similar religious beliefs. Finally, the partners within a couple generally also have similar personality traits. For example, there is a positive correlation in the degree of agreeableness of the partners; more open people are more likely to form a couple with other open people and more cautious or agreeable people tend to partner with other relatively cautious or agreeable people. However, the correlations are less strong than those we have discussed so far. They even become very weak and no longer statistically significant when it comes to extraversion and emotional stability.

Is Like-Seeks-Like Behaviour Age-Related? In Fig. 13.2, we examined the strength of the correlation between individual characteristics such as age, level of education or the BMI of partners in a couple. Figure 13.1 showed that the dating behaviour of young couples is very different from that of older couples. This raises the question as to whether like-seeks-like behaviour is also age-related. For example, we cannot rule out the possibility that some correlations between individual characteristics of partners are different for younger and older couples. 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 -0.10

Below 30

Between 30 and 40

Between 40 and 50

Between 50 and 60

Between 60 and 70

Fig. 13.3 Correlation between characteristics of partners by age group for the male partner in a couple

98

13

Who Forms a Couple with Whom?

To answer this question, Fig. 13.3 shows the correlation between individual characteristics of partners by age group for the male partner in a couple. Regardless of the age of the male partner, the similarities between the partners are greatest in terms of age, level of education and degree of religious beliefs. However, the extent of these correlations varies between age groups. For example, there is a greater correlation for the level of education in older couples than in younger couples.4 Furthermore, the correlation between the ages of partners in the younger and oldest couples is greater than the same correlation for couples consisting of individuals between the ages of 50 and 70.

This finding is not entirely in line with the results in the literature for other countries with more sophisticated techniques; for example, see Schwartz and Maré (2005) for the USA States. We therefore regard it as fodder for further research.

4

How Do Partners Within Couples Spend Their Time?

14

In Chap. 12, we reviewed the time-use patterns of all types of families. In this chapter, we will examine the distribution of time within couples in more detail. This distribution determines how much leisure time remains for each of the partners, which in turn affects the level of well-being of the individual partners. An uneven distribution of childcare or time spent on household tasks can also lead to tension between partners. In this chapter, we only look at couples in which neither partner is retired or on an early leavers scheme and who do not have children over the age of 18 living at home. These couples include a motley assortment of young and older couples, couples with and without children, couples in which both partners work and couples in which neither partner works. Later in this chapter, we will analyse the time use of various specific types of couples.

Time Use of Partners Across All Types of Couples For men and women in these couples, Fig. 14.1 shows the average number of hours per week spent on market labour (including commuting), childcare and household tasks. The last two bars represent the sum of these three activities for men and women. The remaining time is available for leisure activities, sleep and rest, study, helping others (e.g. providing informal care) or personal care. In the rest of this chapter, we will summarise all these activities under the somewhat misleading term “leisure”. The men in the couples spend an average of 44 h a week on market labour, with women spending an average of 29 h a week. Among other things, these figures reflect the fact that women tend to work part-time more than men. We can also see that women spend more time on childcare and other household tasks than men. For example, women spend an average of 9.5 h a week on childcare and 17.5 h a week on household tasks. For men, the figures are 6 h and 7 h a week, respectively. If we © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_14

99

100

14

How Do Partners Within Couples Spend Their Time?

70 60

Hours per week

50 40 30 20 10 0 Market labour (including commung)

Childcare Men

Household tasks

Total

Women

Fig. 14.1 Time use of partners across all types of couples

look at the sum of the time spent on market labour, childcare and other household tasks, we note that men spend 57 h a week on this, while women spend around 56 h a week. On average, therefore, men and women have around the same amount of leisure per week. While men spend more time on market labour, women spend more time on childcare and household tasks. The figures in Fig. 14.1 show the averages for different types of couples. They comprise the time use of couples who are active and inactive on the labour market. In addition to families with children living at home, they also include families that do not have (or no longer have) children living at home. Here, we take a more detailed look at the impact of children living at home on the time use of individuals in couples in which both partners work.

Time Use of Working Couples with and Without Children Living at Home Figure 14.2 shows that men and women in working couples without children spend an average of 56 and 55 h per week, respectively, on market labour and household tasks. For working partners with children, the total time spent on market labour, childcare and household tasks reaches an average of 69 h per week for men and 73 h per week for women. It is striking, but perhaps not surprising, that working partners with children have much less time left for things other than housework and going to work than working partners without children. A more striking result is the varying effect of children on the time use of men and women. As the “total” bars on the right-hand side of the figure indicate, men and women in couples without

Time Use of Working Couples with and Without Children Living at Home

101

70

Hours per week

60 50 40 30 20 10 0 Market labour (including commuƟng) Men without children

Childcare Men with children

Household tasks Women without children

Total Women with children

Fig. 14.2 Time use of working partners in couples with and without children living at home

children have around the same amount of leisure. However, this changes when there are children living at home. The women in these couples have an average of 4 h less leisure per week than the men. It goes without saying that working partners in couples without children living at home do not spend time on children living at home. By contrast, working men in couples with children spend an average of 12 h a week on childcare. For working women, this figure is 17 h per week. We also note relatively large differences between working men and women when it comes to household tasks. On average, working men in couples with or without children spend about 7 to 8 h of their time on household tasks per week, while working women with or without children spend up to about 15 h per week. The conclusion is therefore clear. In working couples without children, the smaller number of hours spent by women on market labour is compensated by the larger number of hours spent on household tasks, with the result that the amount of leisure for men and women is approximately the same. In working couples with children, women also spend relatively more time on childcare, thus giving them less leisure than the man in the couple.

Time Use of Partners by Level of Education The level of education naturally also affects the time use of men and women. Figures 14.3 and 14.4 illustrate this effect. Once again, initially we make no distinction between people who work and those who do not. Figures 14.5 and 14.6 then show the time use of working men and women according to their level of education.

102

14

How Do Partners Within Couples Spend Their Time?

70 60

Hours per week

50 40 30 20 10 0 Market labour (including commung)

Childcare Low

Household tasks Medium

Total

High

Fig. 14.3 Time use of women by level of education

70

Hours per week

60 50 40 30 20 10 0 Market labour (including commung)

Childcare Low

Household tasks

Medium

Total

High

Fig. 14.4 Time use of men by level of education

Figure 14.3 shows that women with a higher level of education spend more time on market labour (including commuting). For example, low-skilled women spend an average of 14 h a week on market labour, women with a medium level of education spend an average of 25 h a week on market labour, and highly educated women spend an average of 38 h a week on market labour. The time spent on childcare also increases as the level of education rises. For example, women with a low level of education spend an average of 7 h a week on their children. For highly

Time Use of Partners by Level of Education

103

70 60

Hours per week

50 40 30 20 10 0 Market labour (including commung)

Childcare Low

Household tasks

Medium

Total

High

Fig. 14.5 Time use of working women by level of education

70

Hours per week

60 50 40 30 20 10 0 Market labour (including commuƟng)

Childcare Low

Household tasks

Medium

Total

High

Fig. 14.6 Time use of working men by level of education

educated women, this rises to an average of almost 11.5 h a week. Some caution should be exercised here, however, as the various groups also include women who do not have children or no longer have children living at home. However, a separate analysis shows that the same phenomena also occur if we restrict ourselves to couples with children living at home.

104

14

How Do Partners Within Couples Spend Their Time?

For time spent on other household tasks, we note the reverse picture. Women with a low level of education spend an average of 20 h a week on household tasks, compared to an average of 14 h a week for highly educated women. This pattern can partly be explained by the fact that partners who work full-time outsource some of their household tasks, such as cleaning or maintenance. However, the reduction in the number of hours spent on household tasks does not compensate for the increase in the number of hours spent on market labour and childcare for highly educated women. After all, the bars on the far right in Fig. 14.3 show that women with a low level of education spend an average of 42 h a week on the sum of market labour, childcare and other household tasks, while women with medium and high levels of education spend an average of 54 h and 63 h a week, respectively. As a result, the amount of leisure available decreases as the level of education increases: women with a low level of education have more leisure than highly educated women. However, we must not forget that what we call “leisure” here also includes time spent on studying, personal care and caring for others. Figure 14.4 shows that the pattern for men is broadly the same as for women. Men with a lower level of education spend an average of 34 h a week on market labour, including time spent commuting. The amount of time spent on market labour increases to an average of 49.5 h per week for highly educated men. For men too, the average number of hours spent on children increases as the level of education rises, albeit somewhat more slowly than for women. For example, men with a low level of education tend to spend just over 4 h a week on their children, while this figure rises to 8 h a week for highly educated men. For household tasks, however, the man’s level of education has little impact: the average fluctuates between 7 and 8 h per week. The overall impact of educational attainment for men and women is also similar for the three categories of time use. More highly educated men spend more time on market labour, childcare and other household tasks and therefore have less leisure than men with lower levels of education. However, the differences are somewhat less pronounced than for women. Figures 14.3 and 14.4 provide further comparisons between the time use of men and women. For each level of education, men spend more time on market labour than women. In terms of time spent on childcare and household tasks, the reverse is true: regardless of their level of education, men spend less time on their children than women with a similar level of education. In the previous two figures, we examined the effect of educational attainment on time use without distinguishing between working and non-working men and women. In Figs. 14.5 and 14.6, we focus more specifically on the time use of men and women in couples in which both partners work. Figure 14.5 shows that the impact of educational attainment on the number of hours spent by women on market labour almost disappears. For example, low-skilled working women spend an average of 40 h per week on market labour, working women with medium levels of education spend an average of 38 h per week and highly educated working women spend an average of 42 h per week. The much stronger impact in Fig. 14.3 is probably explained by the fact that a higher percentage of low-skilled women are not active in the labour market or are

Time Use of Partners by Level of Education

105

unemployed. This phenomenon may also explain why the impact of educational attainment on household tasks continues to exist for working women but is weaker than the average for all women. Working women with low levels of education spend an average of 16 h a week on household tasks, whereas the figure is 13 h a week for highly educated working women. Even if we only look at working women, the time spent on children increases in line with the level of education. While working women with a low level of education spend an average of 6 h a week on childcare, this rises to 12 h a week for highly educated working women. Looking at the sum of all these activities, we note once again that the amount of leisure is lower for highly educated women than for women with a low level of education. However, the impact of educational attainment on the amount of leisure is much less pronounced in Fig. 14.5 than in Fig. 14.3. Whether women are active or inactive on the labour market has a stronger impact on their leisure, relatively speaking, than their level of education. Figure 14.6 shows the time use of men in working couples according to their level of education. The patterns here are less pronounced than those for working women. For example, working men with a low level of education spend an average of 47 h a week on market labour and working men with a medium level of education spend an average of 50 h a week, while highly educated working men spend an average of about 49 h a week. Working men also spend between 7 and 8 h a week on household tasks, regardless of their level of education. However, the man’s qualifications have a relatively strong impact on the average number of hours spent on childcare. Working men with a low level of education spend an average of 5 h a week on their children, rising to 8 h a week for highly educated working men. As for women, the increase in the number of hours spent on market labour and childcare is also reflected in the amount of leisure remaining. On average, highly educated men have almost 4.5 h per week less leisure than men with lower levels of education.

What Do Partners in Couples Spend Their Money on?

15

In Chap. 10, we discussed the average spending patterns of different types of families. More specifically, we illustrated how the total expenditure on non-durable goods and services is divided between food, clothing or housing, for example. However, this distribution of the family’s total expenditure between various different goods and services does not reveal who these goods are intended for. Nonetheless, this distribution of consumption within the family is crucial for the individual well-being of the family members. The MEQIN dataset enables various conclusions to be drawn about this distribution for Belgium. In this chapter, we examine the distribution of expenditure among partners in the same couples whose time use we described in the previous chapter. These are the couples in which neither partner is retired nor on an early leavers scheme and who do not have children aged 18 or above living at home. For the purpose of analysing these budget allocations, it is helpful to classify the goods and services differently than in Chap. 10. Here, we focus on four categories of spending on non-durable goods and services. The first two categories of expenditure are the private expenditure of men and women in couples. Private expenditure only benefits the person for whom it is intended. Examples include expenditure on clothing, food, cigarettes or café visits and other leisure activities outside the family context. The private nature of this expenditure can be illustrated by the fact that the clothes of the woman in a couple are not worn by the man (with some exceptions), or that every bottle of cola drunk by the man can no longer be drunk by his female partner. A third category of expenditure involves expenditure for any children living at home. This includes pocket money, school expenses or spending on clothing or hobbies. In principle, this is also private expenditure that benefits the children. However, we will consider it separately as parents often finance the lion’s share of this expenditure. In addition, it can be assumed that both parents in a couple are concerned about the well-being of their children. As the well-being of the children partly depends on the money spent on them, both parents within the couple benefit from this expenditure. In contrast to the private expenditure described above, it is © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_15

107

108

15 What Do Partners in Couples Spend Their Money on?

not the case that more consumption by one partner implies less consumption by the other. In economic jargon, money spent on children is therefore public (and not private) expenditure. Finally, we have a fourth category of expenditure on other non-durable goods and services. These are goods whose consumption is of a public nature: they benefit all the family members either in full or in part. One example would be the family home. All the family members use the bathroom, living room or kitchen to a greater or lesser extent. To provide an idea of the relative importance of the consumption of housing services compared to that of other goods, we used the rental price of the property for owners and tenants alike. For tenants, this is the actual rent, and for owners, it is the virtual or imputed rent, i.e. the rent that the reference person thinks could be obtained if the property was rented out.1 Other examples of such public expenditure include expenditure on utilities such as gas and electricity and common expenditure such as restaurant visits or family holidays.

Expenditure Within Couples with and Without Children Figure 15.1 shows the average monthly spending for the four categories of expenditure. The first bar always examines the situation of all the couples, while the next two bars zoom in on couples with and without children living at home. The construction of this figure (and the other figures in this chapter) is different from the way in which the figures were presented in Chap. 10. This time, the bars represent the expenditure in euros per month, and the relative percentages are displayed above each bar. The average total expenditure on non-durable goods and services across all types of couples comes to 2888 euros per month. In the case of couples without children, the average expenditure is 2710 euros per month, while couples with children spend an average of 3026 euros per month. The figure also shows that most of the expenditure is spent on public goods such as housing, utilities and other common expenses within the family. This expenditure amounts to an average of 1681 euros per month and accounts for 58% of the total budget spent on non-durable goods or services. Couples with and without children spend an average of 1743 euros (57%) and 1602 euros (59%), respectively, on housing and other public goods each month. The higher amount for couples with children partly reflects the fact that families with children usually live in larger houses or apartments. In addition, these couples with children spend an average of 398 euros per month on child-related expenditure. This represents approximately 13% of the expenditure on non-durable goods or services.

1

Using the mortgage payments would give the wrong impression that the consumption basket of families who no pay off a mortgage does not include housing. See also Chap. 10.

100%

100%

3000

100%

109

3500

57%

58%

2000

59%

2500

13%

8%

20%

14%

17%

500

21%

1000

15%

1500 18%

Expenditure in euros per month

Expenditure Within Couples with and Without Children

0 Private Private Child-related expenditure for expenditure for expenditure men women All couples

Couples without children

Other public expenditure

Total expenditure

Couples with children

Fig. 15.1 Budget allocation across four categories of expenditure

On average, there does not appear to be a significant difference between the private expenditure of men and women (507 and 488 euros per month, respectively). The percentage of private expenditure for both partners is around 18% of the total expenditure on non-durable goods and services across all couples. On average, the private expenditure of men and women in couples with children living at home is lower than that of couples without children living at home. The average difference is 91 euros per month for men and 106 euros per month for women. The higher total expenditure of couples with children is therefore not sufficient to compensate for the higher expenditure on children and public goods such as housing. The relatively similar average private expenditure of men and women seems to suggest that it is unnecessary to focus too heavily on the distribution between partners in a couple. However, this would be a hasty conclusion. There does appear to be a great deal of variation between couples in terms of the percentage of the total private expenditure that goes to women. In some couples, the percentage is low, while in others, it is high. If the woman has a 50% share, this means that half of the total private expenditure goes to her. To illustrate this, we carried out the following exercise (shown in Fig. 15.2). For each couple, we calculated how much of the total private expenditure went to the woman. We then ranked all the couples according to the percentage that went to the woman, from the lowest percentage to the highest. Finally, we divided the couples into four equal groups. The first group contains the 25% of women with the lowest percentage of the private expenditure, while the fourth group contains the 25% of women with the highest percentage. In the first group, an average of 36% of the total private expenditure appears to go to the woman (and therefore 64% to the man). In the fourth group, women account for an average of 62% of total private expenditure. As a result, there is a wide variation in the percentage of total private expenditure that goes to the women. We will return to this in Chap. 16.

110

15 What Do Partners in Couples Spend Their Money on?

70% 60% 50% 40% 30% 20% 10% 0% Quarle 1

Quarle 2

Quarle 3

Quarle 4

Fig. 15.2 Share of women’s private expenditure in total private expenditure

In the following figures, we will examine the average private expenditure of men and women and expenditure on children and other public expenditure for families with different levels of total expenditure and for partners with different levels of education. For these figures, too, one must always bear in mind that the average expenditure conceals a wide variation both between and within families.

Expenditure Within Couples According to Total Expenditure Figure 15.3 illustrates the budget allocation within couples for couples with different levels of expenditure. To come to this figure, this time we first ranked all the couples from the couple with the lowest total expenditure on non-durable goods and services to the couple with the highest total expenditure. Once again, we then divided the ranked couples into four equal groups. The first quartile contains the 25% of couples with the lowest expenditure on non-durable goods and services, while the fourth quartile contains the 25% of couples with the highest expenditure. The figure shows that the total expenditure on non-durable goods and services varies widely across the different expenditure quartiles. For example, couples in the first expenditure quartile spend an average of 1715 euros per month on non-durable goods and services. With an average of 4454 euros per month, couples in the fourth quartile spend much more than twice as much. This increase in expenditure also extends to all the individual categories. However, the difference is most pronounced when it comes to spending on public goods and services such as housing, utilities and other public expenditure. For example, couples in the first quartile spend an average of 963 euros per month on this category of goods, while couples in the fourth quartile spend an average of 2606 euros, or almost three times as much.

100%

111

5000 4500

56%

58%

8%

10%

7%

6%

16%

16%

17%

18%

500

16%

1000

17%

1500

17%

2000

59%

2500

100%

3000

100%

59%

3500

100%

4000

20%

Expenditure in euros per month

Expenditure Within Couples According to Total Expenditure

0 Private expenditure Private expenditure for men for women Quarle 1

Quarle 2

Child-related expenditure Quarle 3

Other public expenditure

Total expenditure

Quarle 4

Fig. 15.3 Budget allocation across four categories of expenditure per expenditure quartile

Child-related expenditure also increases with total expenditure. In the first quartile, an average of 95 euros per month is spent on children compared with an average of 349 euros in the fourth quartile. One could claim that this is because, on average, there are more families without children in the first quartile than in the fourth quartile. However, we also carried out the same calculations only for families with children and found that the effect persists. For example, families with children spend an average of 239 euros per month on their offspring in the first expenditure quartile, while this figure rises to 569 euros per month for families in the fourth expenditure quartile. There are relatively major differences between the private expenditure of men and women across the various quartiles. It is interesting to examine both the percentages (the figures above the bars) and the expenditure in euros (the height of the bars) in more detail. The private expenditure of men and women accounts for 20% and 18%, respectively, of the total expenditure for the couples in the first quartile. In the fourth quartile, the average percentage that goes to the man is 17%, with an average of 16% going to the woman. As the total family expenditure increases, the proportion of private expenditure therefore decreases slightly. This slight percentage drop is accompanied by major absolute differences. For example, the private expenditure of men and women is 341 and 316 euros per month, respectively, for couples in the first expenditure quartile, reaching 770 and 729 euros for men and women in the fourth quartile.

112

15 What Do Partners in Couples Spend Their Money on?

Expenditure Within Couples According to Level of Education The spending pattern of couples also varies according to the educational attainment of the partners in a couple. Figure 15.4 examines the effect of the man’s level of education on the budget allocation, while Fig. 15.5 does the same for the women. The figures show that expenditure in each of the four categories increases with the level of education, both for men and women. Once again, the differences are most pronounced for the category of expenditure on public goods such as housing. Whereas families in which the man has a low level of education (no certificate of secondary education) spend an average of 1235 euros per month on public goods, this rises to an average of 2079 euros per month for families in which the man has a high level of education (at least a degree of higher education). There is also a strong relative increase in child-related expenditure: an average of 179 euros per month is spent on children in families in which the man has a low level of education, with an average of 269 euros per month being spent on children in families in which the man is highly educated. Private expenditure also increases in line with the man’s level of education. However, the differences between the average private expenditure of men and women are not very pronounced. We have already discussed the fact that these averages may conceal major variations between couples.

100% 54%

2000

8%

7%

8%

16%

17%

19%

500

18%

1000

17%

1500

58%

60%

2500

100%

3000

100%

3500

19%

Expenditure in euros per month

4000

0 Private expenditure for men

Private expenditure for women Low

Child-related expenditure

Medium

Other public expenditure

Total expenditure

High

Fig. 15.4 Budget allocation across four categories of expenditure by level of education for men

Expenditure Within Couples According to Level of Education

113

100%

4000

59%

2500

100%

8%

7%

7%

16%

18%

18%

500

18%

1000

17%

1500

57%

56%

2000

100%

3000

20%

Expenditure in euros per month

3500

0 Private expenditure Private expenditure for men for women Low

Child-related expenditure Medium

Other public expenditure

Total expenditure

High

Fig. 15.5 Budget allocation across four categories of expenditure by level of education for women

It is hard to distinguish the picture for the women in Fig. 15.5 from that of the men in Fig. 15.4. Given the like-seeks-like behaviour we discussed in Chap. 13, this is of course not particularly surprising. Figure 15.5 also shows an increase in expenditure in each of the four categories if women have a higher level of education. Once again, the difference is most pronounced for expenditure on public goods.

Who Wears the Trousers?

16

In Chap. 14, we examined in detail how partners in couples divide their available time between market labour, childcare and household tasks. In Chap. 15, we illustrated how the total expenditure on non-durable goods and services is divided between individual private expenditure, expenditure on children (living at home) and other public expenditure relating to the family. Based on the figures in these two chapters, we can conclude that time and money are unevenly distributed between families and between partners within couples. We will try to explain these differences in this chapter.

Distribution of Time and Money Between Partners In this chapter, we will summarise the distribution of money and time for each couple into a single figure for the man and one for the woman. We achieve this by expressing the time and budget allocation of men and women in the form of euros per month. Since public expenditure within the family context benefits all the family members, we will ignore it here for the sake of simplicity. In order to reflect the value of leisure for each partner, we multiply the number of hours of leisure per month by the individual hourly wage. This method of attaching monetary value to leisure is very popular with economists. It is based on the idea that every hour spent on leisure or household tasks cannot be used on the labour market to earn extra income. Individuals, therefore, sacrifice the value of an hourly wage for each hour of leisure. This is sometimes referred to as an opportunity cost. As we stated in Chap. 14, in this context the number of hours of leisure also includes time spent on studying, personal care and caring for others. Allocating a monetary value to leisure allows us to add up the individual monthly private expenses for both men and women. The sum of the values for both partners then becomes the total monetary value of leisure and private expenditure within a couple. The male share is the monetary value of the leisure and private © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_16

115

116

16

Who Wears the Trousers?

70% 60% 50% 40% 30% 20% 10% 0% QuarƟle 1

QuarƟle 2

QuarƟle 3

QuarƟle 4

Fig. 16.1 Women’s share in the sharing rule for time and money within working couples

expenditure of the man divided by that of both partners together. Both shares are expressed as a percentage and add up to 100%. We refer to the percentage of leisure and money that goes to both partners as the sharing rule. It goes without saying that we can only calculate this sharing rule for couples in which both partners work on the labour market, since we can only observe the hourly wage for working respondents. We can now rank all the couples according to the woman’s share in the sharing rule. The higher this share, the more the woman wears the trousers in the couple. The lower the share, the more the man wears the trousers. As in the previous chapter, we can divide all the couples into four equal groups according to the woman’s share in the sharing rule. Here, we note a great deal of variation once again (see Fig. 16.1). In the first group, the average share of women in the sharing rule is 36%. This share is about 61% for families in the fourth group. It is interesting to compare these figures with those in the previous chapter, where very similar results were found for the distribution of private expenditure without taking the value of leisure into account.

Distribution of Time and Money in Relation to Relative Hourly Wage An alternative method of illustrating the unequal distribution of time and money within and between families can be found in Fig. 16.2. On the horizontal axis, this figure shows the hourly wage of the woman divided by the hourly wage of the man. This is called the relative hourly wage or wage rate. If the wage rate reaches a value higher than 1, the woman has a higher wage than the man in the couple. On the

Distribution of Time and Money in Relation …

117

80% 70%

Women's share

60% 50% 40% 30% 20% 10% 0% 0

0.5

1

1.5

2

2.5

3

Wage rate

Fig. 16.2 Women’s share in the sharing rule within working couples in relation with the relative hourly wage

left-hand side of the axis, we find women with a low wage compared to that of the man in the couple. On the right-hand side, we find the reverse situation: that of women with a high wage compared to the man in the couple. On the vertical axis, we show the woman’s share in the sharing rule. A low percentage means that the woman has relatively little leisure and individual consumption compared to the man in the couple. A high percentage implies an uneven distribution of time and money in the woman’s favour. A 50% share on the vertical axis implies that each partner in the family has an equal share of the total available resources for time and private consumption. Each dot on the figure shows the position of an observed couple in our data. Most of the observations hover around 1 on the horizontal axis, meaning that the hourly wages of women and men do not differ significantly. This was to be expected based on the like-seeks-like behaviour that we described in Chap. 13. However, in some observations the hourly wage of the woman is less than half that of the man (these are the observations closer to the vertical axis). Conversely, in some families the woman’s hourly wage is at least twice as high as that of the man (the observations further away from the vertical axis). Half of the observations on the vertical axis are between 43 and 54%. However, there are a few outliers in both directions. In 5% of cases, the woman’s share in the sharing rule is less than 30%. Conversely, women gain over 2/3 of the share in 5% of cases. There are slightly fewer outliers at the top than the bottom.

118

16

Who Wears the Trousers?

Figure 16.2 offers an important new insight. It shows that the higher the woman’s relative wage, the greater her share of the time and money available for private consumption. This is clearly illustrated by the scatter plot of observations and the dotted line that passes through it, representing the average trend. For example, the average share of women in families in which the woman earns half the hourly wage of the man is less than 35%. If we look at families in which the woman earns twice the hourly wage of the man, we see an average share of more than 64.4%. It is evident that there are relatively major differences in the way in which time and private consumption are distributed within the families. The pattern discussed above does not appear to be coincidental. Numerous studies within the field of family economics have found that the sharing rule depends on the partners’ relative wages. Incidentally, other variables also affect the male and female share of a family’s available resources. For example, the study carried out by Pierre-André Chiappori, Bernard Fortin and Guy Lacroix (2002) using American data is rather intriguing. Different US states sometimes have very different laws on divorce. These laws are relatively beneficial to women and more detrimental to men in some states, whereas the reverse applies in others. Furthermore, there is a relatively large variation in the gender ratio in the USA. This shows the proportion of males to females within a certain age group and ethnic group. Chiappori, Fortin and Lacroix (2002) demonstrated that the share that goes to women within families depends on how favourable the divorce laws are for women: the more favourable the legislation is towards women, the greater their share. The study also shows that the lower the gender ratio (i.e. relatively fewer women), the greater the share of time and private consumption that goes to women within a family.

Distribution of Time and Money in Relation to Full Family Income Figure 16.3 illustrates how the woman’s share of time and private consumption in working couples varies with the full family income. This full family income is defined as the sum of the maximum income that a couple could potentially earn per month if both partners spent all their available time on market labour at their current hourly wage, along with income from assets. It goes without saying that full family income is a theoretical concept and, for all kinds of reasons, is not actually achievable. However, the concept is extremely suitable as a measure of the potential purchasing power of families. The observed family income is much less suitable for this, partly because it depends on the choice of how many hours both partners in the couple work. The figure shows that the woman’s share does not vary greatly with the full family income of a couple. The dotted line showing the average trend plots a slightly downward course. This means that on average, the share of women in potentially poorer families is not much higher than the share of women in

Distribution of Time and Money in Relation to Full Family Income

119

100% 90% 80%

Wonen's share

70% 60% 50% 40% 30% 20% 10% 0% 7500

10000

12500

15000

17500

20000

22500

25000

Full income (euros per month)

Fig. 16.3 Women’s share in the sharing rule within working couples in relation to the full family income

potentially richer families. Of course, major variations are still found in the distribution of time and private consumption in Fig. 16.3. However, these variations depend on factors other than full family income. These include the relative wage, for example, the impact of which on the distribution of time and money was illustrated in Fig. 16.2.

Marriage Market We illustrated above that the sharing rule depends on factors such as relative wage. One possible explanation for this could be the collective model for the consumption behaviour of couples, proposed by Pierre-André Chiappori (1988). This approach assumes that men and women may have different preferences for how time and consumption are distributed within families, both for private and public goods. The ultimate distribution of time and consumption within the family, represented by the sharing rule, then results from a bargaining process between the partners. The better someone’s bargaining position, the greater his or her share in the sharing rule. The following questions could then be asked: Why does the sharing rule depend on the bargaining position of the partners in a couple? What exactly determines someone’s bargaining position? Gary Becker (1973, 1974) suggested that the “marriage market” could play a major role here. More specifically, he assumes that individuals look for the best possible partner for them on the marriage market. Although emotions naturally play a significant role, the specific advantages of

120

16

Who Wears the Trousers?

marriage (or cohabitation) must certainly be taken into account too. These include companionship, the opportunity to raise children together and economies of scale. The latter has already been mentioned several times in this section: public expenditure on items such as rent or utilities benefits the whole family while not depending, or depending only to a lesser extent, on the number of individuals within the family. Finally, marriage or cohabitation also makes it possible to spread risks. If one of the partners in a couple becomes unemployed, the family can still rely on the income of the partner who has not become unemployed, in addition to unemployment benefits that are often lower than the lost labour income. In a recent article by Laurens Cherchye, Thomas Demuynck, Bram De Rock and Frederic Vermeulen (2017), the collective model was merged with the concept of the marriage market. Their study demonstrated that someone’s bargaining position can be defined as how attractive that person is on the marriage market, in the sense that he or she has potentially more or fewer alternatives to his or her current partner. It is evident that someone with a high level of productivity (a high hourly wage) occupies a stronger position on the marriage market than someone with low productivity. In turn, this implies that a person with a higher relative wage has a more dominant position within the family and will be able to influence the distribution to his or her advantage. Figure 16.2 illustrates this correlation.

Part III

Who Deserves Special Attention?

In social policy, some groups of people deserve more attention than others, as they are generally considered to be the most vulnerable. In this part, we examine the circumstances of single-parent families, children and the elderly in turn.

Is Life Harder for Single-Parent Families?

17

Single-parent families can rely on particular attention in the literature as they are often over-represented in the poorer segments of the population. This is also the case in the MEQIN data: while the general risk of poverty is 14.5% (as described in Chap. 5), the same risk of poverty among single-parent families is no less than 24.5%. It is therefore important to gain a clear picture of the specific needs of these families. We first briefly describe some of the characteristics of the parents in single-parent families, before comparing the overall well-being of single- and two-parent families. We reach the hardly surprising conclusion that single-parent families achieve a lower level of well-being. We then focus on the dimensions of life in which these differences in well-being are most pronounced. Somewhat surprisingly, this is not the case for the quality of housing. The employment situation, however, turns out to be very important. Finally, we focus on the well-being of the children in both family types.

Who Are the Parents in Single-Parent Families? Single-parent families are traditionally defined as families in which a single parent is responsible for the upbringing and financial support of one or more children. We have opted to focus on families in which the parents are not retired or on an early leavers scheme and only on families with children under the age of 18. Table 17.1 compares various socio-demographic characteristics of the parents in single- and two-parent families. The parents in single-parent families are mainly women: almost 82.8% of single parents are women. Single-parent families have fewer children on average, 1.6 compared with 1.9 in two-parent families. On average, we also note that parents in single-parent families are more than two years older than parents in two-parent families. This difference stems from two contradictory effects. Firstly, in most cases, people initially become parents in a © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_17

123

124

17

Is Life Harder for Single-Parent Families?

Table 17.1 Socio-demographic characteristics Female Average age Number of children Highest degree: low Highest degree: medium Highest degree: high

Single-parent families

Two-parent families

82.8% 40.9 1.6 19.9% 40.2% 39.8%

50.0% 38.6 1.9 16.4% 34.8% 48.8%

two-parent family before the status changes to that of a single-parent family. As a result, the average age of the parents in single-parent families is higher. Secondly, people are often only single parents on a temporary basis as many later find a new partner. In itself, this will cause the average age of the parents in two-parent families to increase in relative terms. The average age difference in our data suggests that the first effect dominates the second. However, the parents in the two groups mainly differ in terms of their level of education. Table 17.1 shows that the average level of education is higher for parents in two-parent families.

Do Parents in Single-Parent Families Have Lower Levels of Well-Being? We could expect the level of material welfare to be lower among parents in single-parent families. After all, single parents are often divorced and divorce often leads to a drop in material welfare. Indeed, our data confirms that single-parent families achieve lower average levels of well-being. Table 17.2 shows the results for three different indicators of well-being. The first two indicators were also discussed in Chap. 5. The first indicator is disposable income. After correcting for differences in family size using the modified OECD equivalence scale (the correction already applied several times in the previous chapters), we can see that the income of single-parent families averages 1346 euros per month, just 70% of the average income of two-parent families. The second indicator is the extent to which families cannot afford certain necessary goods and services. In Table 17.2, we examine five relevant goods and services. For each of these goods, there is a greater chance that single-parent families will not be able to afford them. The biggest differences are found in the following areas: the ability to go on holiday for one week each year, to replace worn furniture and to buy a car. For example, 16.3% of single-parent families would like to buy a car but cannot do so due to a lack of funds; for two-parent families, this figure is only 2.1%.

Do Parents in Single-Parent Families Have Lower Levels of Well-Being?

125

Table 17.2 Income, deprivation and life satisfaction in single- and two-parent families Disposable income (euros per month) Percentage of families that cannot afford: Heating in winter Meat, fish or an equivalent every two days One-week holiday every year Replacement of worn or damaged furniture Car Average life satisfaction (from 0 to 10)

Single-parent families

Two-parent families

1346

1913

6.3% 6.0% 34.6% 37.4% 16.3% 6.9

3.0% 2.4% 19.3% 19.2% 2.1% 7.7

In Chap. 1, we saw that income is not an ideal measure of well-being. We may also wonder, for example, whether subjective life satisfaction differs between these two groups of parents. This third indicator of well-being is discussed in more detail in Chap. 21, but we are already using it here. Table 17.2 shows that average life satisfaction is lower in single-parent families (6.9 vs. 7.7 on a scale from 0 to 10).

Which Dimensions of Well-Being Make the Difference? We can now go a little deeper and also include non-material aspects of well-being in the analysis. As in the previous section of the book, we will be looking at health, housing quality, time use and employment situation. Our results in Table 17.3 show considerable differences between the health levels of parents in single- and two-parent families. Parents in single-parent families assess their health less favourably and are also less satisfied with their health than the other parents (6.9 compared to 7.7 on a scale from 0 to 10). Rather than an objective difference in health status, this difference in perception could also mean that parents in single-parent families are more concerned about their health. However, more objective indicators also point to poorer health on the part of single parents. For example, in the third row of the table, we can see that single parenthood is linked with a much higher risk of chronic illness. Of course, this does not mean that single parenthood in itself leads to a greater risk of developing a chronic Table 17.3 Health in single- and two-parent families Single-parent families Health (on a scale where 5 is very good and 1 is very bad) Average level of satisfaction with health (from 0 to 10) Chronic illness (%)

Two-parent families

2.2

2.6

6.9

7.7

40.7

19.6

126

17

Is Life Harder for Single-Parent Families?

Table 17.4 Housing in single- and two-parent families

Average level of satisfaction (from 0 to 10) Quality of housing (from 0 to 100) Size of housing (from 0 to 100) Quality of the living environment (from 0 to 100) Availability of services (from 0 to 100) Relationships with the neighbours (from 0 to 100) Percentage of owners

Single-parent families

Two-parent families

7.9 90.6 87.3 72.5

8.0 88.7 85.9 75.2

79.1 76.4

77.6 73.7

53.5%

80.4%

condition. On the contrary, it could be the chronic illness itself that increases the chances of becoming a single parent. However, this does mean that an effective approach to chronic diseases requires a special focus on single-parent families. The second dimension is the quality of housing. As Table 17.4 shows, there is little difference between single-parent and two-parent families both for subjective satisfaction with housing (7.9 compared to 8.0 on a scale from 0 to 10) and the objective characteristics of housing: sometimes single-parent families score worse here, e.g. for the quality of the living environment, and sometimes they score better, e.g. for social relations in the neighbourhood or the quality of the housing itself. Although other indicators point to significant differences, these only relate to indirect measures of the quality of housing. The percentage of owners is one of the indicators to which we devoted a great deal of attention in Chap. 11. We find that there are many more owners among two-parent families than single-parent families. Nonetheless, the differences in Table 17.4 are not entirely in line with the findings described in Chap. 11. It may be that the housing situation changes slowly, thus keeping the quality of housing at the level achieved when both parents were still present in the family for a while longer. Home ownership appears to offer an additional guarantee of not ending up in poverty. Incidentally, this is also confirmed by another result in the MEQIN survey: 24% of single-parent families reported having difficulty paying their housing and utility bills, compared to only 17% for two-parent families. From this perspective, it would therefore seem particularly important to help poor families (whether single-parent or two-parent families) to acquire sufficient capital to act as a buffer in the event of setbacks. The last dimension of well-being that we highlight here is the work situation, the time spent on labour and time use in general. Table 17.5 shows that parents in single-parent families are less likely to have paid work than other parents (67.0 vs. 84.6%). Of the single parents with paid work, the percentage who works part-time is also much higher than among working parents in two-parent families. However, this is no longer the case if we restrict ourselves to mothers: for mothers who work, the percentage who works part-time is lower among single-parent families than two-parent families (34.2 vs. 40.2%).

Which Dimensions of Well-Being Make the Difference?

127

Table 17.5 Work and time use in single- and two-parent families

Paid work (% of parents) Part-time work (% of parents with paid work) Part-time work by the mothers (% of mothers with paid work) Would like to work half a day more (mothers only) Household tasks (hours per week) Leisure (hours per week)

Single-parent families

Two-parent families

67.0% 31.0% 34.2%

84.6% 23.2% 40.2%

68.7% 14.2 18.3

37.6% 12.3 14.2

We might wonder whether this lower number of working hours could be regarded as a voluntary choice (e.g. because more attention is paid to the children) or whether parents of single-parent families would actually like to work more. The MEQIN survey therefore asked each interviewed adult who worked part-time the following question: If you were offered the chance to work half a day a week more in your current job, would you accept this if it meant you could adjust your consumption accordingly? If necessary, you can spread this extra half day across your working week. Your workload would also be adjusted to the new regime, giving you the equivalent of an extra half day’s work.

Almost 69% of the mothers in single-parent families who worked part-time answered this question in the affirmative, compared to just 38% of mothers in two-parent families. This suggests that mothers in single-parent families have fewer job opportunities than other parents. Finally, Table 17.5 provides information about the average number of hours spent by parents on household tasks and leisure. These results can be compared with those in Figs. 12.1 and 12.2. We can see that although parents in single-parent families spend more time on household tasks than the other parents (14.2 vs. 12.3 h), it is clearly less than twice the time, meaning that the total number of hours spent on household tasks is greater among two-parent families. Finally, we note that parents in single-parent families have about 4 h more leisure than the other parents.

Are There Differences When It Comes to the Children? Finally, we take an initial look at the situation of children in single-parent families. As income in these families is lower, it is not surprising that less is spent on children (see Table 17.6). However, we have already seen that there are also fewer children in single-parent families. The second row of Table 17.6 shows that expenditure per child does not differ significantly between the two groups and is even slightly higher in single-parent families.

128

17

Is Life Harder for Single-Parent Families?

Table 17.6 Time and money spent on each child in single- and two-parent families

Expenditure on children (euros per month) Expenditure on children per child (euros per month) Time spent on each child per adult (hours per week)

Single-parent families

Two-parent families

344 239

398 224

10.9

9.0

The biggest difference between children in single-parent and two-parent families relates to the time spent by the parents with each child. As we can see from Table 17.6, the average parent in a single-parent family spends a little more time with his or her children than a parent in a two-parent family, but much less than double. These results shed an initial light on the circumstances of children in vulnerable families. We will go into this in more detail in the next chapter.

Which Children Grow up in Poverty?

18

According to recent studies, about one minor child in five grows up in poverty in Belgium.1 In an average class of twenty children, this means that four children are growing up in a family living below the poverty line. These figures often conceal a complex reality and a wide gap between the children growing up in poverty and the other children in the class. In this chapter, we will try to shed light on this gap for various different dimensions of well-being. To this end, we divide the children in our data into four groups based on the disposable income of the family in which they are growing up. Here, we opt for a system that differs from the one applied in the previous chapters. The first group consists of the children from poor families. As discussed in Chap. 5, a family is classified as poor if the disposable family income (after correction for family size) is below the poverty line of 972 euros. The second group includes all the children from families with an income between the poverty line and twice that amount. We refer to this group as the low middle class. Children from families with an income between two and three times the poverty line form the group that we refer to as the high middle class. Finally, children from families with an income higher than three times the poverty line are regarded as rich. Figure 18.1 shows the distribution of minor children between these four groups. In our data, we can also see that about one child in five comes from a poor family and that most children (almost half) are growing up in the vulnerable low middle class. The lowest number of children comes from the rich income group. This chapter focuses in more detail on the circumstances, health and housing of children in various income groups. We supplement the picture by looking at how

1

For example, see the contribution of Frank Vandenbroucke and Julie Vinck in the Belgian Journal of Social Security in 2015.

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_18

129

130

18 Which Children Grow up in Poverty? 50%

Percentage of children

40%

30%

20%

10%

0%

Poor

Low middle class

High middle class

Rich

Family income group

Fig. 18.1 Distribution of children between the four income groups

much time and money are spent on the children in various income groups. Finally, we examine the satisfaction of the children in various income groups.

Poverty Is Hard In 2009 and 2014, SILC (the main European survey on income and living conditions) paid particular attention to “material deprivation” in children. Parents were asked whether their children had access to certain age-specific items and whether they could participate in certain activities. We also asked similar questions in the MEQIN study. At the end of the regular interview, the parents who took part in the survey were given an additional questionnaire for each minor child in their family. Just over 60% of the parents returned this additional questionnaire to the research team. Unlike the SILC survey, there is a separate questionnaire for each child. As in Part II, to a certain extent, this allows us to look behind the scenes for each family as we have information about the situation of each individual child. Among other things, the questionnaire includes 13 questions on material deprivation. The principle is the same as for the deprivation questions relating to the circumstances of the adults, as used in Chap. 5, but the questions have been adapted to the situation of children. The left-hand column of Table 18.1 shows these questions. Figure 18.2 shows the number of deprivations for children in different income groups, i.e. the average number of items that children in each of the groups have to do without. It is striking that children in the poor income group miss out on far more items than children in the other groups.

Poverty Is Hard

131

Table 18.1 Deprivation in the four income groups (in %) Poor (%) New (not second-hand) clothing Two pairs of shoes (including one pair of closed shoes) Vegetables and fruit at least once a day At least one meal a day containing meat, chicken, fish or a vegetarian substitute Age-specific books other than textbooks Toys for outdoor use such as a bicycle, roller skates, skateboard, etc. Age-specific indoor toys or games A suitable place to study or do homework (sufficiently spacious and quiet) Regular participation in leisure activities outside the family Celebrating life events such as birthdays and religious ceremonies Occasionally inviting friends round to play or for a meal One-week holiday every year Participation in paid school activities such as school trips and excursions

Low middle class (%)

High middle class (%)

Rich (%)

12 1

1 0

0 0

0 0

0 8

0 0

0 0

0 0

3 5

2 2

0 0

0 0

2 5

1 1

0 0

0 0

8

2

0

0

6

2

0

0

6

1

0

0

52 6

12 0

2 0

0 0

Looking at the individual questions in Table 18.1, we note that 52% of children in the poorest income group do not have annual holidays. This is much less often the case for children in one of the other income groups. In addition, almost 13% of children who live in poverty do not have access to new clothing for financial reasons and over 6% cannot celebrate a special life event (e.g. a birthday party). They often fall through the net even for school trips. Although children in the low middle class are much less deprived, we can still identify a clear difference between them and the higher-income groups.

Health and Housing Quality We will now examine the health and housing of the children in various income groups. In the left-hand panel of Fig. 18.3, we show the results for a general health score measured on a scale between 1 and 5. We note that children growing up in poor families are in worse health than children growing up in rich families. However, their health does not differ significantly from that of middle-class children. Further research based on the supplementary questionnaire for children shows

132

18 Which Children Grow up in Poverty?

Number of deprivations

1.5

1

0.5

0 Poor

Low middle class

High middle class

Rich

Family income group

4.6

85

4.4

80

Quality of housing

Children's health

Fig. 18.2 Average number of deprivations of children in the four income groups

4.2

4

3.8

75

70

65 Poor Low middle High middle Rich class class

Family income group

Poor

Low middle High middle Rich class class

Family income group

Fig. 18.3 Health and housing of children in the four income groups

that the health difference between poor and rich children is not expressed so much in physical terms, but that the children in the poorest income group mainly achieve lower scores for the behavioural or psychological aspects of health. For example, parents in the poor income group are more likely to report problem behaviour by their children: they find it hard to concentrate, tend to lie more often and argue more frequently.

Health and Housing Quality

133

Although healthcare is certainly not the only (and probably not even the most important) determinant of health, it is still reasonable to assume that the health outcomes of children in the poorest families can be linked to the results for the accessibility of care shown in more detail in Chap. 8. These results showed that many poor families find it difficult to fit healthcare expenses into the family budget and, even more strikingly, that they regularly have to postpone urgent healthcare for financial reasons. In the right-hand panel of Fig. 18.3, we look at the quality of the housing in which the children are growing up. The figure shows a remarkable difference in quality between the housing of the poor group and all the other income groups. Children who grow up in a family in the poorest income group are more likely to live in housing that suffers from a leaking roof, damp walls or rotting woodwork. In Chap. 7, we already pointed out that this adversely affects the health of adults. There is no reason to believe that this would be any different for the health of children. Although we established in Chap. 17 that housing quality does not pose many problems in the specific situation of single-parent families, this positive finding therefore does not apply to poor families in general. After all, the latter group also includes families in long-term, structural poverty. In addition, about one-sixth of the poor families in the survey indicated that the payment of the rent or mortgage had been postponed at least twice in the past 12 months for financial reasons.

How Much Do Parents Invest in Their Children? Families do not just invest money in their children, they also invest time. We will therefore briefly review the results we discussed in Chaps. 10 and 12. The left-hand panel of Fig. 18.4 shows the time use of the adults in the family. Achieving a good balance between work and family is a challenge for all the income groups, but especially for the low middle class. In a normal week, the adults in this group spend the least time with their children, an average of one hour less than the poor and almost two hours less than the richer income groups. The low middle class spends more money on their children than the families in the poor income group, however, as shown by the right-hand panel of Fig. 18.4. The more privileged groups spend both more time and more money on their children. However, there is no noticeable difference between the amount of time and money spent by the high middle class and the rich income group on their children. Figure 18.4 suggests a considerable dilemma for the lower-income groups. Although more money is spent in the low middle class than in the poorest income group, this seems to be at the expense of the time spent on their children. It appears that these parents face a difficult balancing act.

18 Which Children Grow up in Poverty? 16

14

12

10

Poor

Low middle High middle Rich class class

Expenditure per child (euros per month)

Time spent on children (hours per week)

134

300 250 200 150 100 50

Poor Low middle High middle Rich class class

Family income group

Family income group

Fig. 18.4 Investment of time (left) and money (right) by adults in the family on their children in the four income groups

As Long as They’re Happy … Many children in the poor and low middle classes are therefore deprived in many areas of life. But does this gap actually make the children unhappier or less satisfied? To answer this question, we examine how the parents answered the question in the additional questionnaire as to how satisfied they think their children are with their own lives. The responses were reported on a scale between 1 and 5, where 1 means “not at all satisfied” and 5 means “very satisfied”. Figure 18.5 shows the

Children's satisfaction levels

4.4

4.2

4

3.8 Poor

Low middle class

High middle class

Rich

Family income group

Fig. 18.5 Children’s satisfaction levels according to parents in the four income groups

As Long as They’re Happy …

135

answers of the parents in the four income groups. The figure shows a clear trend: children growing up in a poorer income group are less satisfied than children in richer income groups. Once again, there seems to be no noticeable difference between children from the high middle class and rich income groups. Some caution is advisable here, however. The question about the children’s satisfaction was not answered by the children themselves, but by their parents. Parents from lower-income groups may project lower life satisfaction onto their children because of the family’s income situation, or they may generally have a more pessimistic outlook. However, this does not make the result in Fig. 18.5 any less relevant.

A Nice Retirement?

19

As a result of the ageing population, the number of elderly people in our society is rising sharply. Most of them are no longer active on the labour market and therefore need to live on other sources of income. We are mainly referring to pensions here, as well as investment income. It is generally accepted that Belgian pensions for private sector employees are on the low side. Moreover, it is often assumed that pensions can only remain affordable if people work for a longer period on average. At the same time, the health status of the elderly tends to deteriorate, leading to higher care costs. Concerns about pensions and rising care costs are therefore causing concern about the future quality of life of the growing group of elderly people in our society. Although our sample does not allow us to say anything about the evolution over time, we can provide a snapshot of the circumstances of the elderly in 2016. We do so in this chapter. We first look at the number of elderly people in our society and then chart their circumstances.

How Many Older People Are There? Figure 19.1 shows the population pyramid for the adult Belgian population based on the MEQIN dataset. The first group, the “young people”, are all people under the age of 50. Just under half (47%) of the adults are young people. They form the broad base of the pyramid. There are about as many women as men in the group of young people. We refer to all other people as the “elderly”.1 They form the subject

1

The Survey of Health, Ageing and Retirement in Europe (SHARE) is a specific survey that only surveys people over the age of 50, i.e. everyone except the first group in our classification.

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_19

137

138

19

A Nice Retirement?

>80 70-79 60-69 50-59 18-49 -30%

-20%

-10%

0% women

10%

20%

30%

men

Fig. 19.1 Distribution of adults across the five age groups

of this chapter. The 50-somethings form the second group. For our analyses in the rest of this chapter, we will sub-divide the group of sixty-somethings into the group that had not yet reached the retirement age of 65 (applicable in 2016), and those over the age of 65. People in their seventies and those aged over 80 constitute the top two groups of the pyramid. We can see that there are slightly more women than men in the older age groups. The proportion of women in the dataset increases with age because females have a higher life expectancy.

The Situation of the Elderly In Chaps. 6 and 8, we already discussed our findings on the relationship between age, and health and health costs. We will summarise them again briefly here. Up to the age of about 50, health deteriorates across all health dimensions except “emotional well-being”. Between fifty and sixty-five years of age, health status does not change significantly; it then starts to deteriorate again. The age-related effect for the dimension of “emotional well-being” is much less clear. In fact, between the ages of 60 and 70, people seem to feel quite well. The deteriorating overall health situation of the elderly does indeed lead to higher healthcare costs. Moreover, older people find health important and do not tend to postpone care even if it starts to place a heavy burden on their budget. We note that elderly people go to the dentist and the physiotherapist a little less often, but when they do visit the physiotherapist this is often for intensive treatment. The need for home care also increases sharply with age. We can now combine this health-related information with information on other important life dimensions as presented in previous chapters. Table 19.1 provides a general overview of the life situation of different age groups. The first block of this

The Situation of the Elderly

139

Table 19.1 Life situation of different age groups Everyone 18–49 50–59 Demographics % in a relationship 71.4 % in a residential care 1.3 centre (in our sample) Work situation % not working 43.3 % unemployed 5.8 % full-time work 37.9 % part-time work 13.0 Average number of hours 5.5 of informal care Incomes % poor 12.5 % low income 34.6 % average income 25.6 % rich 27.3 Average pension (euros 1456 per month) % that can pay an 74.5 unexpected expense of 1000 euros from their own resources Housing % owner without 36.6 mortgage % owner with mortgage 33.6 % tenant on private 18.3 housing market % tenant in social 6.3 housing Housing quality (0–100) 91.7 Housing size (0–100) 91.1 Quality of environment 76.7 (0–100) Accessibility of facilities 77.0 (0–100) Social relationships in 75.2 the neighbourhood (0– 100) Satisfaction with housing 8.1 situation (0–10) Well-being % not deprived 67.8 Life satisfaction (0–10) 7.38

60–64

65–69

70–79

80+

74.4 0.2

73.2 0.0

77.2 0.3

74.1 0.0

66.7 2.0

38.0 15.6

17.0 9.9 57.2 15.8 4.3

25.9 5.8 46.8 21.5 5.9

75.1 0.1 15.3 9.5 10.0

94.5 0.0 2.4 3.1 9.5

97.1 0.0 1.1 1.7 5.0

100.0 0.0 0.0 0.0 2.7

15.5 26.9 28.4 29.3 769

10.6 32.2 24.1 33.2 1222

8.7 39.5 25.9 25.9 1455

8.4 41.2 27.4 22.9 1411

9.6 47.8 19.3 23.3 1583

12.1 58.9 18.2 10.7 1370

69.6

77.0

76.5

80.7

83.4

78.5

12.7

41.3

61.5

68.4

76.1

61.0

52.1 23.9

33.9 13.5

12.7 11.7

7.6 16.2

3.7 10.9

3.6 15.8

5.5

7.6

8.4

4.0

5.7

9.8

89.9 87.1 75.1

92.0 93.8 78.0

92.9 94.7 79.2

94.3 95.8 78.8

95.0 95.6 77.4

93.5 94.7 77.5

79.4

76.3

74.7

75.1

74.7

71.1

73.4

75.5

77.5

79.0

78.0

75.8

7.9

8.1

8.3

8.3

8.3

8.2

62.3 7.38

70.9 7.25

69.3 7.41

72.6 7.53

79.7 7.51

70.2 7.36

140

19

A Nice Retirement?

table provides demographic information. Older people are more likely to live alone. In the second row of the table, we provide the answers to the question “Are you currently in a relationship?” The number of positive responses decreases significantly from the age of 70 upwards. The third row shows the percentage of various age groups that live in a “residential care centre”. The percentage of people aged over 80 in this situation is almost 16%. Although it is reassuring that our figures come very close to the official figures,2 this does not alter the fact that our sample is small. Our sample only includes 43 people living in a residential care centre; 32 of them are over 80 years old. The second block of results in the table shows the employment situation of the elderly. These results can be compared with those in Chap. 9. From the age of 60 upwards, the percentage of non-working people starts to rise sharply: in our sample, only 15.3% of people between the ages of 60 and 64 work full-time and 9.5% work part-time. Hardly, anyone in this age group is looking for work. The percentage of full-time and part-time workers in their fifties is also only 68%, and the number of unemployed people (non-working people who are seeking work) is 5.8%, while almost 26% are not working and not actively seeking work either. Another striking result is the average number of hours spent on informal care (both inside and outside the family). It appears that people aged between 60 and 70 provide an average of 10 h of informal care per week, probably to partners or parents who require assistance. This average conceals major differences between individuals. It goes without saying that people who provide a great deal of informal care can hardly be expected to actively look for work on top of this. The third block of the table examines incomes by age group. From the age of 60 and certainly 65, the pension becomes the main source of income. The average net pension remains fairly stable at around 1400–1600 euros per month from the age of 60 upwards.3 Although this pension can of course be combined with other sources of income, its level still largely determines the income position of the elderly in Belgium. We can see that the elderly are not rich, but not poor either. Overall, the risk of poverty (i.e. a family income that is less than 60% of the median income) is higher among young people than the elderly.4 At the same time, however, young people also account for a higher proportion of the richest group (defined here as the top 25% of the income distribution). As a result, there is less income inequality among the elderly than among young people. Generally speaking, Belgian

According to official figures from the Flemish health insurance and the Flemish government agencies that develop, produce and publish official statistics, 16.3% of people over the age of 80 lived in a residential care centre in 2015. 3 The averages shown in Table 19.1 refer only to people who report a pension income. 4 Attentive readers will note that the percentage of people below the poverty line (12.5%) differs from the figure in Chap. 5 (where 14.5% were classified as poor). This difference is due to different sub-group being considered. We are looking at the adult population here, whereas, in Chap. 5, we examined the population as a whole. Poverty rates among families with minor children therefore appear to be disproportionately high. 2

The Situation of the Elderly

141

pensioners can enjoy a relatively peaceful old age: we find that more than 80% of people over 65 live in a family that can pay an unexpected expense of 1000 euros from their own resources. This percentage is significantly higher than in the younger age groups. Housing partly explains this phenomenon, as can be seen in the fourth block of the table. More than two-thirds of people over 65 live in a family that owns the family home outright (i.e. with no mortgage). In the oldest groups, this figure even reaches 76%. Young people are more likely to live in a family that rents the family home on the private housing market and, more importantly, in a family that owns the family home but has not yet paid off the mortgage. The latter applies to over half of people below the age of 50. The elimination of the mortgage burden certainly makes an important contribution to the quality of life of the elderly. Moreover, the elderly also feel that the quality and size of their home is above average. As already indicated in Chap. 11, although somewhat less impressed with the accessibility of facilities such as shops or public transport in the vicinity of their homes (they may well have higher demands in this area due to their limited mobility), they are happier with the social relationships in their neighbourhoods: they can count on the help of their neighbours more than young people can, trust their neighbours and generally feel that the people in their neighbourhood get along well. Overall, they are more satisfied with their housing situation than people below the age of 60. In the last block of the table, we can see that older people are generally satisfied with their lives. Pensioners below the age of 80 score better here than all the other age groups. This satisfaction has an objective basis, as evidenced by the deprivation score that we already used in Chap. 5 which shows how many essential items older people cannot afford (such as going on holiday for a week every year, having a colour TV and a washing machine). Almost 38% of people below the age of 50 are deprived of at least one of these items, compared to just over 20% of people in their seventies. On average, therefore, the elderly in Belgium are doing quite well. However, this optimistic conclusion must be accompanied by the following three qualifications. Firstly, a more accurate examination of the table reveals that people over 80 are worse off in every dimension than those in their seventies. We have already mentioned that 16% of them live in residential care centres (and therefore may have sold their own home first), but the number of observations in our sample is too small to make statements about the quality of life in residential care centres. This is generally true of statements about the 80+ age group: our results might suggest that the elderly in society are worse off than the other age groups, but because the 80+ age group in our sample is not particularly large, this cannot be deduced with great certainty from our study. Incidentally, the (possibly worse) life situation of people in the 80+ age group could be a specific illustration of a second qualification. In this chapter, we have only looked at the average results for different

142

19

A Nice Retirement?

age groups. This average does not teach us much about the possibility of abject poverty among the elderly in some age groups. Finally, we do not know whether our results are due to the age of the people in our sample or the time they were born. After all, everyone over the age of 80 in our study was born just before the Second World War. We cannot tell whether their current life situation is due to their age or the circumstances during the early years of their lives. In Chap. 7, we already discussed the great importance of the early years of life on health in later life.

Part IV

Towards a Measure of Individual Well-being

In Part II of this book, we have already examined well-being from various different perspectives. In Chaps. 4, 5 and 10, we considered the material welfare of Belgians based on income and consumption indicators. In Chap. 11, we focused, in particular, on the quality of their housing. In Chaps. 6–8, we considered their state of health, and in Chaps. 9 and 12 we looked at their employment situation in particular and time use in general. Although each of these individual perspectives provides some interesting insights, we need a multidimensional viewpoint to truly convey the multifaceted nature of well-being. By taking this viewpoint, we reveal all these perspectives at the same time, as well as the links between them. In this part, we will look for a measure that summarises individual well-being. Firstly, we examine the phenomenon of cumulative deprivation. We then ask ourselves whether Belgians are satisfied with their lives and whether life satisfaction can be regarded as a general measure of well-being. Since we don’t think it can, we then formulate an alternative: the “equivalent income”.

Who Suffers from Cumulative Deprivation?

20

In Part II, we already gave several examples of pairwise correlations between poor scores for various dimensions of well-being (health and income, income and job quality, housing and health, etc.). In this chapter, we examine the correlation between three important dimensions of well-being (material welfare, health and housing) in a more systematic way. We measure the material welfare of Belgians based on their disposable income, as in Chap. 4. This income reflects their ability to acquire material living standards. As a health index, we take the average of the different health dimensions introduced in Chap. 6; when it comes to housing quality, the index is the average of the various characteristics in Chap. 11. We start our analysis by dividing the respondents into three groups of equal size for each of these dimensions. Each respondent is awarded a score for each dimension. This score is “low”, “medium” or “high”. The combination of these three scores gives us a first rough overview of individual well-being. For example, some people achieve high scores for income but low scores for health and medium scores for housing. For these people, the high score in one dimension may compensate for the lower scores in the other dimensions. People who achieve low scores for all three dimensions at the same time suffer from what we call cumulative deprivation. These people are in a particularly precarious situation as their poor score in one dimension is further aggravated by poor scores in the other dimensions. Not only do they belong to the poorest group, they also find themselves in the group with the worst health and housing.

How Much Cumulative Deprivation Is There in Belgium? Even in a hypothetical world in which there is no systematic correlation between the dimensions, and the scores in the three dimensions are awarded randomly, there would still be a number of people suffering from cumulative deprivation. It’s not hard to figure out how many people that would be. A total of 27 different © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_20

145

146

20

Who Suffers from Cumulative Deprivation?

combinations of scores are possible. All of these combinations are listed in Table 20.1. In a random world, each score would be equally common, i.e. it would occur in just under 3.7% of cases. The last column of Table 20.1 shows how often each combination is actually observed in real life. We can see that most combinations do indeed occur in about 4% of cases, sometimes slightly more and sometimes slightly less. However, the combination in the first row of the table immediately stands out. This is the combination in which people belong to the group that achieves low scores for income,

Table 20.1 Correlation between income, health and housing scores

Income

Health

Housing

Percentage (%)

Low Low Low Low Low Low Low Low Low Medium Medium Medium Medium Medium Medium Medium Medium Medium High High High High High High High High High

Low Low Low Medium Medium Medium High High High Low Low Low Medium Medium Medium High High High Low Low Low Medium Medium Medium High High High

Low Medium High Low Medium High Low Medium High Low Medium High Low Medium High Low Medium High Low Medium High Low Medium High Low Medium High

7.7 4.5 3.5 4.4 3.2 2.2 3.1 2.4 2.4 3.6 2.8 3.6 3.8 3.8 4.1 2.9 4.1 4.9 2.1 2.9 2.7 3.2 4.5 4.6 2.5 5.3 5.2

How Much Cumulative Deprivation Is There in Belgium?

147

health and housing. In other words, this is the group in which cumulative deprivation occurs. No less than 7.7% of all Belgians suffer from cumulative deprivation, twice as much as we would expect in an entirely random world. This remarkable finding points towards a strong correlation and interaction between these three dimensions of well-being, especially for those in the low-scoring groups. It is hard to establish exactly how this interaction works. A precarious income position may make people sick and land them in low-quality housing, for example. Poor housing could also make people sick, putting them in a precarious income position. There may be a complex interplay between the various factors that affect these dimensions, bringing about a correlation in the scores achieved for these dimensions.

Who Suffers from Cumulative Deprivation? Who are the 7.7% of the Belgians suffering from cumulative deprivation? To find out, we will take a closer look at three of the 27 groups. These are the groups that achieve equally high (or low) scores for the three dimensions of well-being. In Table 20.1, these groups are indicated in grey. The “low” group (containing 7.7% of the people) achieves low scores for all dimensions and constitutes the group suffering from cumulative deprivation. The “middle” group (3.8% of the respondents) occupies a middle-of-the-road position for each dimension. The “high” group contains the 5.2% of people who belong to the group that scores highest for all three dimensions. In Fig. 20.1, we compare the characteristics of the people suffering from cumulative deprivation with the characteristics of the other two groups. The top left-hand panel shows the percentage of women in the three groups. We can immediately see that the low group, i.e. the group suffering from cumulative deprivation, contains many more women than the other two groups. The second panel, at the top right, shows the percentage of pensioners (or people who have taken early retirement) in the three groups. We note that slightly fewer, around one in four, of the people with cumulative deprivation are pensioners. This share is remarkably higher than the share in the group that achieves high scores for every dimension. It is also striking that the middle group contains a large group of pensioners. This is entirely in line with all the other findings that we described in Chaps. 5 and 19: the elderly are not necessarily the worst off in our society. The third panel, middle left, shows that around 25% of people with cumulative deprivation are in a relationship, regardless of whether they live with the person concerned. In the other two groups, about 70% of people are in a relationship. A relationship may possibly protect people from the phenomenon of cumulative deprivation through spreading the risk of income loss, or providing informal healthcare between partners. However, the correlation may also work in a different way: the phenomenon of cumulative deprivation may also make potential partners less attractive, leaving them alone for longer, a mechanism that we pointed out at

148

20

Who Suffers from Cumulative Deprivation? 45%

70%

40%

60%

35% 50%

Pensioners

Women

30% 40% 30%

25% 20% 15%

20%

10% 10%

5%

0%

0% Low

Medium

High

Low

Medium

High

Low

Medium

High

Low

Medium

High

25%

80% 70%

20% 50%

Migrant

Relaonship

60%

40% 30%

15%

10%

20% 5% 10% 0%

0% Low

Medium

High 16%

60%

14%

50%

Unemployed

Low-skilled

12% 40% 30% 20%

10% 8% 6% 4%

10%

2%

0%

0% Low

Medium

High

Fig. 20.1 Characteristics of the three groups with equal scores in the three dimensions

the end of Chap. 16 and referred to as the “marriage market”. The next panel, middle right, looks at the country of birth of the people in the three groups. The difference is striking: among the people in the low group, just over 20% of people have first-generation migrant status, compared to only 5% in the high group.

Who Suffers from Cumulative Deprivation?

149

9

Life sasfacon

8

7

6

5 Low

Medium

High

Fig. 20.2 Life satisfaction of the three groups with equal scores in the three dimensions

In the panels at the bottom, we look at the educational attainment and labour market status of the people suffering from cumulative deprivation. The left-hand panel shows that over half the people in the group with cumulative deprivation have low levels of education. The right-hand panel shows that around 15% of all people with cumulative deprivation are unemployed, i.e. not working but actively seeking work. This percentage is much lower and sometimes even negligible in the other groups. Finally, in Fig. 20.2, we consider the impact of cumulative deprivation on life satisfaction. People who achieve the highest scores for the three dimensions are far more satisfied with their lives: on average they score no less than two points higher (on a scale from 0 to 10) than the Belgians in the lowest group. In the next two chapters, we will discuss in more detail the explanation for this and the relevance of life satisfaction. In any case, we can already conclude here that the phenomenon of cumulative deprivation is widespread in Belgium. Remarkably often, the people in this precarious group are single women, low-skilled and/or unemployed. Unfortunately, this phenomenon remains below the radar in standard aspect-by-aspect analyses of policy indicators.

How Happy Are We?

21

To find out how happy Belgians are, we must first dive deeper into the underlying question as to how concepts such as happiness and life satisfaction can actually be measured. We then examine what gives the average Belgian a high level of life satisfaction, and how satisfied he or she is with various facets of life.

Happiness and Life Satisfaction Influenced by findings in the field of psychology, many social scientists distinguish between the concepts of happiness and life satisfaction.1 Happiness is a fluctuating emotion that is influenced by the constant flow of major and minor events in life. We feel happy at the birth of a child, we are less happy when our favourite football team loses and then cheer up again when we feel the first rays of sun in spring. Life satisfaction, on the other hand, is a rather more reflective opinion that is based on a more detached analysis of every aspect of life. We are probably satisfied with certain aspects of our life (our relationship, for example) and somewhat less satisfied with other aspects (our job and the associated salary, for example). When respondents are asked about their life satisfaction, we assume that they weigh all relevant aspects of a good life against each other. Are their lives working out as they expected? Are they achieving the goals they consider important? These two concepts, therefore, indicate very different aspects of our internal disposition. People who are being exploited by their boss or get caught up in a traffic jam may feel very bad at the time, but it doesn’t mean that they also think their whole life is worthless. For example, take our attitude towards children.

1

For example, see Diener (1994).

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_21

151

152

21

How Happy Are We?

There are moments in the day when children (both younger and older) can be extremely irritating and make their parents feel as though they might explode. Parents certainly don’t feel good at the time. However, these same parents may at the same time feel that their lives would be much less worthwhile without their children (even if there are more stressful moments than happy moments). Most studies measure happiness or life satisfaction in a very simple way: the respondents are asked to indicate on a scale between 0 and 10 how happy or satisfied they are with their life. These kinds of questions are asked in surveys all over the world.2 Even though, at first glance, such rudimentary questions might seem a little too simple to measure a complex concept such as happiness, the answers in all these surveys show a remarkably stable picture. There is also a very strong correlation between the answers to the two questions in the large samples, despite the fact that happiness and life satisfaction must be carefully distinguished from a conceptual point of view. Some researchers (such as Daniel Kahneman) attribute this to the flawed questioning in these large samples and argue for a different method of measuring the first aspect (happiness) that they also consider to be the most psychologically relevant. Popular methods include the “day reconstruction” or “experience sampling” method, in which respondents are asked to keep a diary or to indicate at regular intervals how they are feeling at that time. One important study, for example, finds that people mainly have positive feelings when they are having sex or when chatting with friends and relatives in a relaxed way. They feel the worst at work or (even more strongly) in traffic jams or when they are alone or with their boss (Kahneman et al. 2004). We have not collected any data using this specific method of measuring happiness and only discuss the answers to the simple standard questions in this book. Figure 21.1 shows the distribution of the answers to these standard questions. We show the answers for happiness in the left-hand panel and the answers for life satisfaction on the right. Both figures look very similar. The vast majority of respondents use the top half of the scale and most Belgians answer 8 to both questions. On average, Belgians give a score of 7.56 for happiness and 7.38 for life satisfaction. These figures are in line with other studies for Belgium. Around 60% of all respondents give exactly the same score for their happiness as for their life satisfaction. This suggests that respondents are either as happy as they are satisfied with their lives, or that they do not distinguish between the two concepts as strongly as social scientists do. Here, too, our results are in line with previous research. In the rest of this chapter, we will take a closer look at the answers to the reflective question of life satisfaction. The answers to the specific question about happiness are very similar and are therefore not shown.

2

One example is the annual World Happiness Report.

What Makes Belgians Satisfied with Their Lives?

153 40%

Percentage of Belgians

Percentage of Belgians

40%

30%

20%

10%

0%

0

1

2

3

4

5

6

7

8

9 10

30%

20%

10%

0%

Happiness

0

1

2

3

4

5

6

7

8

9 10

Life satisfaction

Figure. 21.1 Happiness and life satisfaction in Belgium

What Makes Belgians Satisfied with Their Lives? Money doesn’t buy happiness, as the saying goes. But are wealthy Belgians really unhappier than their less affluent compatriots? To answer this question, the left-hand panel of Fig. 21.2 shows the average life satisfaction at various points in the income distribution. To construct this figure, we measured disposable income as described in Chap. 4. Poor Belgians (the dwarves in the parade) are on the left and wealthy Belgians (the giants) are on the right in the figure. The 972-euro poverty line, described in Chap. 5, is represented by the vertical line. The black curve shows the average life satisfaction for 12 specific income groups in our sample of Belgian adults. We note that the curve is rising and that richer people are happier with their lives than poorer people. However, the effect of extra income on happiness decreases as people become richer. It is striking that people living below or around the poverty line are less satisfied than rich people and middle-class people. Even if having more money makes Belgians only a little more satisfied on average, living in poverty clearly makes Belgians less satisfied. In Chap. 1, we already pointed out the apparent paradox that income growth over time increases satisfaction only slightly (or not at all), whereas at a certain point in time and within a certain society, there is indeed a strong correlation between income and life satisfaction. We may have to seek the explanation in the phenomenon of comparison with the income of social reference groups, which can therefore partly explain our results too. On the first day of the new year, we usually wish each other good health and happiness. Are healthy Belgians more satisfied with their lives? In order to answer this question, we use the general health scores from Chap. 6 and review the average life satisfaction for people with different health scores once again. In the right-hand panel of Fig. 21.2, we can see that unhealthy Belgians (on the left in the figure) are

154

21 9

How Happy Are We?

9 8

Life sasfacon

Life sasfacon

8

7

6

7 6 5 4 3 2

5 0

1000

2000

3000

Disposable income

4000

5000

0

10 20 30 40 50 60 70 80 90 100

Health score

Fig. 21.2 Life satisfaction and disposable income (left) and health (right)

less satisfied than healthy people (on the right). There is a major difference, more than 5 points on the scale between 0 and 10. Good health appears to be an important factor for life satisfaction (and happiness). Several other factors that contribute to higher life satisfaction are shown in Fig. 21.3. Each panel of the figure shows the average life satisfaction for two groups. In the first panel, we find that the group of Belgians with a relationship are more satisfied than the group without a relationship. Unemployed people are significantly less satisfied with their lives than people who work. This is not (fully) explained by the lower income of the unemployed. Statistical analysis confirms what has already been revealed in a great deal of sociological research: unemployment would reduce life satisfaction even if the unemployed were able to retain their former income. This is probably due to the fact that being active on the labour market leads to social inclusion and recognition. Finally, there appears to be no difference between men and women when it comes to average life satisfaction. This is a striking result by comparison with the previous chapter, where we found a major difference in cumulative deprivation between men and women. All these effects are in line with the findings of other studies for Belgium and other Western countries. If we compare happiness or life satisfaction for respondents of different ages, we notice a roughly U-shaped correlation. On average, young Belgians become less satisfied each year until a low point is reached. After this, they become a little more satisfied each year. Age is naturally linked with many other factors: examples include income, health and employment status. Keeping all these other factors constant, we find that on average people in their early forties are the least satisfied age group in Belgium. All these insights are so intuitive that they seem almost self-evident. A healthy dose of common sense and some everyday insights into human psychology could probably have brought us as far as this analysis that was based on a survey.

What Makes Belgians Satisfied with Their Lives?

155 8

Life satisfaction

Life satisfaction

8

7

6

5

7

6

5 No

Yes

No

Relationship?

Yes

Unemployed?

Life satisfaction

8

7

6

5 No

Yes

Female?

Fig. 21.3 Life satisfaction for various demographic groups

However, we learn more from these results than just what makes Belgians happy. We can also calculate how much happier they become on average following certain changes in their lives. For example, we note that the impact of being unemployed is greater than the impact of not being in a relationship, or that the impact of an increase in income from the poverty line (around 972 euros) to a disposable income of 4000 euros is about the same as the impact of an improvement in general health from 60 to 100. Bringing these insights together shows us how Belgians weigh up various aspects of their lives against each other. This gives us an interesting insight into the answer to the old question as to what constitutes a good life. In Chaps. 23 and 24, we will introduce another—more direct—approach for this.

156

21

How Happy Are We?

(Dis)satisfaction with a Partial Aspect of Life Not only did we survey Belgians about their happiness and overall life satisfaction, we also asked them how satisfied they were with various partial aspects of their lives. We use the same simple and somewhat rudimentary method here, based on a scale between 0 and 10. Figure 21.4 shows the average answers for satisfaction with their consumption level, health, housing and job. For reference, we show the average overall life satisfaction (7.38) on the far left of the figure. We can see that, on average, Belgians are least satisfied with their level of consumption.3 They are most satisfied with their housing. This last finding is in line with what we found in Chap. 11: in general, Belgians are quite satisfied with the objective characteristics of their home. The difference in satisfaction between the two partial aspects is fairly large and exceeds one point on the scale between 0 and 10. Average satisfaction with health is slightly lower than average overall life satisfaction, while satisfaction with the current job is slightly higher. However, we would like to point out that the question of satisfaction with the current job was only asked to respondents who worked at the time of the survey and the question of satisfaction with housing was only asked to the reference person within each family. In Table 21.1, we examine the percentage of people who are dissatisfied with one of the facets of their lives. Here, we look at the respondents who gave a score of less than or equal to six on a scale of 0 to 10. As we might expect on the basis of Fig. 21.4, the percentage of Belgians who are dissatisfied with their housing is lowest (9.0%), while the figure is highest for personal consumption (31.2%). The subsequent rows of Table 21.1 can be used to compare the results for the whole of Belgium in the first row with specific groups within Belgium. Women seem somewhat less satisfied with each facet, although they are about as satisfied with their lives in general. This finding suggests that women may be less dissatisfied with one or more other facets (not covered here) when thinking about their lives in general. Belgians who are in a relationship are less dissatisfied with their lives in general, except when it comes to their jobs. People with a migrant background, however, are much more dissatisfied with their lives, their housing and— above all—their personal consumption. Individuals with low levels of education are also more dissatisfied with their lives in general and the facet of health in particular. This finding is in line with what we saw in Chap. 6, where we noted that those with low levels of education are also less healthy. Unemployed people are significantly more dissatisfied with their lives in general. The high number of unemployed people who are dissatisfied with the facet of personal consumption is also striking. No less than 64% of unemployed respondents gave a score of less than or equal to six on a scale of 0–10. Finally, we note that pensioners are less dissatisfied in every 3

The consumption level refers to the personal consumption, i.e. private consumption and an equal allocation of public consumption between the family members (see Chap. 12 for the definition of public and private consumption within a family). However, as we surveyed respondents about their level of satisfaction with the housing separately, we did not include the housing expenses or imputed rent in personal consumption. The concept of personal consumption is discussed further in Chap. 24, where we propose an alternative measure of well-being.

(Dis)satisfaction with a Partial Aspect of Life

157

8.5

Satisfaction

8

7.5

7

6.5

6 Life

Consumption

Health

Housing

Job

Fig. 21.4 Satisfaction with partial aspects of life

Table 21.1 Percentage of dissatisfied Belgians for each facet Percentage of Belgians who are dissatisfied with their … Life (%) Consumption (%) Health (%)

Housing (%)

Jobs

Everyone Female In a relationship Migrant Low-skilled Unemployed Pensioners

9.0 10.4 7.8 19.0 11.1 21.9 6.1

15.6% 16.3% 16.5% 20.6% 21.7% / /

21.0 20.9 17.4 31.8 27.9 41.7 20.3

31.2 32.8 28.4 46.7 39.4 64.4 26.7

23.6 27.2 21.6 25.5 30.9 29.4 28.8

area with the exception of health. This phenomenon is also consistent with our findings from Chap. 6, where we observed a clear link between health and age. On average, these results are in line with the descriptions of life situation in previous chapters in more objective terms.

As Long as We’re Happy …?

22

On average, Belgians with a decent income, good health, a relationship and work are happier than others. So couldn’t we simply use reported happiness as an indicator for steering or evaluating policy? If all citizens want to be happy, it would be very easy to make political decisions using an overall happiness score as a guiding policy indicator. A government that wishes to eliminate dissatisfaction could focus on the individuals in Table 21.1. It would only need to quantify how to create as much happiness as possible for its citizens using its limited resources. As a result, a large-scale survey and powerful computer could replace long and difficult political discussions on policy priorities. This vision of policy and its objectives echoes the thinking of eighteenth-century utilitarian philosophers and economists such as Jeremy Bentham. They felt that the government’s task was to create the maximum benefit or happiness for its citizens. Simply asking about happiness or life satisfaction in large-scale surveys and modern statistical analyses of these surveys therefore seem to be reviving this old “utilitarian” tradition.1 “Feelings” of happiness are certainly important in life. It is hard to imagine that someone who feels terrible all the time would eventually achieve a high level of well-being. However, happiness is not the only thing that drives us. People can make an effort for things they care about, even at the expense of temporary feelings of happiness. For example, informal carers who take care of a family member in need often experience this as a valuable—yet emotionally challenging—activity. Some migrants leave everything they own in their country of origin, often at a high emotional cost, hoping to guarantee a better life for their children. People decide to have children because they think it will enrich their lives, while probably realising

For example, see the article by Kahneman, Wakker and Sarin (1997) with the title: “Back to Bentham?”.

1

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_22

159

160

22 As Long as We’re Happy …?

that bearing responsibility for children also generates a lot of stress and tension. Therefore, happiness is not everything. In addition, in this book, we are looking for a measure of individual well-being that can be used to determine priorities in social policy. As a result, we need to be able to compare the well-being of different people. In Chap. 1, we already stated that from this perspective various questions could be asked about the use of happiness or life satisfaction as a measure of well-being. These questions relate to the importance of expectations and aspirations. If we can see that people adapt, at least in part, to their income situation, does that make the income situation less important as a policy objective? We think not. Expectations and aspirations also play a significant role in happiness for dimensions other than income. Here too, people adapt. Perhaps the most striking examples of this can be found in the health area. For example, many patients who suffer from locked-in syndrome appear to adapt so well to this situation that they achieve a reasonable level of life satisfaction once again (Bruno et al. 2011). Locked-in syndrome is an unenviable condition in which affected patients suffer from severe paralysis of the muscles of the face and limbs, only allowing them to make vertical eye movements. This makes communication with the outside world very difficult. At the same time, however, these patients are not in a coma and remain fully aware of what is happening around them. These patients, therefore, appear able to adapt to very negative circumstances. However, the fact that some patients adapt does not mean that they would rather remain ill than recover. Healthy people would probably prefer not to suffer from this syndrome either, even if they know that they would, on the whole, adapt. It seems obvious that it would be better for society to invest in research to cure or prevent this syndrome, rather than giving it a very low priority and counting on people’s ability to adapt. Happiness and life satisfaction are therefore determined not only by people’s more or less objective circumstances, but also by the extent to which they can adapt to their circumstances or by their expectations for their future lives. Personality characteristics and social background play a major role here. Is it desirable to prioritise pessimistic people who give a lower satisfaction score for every life situation when allocating scarce public funds? Probably not. Amartya Sen gives the example of a housewife with a miserable life who has adapted to her hopeless situation, has low expectations for the rest of her life and therefore still gives a high score for her happiness or life satisfaction. This high score does not make her situation desirable; it only shows that this woman has a remarkable ability to adapt. We can illustrate this argument with results based on our MEQIN data. We will first look at the impact of personality traits on happiness and then at the role of expectations. Finally, we illustrate how, despite the overall positive relationship between happiness, and income and health (which we described in Chap. 21), for a significant number of people happiness still shows little correlation with objective dimensions of well-being such as health, housing quality and consumption.

22

As Long as We’re Happy …?

161

Personality and Life Satisfaction Personality can be described on the basis of five dimensions that apply to a greater or lesser extent to a person. These dimensions are openness to experience, conscientiousness, extraversion, agreeableness and neuroticism/emotional stability.2 Figure 22.1 shows that some personality traits do indeed go hand in hand with higher life satisfaction. On average, extrovert respondents and respondents who score higher for the personality trait “emotional stability” tend to be happier and more satisfied with their lives. Neuroticism/emotional stability is connected with the “emotional well-being” that we described in Chap. 6 as a component of health, and as such can be a cause of social concern. However, it seems very difficult to defend the idea that we should, in social policy, focus more on introverts than extroverts. Incidentally, what we have warned about a few times in this book also applies here: correlation does not necessarily imply causation. It may just as well be the case that a lower level of well-being leads to a poor self-image and that this, in turn, leads to greater emotional instability, or that people become more introverted as a result, as vice versa.

Level of Education and Expectations To measure aspirations, we also included three “vignettes” in the MEQIN survey. These vignettes describe an imaginary situation based on a specific level of consumption and health. Table 22.1 describes these three situations. We then ask the respondents how satisfied they would be if they were to wake up the next day in this imaginary situation. As the vignettes are exactly the same for all the respondents, this question allows us to chart how they assess different life situations. This shows us how ambitious their expectations are. After all, a respondent with high expectations will be less satisfied and will give lower scores for the same vignette than a respondent with more modest expectations. The results are summarised in Fig. 22.2. Here, we organise individuals according to their level of education. If there were no systematic differences between the different levels of education, on average each vignette would be assessed in the same way and we would have a perfectly horizontal curve for each imaginary situation. For a positive life description (imaginary situation 3 with high consumption and very good health), there are indeed few differences between the respondents with different levels of education. However, when the vignettes describe worse imaginary situations, the respondents with lower levels of education (on the left in the figure) give higher scores than highly educated people (on the

2

In the MEQIN survey, we used the TIPI questionnaire developed by Gosling et al. (2003). This scale asks people to what extent they agree with ten statements about their personality.

22 As Long as We’re Happy …?

162 8

Life satisfaction

Life satisfaction

8

7

6

5

Less

7

6

5

More

Less

Extrovert? 8

Life satisfaction

Life satisfaction

8

7

6

5

More

Agreeable?

Less

7

6

5

More

Conscientious?

Less

Life satisfaction

8

7

6

5

More

Neurotic/emotionally stable?

Less

More

Open?

Fig. 22.1. Life satisfaction and personality traits

Level of Education and Expectations

163

Table 22.1. Three imaginary situations Imaginary situation 1 Imaginary situation 2 Imaginary situation 3

Personal consumption (euros/month)

Health

500 750 1000

30 60 90

8

Life satisfaction

7 6 5 4 3 2 1

Low

Medium

High

Level of education Imaginary situation 1

Imaginary situation2

Imaginary situation 3

Fig. 22.2. Life satisfaction in three imaginary situations

right in the figure); the latter therefore feel that they would be less satisfied in this imaginary situation. For the first and second imaginary situation, the difference is approximately one point on a scale of 0–10. This effect is relatively major: for example, it is greater than the difference in satisfaction we saw between Belgians in a relationship and not in a relationship. We also find a similar pattern when ranking respondents on the horizontal axis according to their income or health. These findings strongly suggest that highly educated, wealthy and healthy Belgians have greater ambitions for their lives than low-skilled, poor or unhealthy Belgians. They certainly show that there are systematic differences in the way in which respondents interpret the numbers on the response scale of 0–10. The same score for life satisfaction can therefore represent very different life situations. As a result, we must therefore correct the hope that simply asking about happiness or life satisfaction can provide a simple indicator for good and modern policy. Happiness scores depend on personality traits and, moreover, differences in aspirations mean that respondents will each assign a different meaning to the response scale for life satisfaction. As a result, we cannot simply compare scores

164

22 As Long as We’re Happy …?

between different people to determine their well-being. In the following chapters, we will look for an alternative measure of individual well-being.

Poor but Happy In Fig. 21.2, we illustrated that in general there is a positive correlation between life satisfaction and income and health. A closer look at these figures reveals that this correlation is considerably weaker for higher incomes and the most common health scores (more than 85% of Belgians have a health score of 40 or more). In addition, this general correlation conceals a remarkable observation that we are illustrating here: a significant group of people who suffer from cumulative deprivation and are among the most deprived in terms of health, housing quality and material welfare indicate a higher level of satisfaction than a large majority of the people who score highest for these three dimensions. To divide people into groups according to their score for these three dimensions, we followed the same approach as in Chap. 20. For each dimension, we divided the respondents into three equal groups and then selected the group of people that belong to the lowest category for all three dimensions, the group of people that belong to the middle category and those that belong to the category with the highest score for all three dimensions. For health and housing, we used the same average scores as in Chap. 20.3 In contrast to Chap. 20, here, for material welfare, we use personal consumption instead of income. After all, we saw in Chap. 15 that the goods on which the income is spent do not benefit all the family members equally. This personal consumption will also play an important role in the alternative measure of well-being that we propose in the next two chapters. It includes the private consumption of family members (food, clothing, personal care, transport and leisure; see Chap. 15) and a proportionate share of the public expenditure on utilities, recreation and eating out. We do not include housing consumption here as we regard housing quality as a separate dimension, not expressed here in monetary terms. More details about this personal consumption can be found in Chap. 24. Table 22.2 shows that almost 9% of Belgians who belong to the most deprived group according to these three dimensions are nonetheless very satisfied with their current life situation (a score of 9 or 10). As a result, they indicate a higher level of life satisfaction than no less than 72% of people who score highest for all three of these dimensions. Despite possible positive correlations in general, objective dimensions of well-being do not appear to correspond precisely with life

As in Chap. 20, these figures are based solely on the answers of the reference people in the family, as they are the only people who answered the questions about housing quality and living environment.

3

Poor but Happy

165

Table 22.2 Objective dimensions of well-being versus life satisfaction Score for personal consumption health housing quality

Life satisfaction 0–8 (%)

Life satisfaction 9–10 (%)

% of the population (%)

Low, low, low Medium, medium, medium High, high, high

91.2 67.3 72.3

8.8 32.7 27.7

7.5 3.6 5.8

satisfaction. Those who regard life satisfaction as a good indicator of well-being would therefore argue that these 9% of the most deprived are better off than the 72% of those who are actually best off in terms of health, housing quality and material welfare.

How Do We Measure Well-Being?

23

The guiding principle of this chapter, and the whole book, is that various different aspects of life must be taken into account when determining individual well-being. In Part I of this book, we provided individual descriptions of each of these different aspects. Although this is already very useful in itself, it is still necessary for some purposes to bring these different elements under a common denominator. Indeed, a policymaker who examines the various aspects and policy indicators separately, as if they were different lights and gauges on the policy indicator dashboard, is overlooking an important aspect of well-being.1 This is the phenomenon of cumulative deprivation that we illustrated in Chap. 20. In order to take the correlation between the various aspects into consideration, we need an overall measure of individual well-being. Let us suppose that the policymaker wishes to develop a policy to increase the well-being of people in our society. How should the priorities be defined? Which aspects of life should be given particular importance? And if the policymaker mainly wishes to improve the situation of the weakest members of society, how can it be established who achieves the lowest levels of well-being? In the previous chapter, we saw that serious criticism can be levelled at the temptingly simple answer to use a personal assessment of one’s own happiness as a measure of well-being. It is easy to criticise, of course. But is there an alternative? In the last two chapters of this book, we introduce an alternative approach to the measurement of well-being. In this chapter, we explain this alternative approach in an intuitive way based on various concrete examples. In Chap. 24, we then apply this approach using the MEQIN dataset.

1

The social indicators used by the European Commission are one example of such a dashboard. The scientific substantiation of these indicators is described by Anthony Atkinson et al. (2002).

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_23

167

168

23 How Do We Measure Well-Being?

An Example of the Alternative Approach In Table 23.1, we describe the life situations of four fictitious people: Anne, Benny, Chantal and Dirk. Which of the four people should the policymaker prioritise? This is a tough choice. Although Dirk has a very low income, he is extremely healthy and socially well integrated. He also really knows how to enjoy life. Although a higher income would certainly improve his situation, should he be given priority over Benny who can get by on his low income but is constantly in pain and often depressed? Or would it be better to expand the mental healthcare system so that Benny could get better help? Expanding the mental healthcare system could also benefit Anne, who seems to attach so much importance to her career and high income that her health and happiness are suffering and she barely has any friends left. The retired and lonely Chantal also seems to have no need for a higher income: for her, good primary healthcare and a policy that stimulates social contacts are especially important. This example illustrates yet again that while happiness is important, it is not everything. Although Anne is not happy, few of us probably think that she should be given priority. Chantal has adapted to her fate and is neither happy nor unhappy, yet her chronic illness and loneliness seem to be factors that the policy could (or should) respond to. It goes without saying that income cannot be regarded as the only important factor either: in this case, we would need to give Dirk higher priority than Benny, with his serious health problems, while Chantal would be left entirely to her own devices. These conclusions are unlikely to correspond to most people’s intuition. If we do not wish to focus solely on a single row in the table, how can we bring the various aspects of the lives of these four people under a common denominator? It should be obvious that it is impossible to simply add up the different rows in the table. Although income can be measured in euros, health and work situation are expressed in very different units. However, as illustrated earlier in this book, we can also try to assign a score to the non-material aspects of life. If we do so on a scale from 0 to 100, where 0 stands for the worst possible situation and 100 for the best possible situation, this could generate a table like Table 23.2. Table 23.1 Life situations of Anne, Benny, Chantal and Dirk Income Health Work situation Social integration Happiness

Anne

Benny

Chantal

Dirk

Very high Decent, a few complaints Stressful and challenging job Few friends

Low Constant physical pain Disabled, but would like to work Lots of friends

High Chronically ill

Unhappy, does not enjoy life

Unhappy and often depressed

Not happy and not unhappy

Very low Extremely healthy Low-paid, part-time job Lots of friends Happy and enjoys life

Pensioner Lonely

An Example of the Alternative Approach

169

Table 23.2 Life situations of Anne, Benny, Chantal and Dirk quantified Income Health Work situation Social integration Happiness Sum of non-material aspects

Anne

Benny

Chantal

Dirk

6000 euros 80 60 10 70 220

1000 euros 40 70 95 40 245

4500 euros 50 90 10 80 230

500 euros 100 100 100 100 400

Does it make sense to just add up these scores and decide that for the non-material aspects of life, Chantal is better off than Anne (with totals of 230 and 220, respectively) and that the sick (but sociable) Benny is even better off (with a total score of 245) and Dirk reaches unrivalled heights with his score of 400? Although it often happens in practice, simply adding up the scores actually seems like a completely arbitrary operation. In addition, we would still need to establish how these total scores for the non-material aspects should be combined with income. After all, Anne and Chantal have enough income while Dirk is poor.

Different Opinions on the Good Life: Willingness to Pay Simply adding up scores is too simplistic. After all, people have different opinions on what they consider important for a good life. To some extent, this is also reflected in their choices. Anne may choose a stressful job because she loves driving around in her red sports car (admittedly alone), while Dirk has moderate material needs and enjoys life in a completely different way by spending more leisure with his numerous friends. It would seem very strange to judge the well-being of these people without taking into consideration their own views on what is important in life. We wish to give people equal opportunities to realise their own life project after all, but people can have different views on what this life project looks like. How can we take these differences into account? At first glance, there appears to be an easy answer: why not just ask people how important the various aspects of life are to them, for example, by giving them a weighting between 0 and 5? We also included these kinds of questions in our MEQIN questionnaire, but it is not always easy to interpret the answers. Both Anne and Dirk could attach a low weighting to income: Dirk because he doesn’t think material consumption is particularly important and Anne because she has enough money anyway and would now like to improve other areas of her life. Chantal might attach a low weighting to health, because she knows that she can’t change anything about her chronic illness and has learned to live with it. And Benny might attach a lower weighting to all the various dimensions of life than the others, because he’s so depressed that he doesn’t really

170

23 How Do We Measure Well-Being?

care much anyway. Although these weightings certainly mean something, it is very difficult to compare them between individuals. Is there another, more promising approach? One possibility would be to depart from the idea that people’s choices reveal what they consider important. Let us return to Anne and Dirk and assume that they are about the same age and have the same opportunities on the labour market. However, their choices are very different. These choices teach us something about what is important to them. This takes a very concrete form: we can deduce that Dirk “sacrifices” 5500 euros per month in exchange for better health and more leisure with his friends. For Anne, better health, more friends and a less stressful job are worth less than 5500 euros, otherwise, she would try to change her life by looking for a different job. However, this option is not always applicable. After all, we assumed here that Anne and Dirk have complete freedom of choice. This is of course not always the case. People do not have the same opportunities on the labour market because they are less or more productive, because they belong to a discriminated group or simply because they live in a different region and would find it difficult to move. We do not choose some of the important events in our lives. For example, Benny might be depressed and subject to physical pain because he has an unfavourable genetic profile or because he was raised in a very poor family. Chantal’s chronic illness is not a free choice either. From the description of their condition alone, it is therefore not possible to deduce that health is less important to them. However, we can set up a hypothetical exercise that would give us a similar answer. We can ask Benny how much income he would be willing to sacrifice (if he had the choice) to become perfectly healthy again and have an interesting job. We can also ask Chantal a similar question. These are difficult questions and the answers must therefore be interpreted very carefully. In principle, they give us an insight into the relative importance that people attach to the various different aspects of their lives. We call the answer to these questions the willingness to pay. This concept plays an important role in our proposed measure of individual well-being. In the next chapter, we will describe in detail exactly how we measured the willingness to pay for some aspects of life in the MEQIN survey. Here, we will discuss several general characteristics of the concept of “willingness to pay” for one (important) aspect of life: health. Firstly, we would expect people to have a greater willingness to pay for perfect health as they become sicker and further removed from the ideal of perfect health.2 People (like Dirk) who are already in perfect health would have a willingness to pay that is equal to zero. This is a very good illustration of how this data should be interpreted. Just because Dirk is not willing to give up income to be perfectly healthy does not mean that he doesn’t care about health; on the contrary, in fact. We have already seen that he actually misses out on 2

By the concept of perfect health, we refer to the situation in which the health index reaches the maximum, or a score of 100. Although a person in this situation does not have any health problems of the kind we ask about in the MEQIN survey, it does not mean that they can run a marathon or reach other top sporting achievements.

Different Opinions on the Good Life: Willingness to Pay

171

a large increase in income because he considers the other aspects of life so important. The willingness to pay as measured here is one way of placing a monetary value on the loss of well-being due to illness. This loss of well-being is a combination of two factors. Firstly, the extent to which the actual health situation differs from perfect health (which can be described objectively) and, secondly, how much the person is suffering as a result (this depends on his or her individual views on the relative importance of health). Some people may be outraged by this approach, because they feel that this method is designed to put a price on health. There’s no market for health and we want to keep it that way, right? However, their outrage is based on a misunderstanding. The willingness to pay reflects how much income people would be willing to sacrifice for perfect health, but we might as well as ask the opposite question: how much health would people be willing to sacrifice for a higher income? What we wish to do is balance different aspects of the good life against each other in a way that respects the individual views on the relative importance of these aspects. The choice of income as a “unit of measurement” is not substantial: it is simply much easier to answer a question about willingness to pay than a hypothetical question about sacrificing health, because we are used to making calculations in monetary terms. If such a balance between the various aspects of a good life is regarded as pointless, the method is naturally also pointless. However, we consider this a somewhat strange view as people constantly make these kinds of considerations in their own lives. The choice between doing an hour’s overtime and playing sport with friends involves finding a balance, as it were, between extra income and extra health and social contacts. Policymakers certainly cannot afford not to weigh up the various aspects of a good life. When faced with the difficult question as to which of the four people in our example should be given priority, they inevitably make such an assessment. However, we must exercise caution when interpreting the concept of willingness to pay. In our example, we were able to deduce that Dirk sacrifices 5500 euros for a quiet and healthy life. It’s not just about health, but let us assume for a moment that his health is worth 2500 euros to him. Now, let us look at Benny’s situation: he is depressed and in pain, but his income is only 1000 euros and he needs part of it to survive. It is therefore impossible for him to give up 2500 euros in exchange for better health. However, this does not mean that his health means less to him than it does to Dirk. In principle, Dirk had the opportunity to earn a lot more but Benny does not have this opportunity. People with a higher income are in a better position to sacrifice a larger proportion of their income. We could also formulate this differently: because they have a higher income, each euro is less important to them. When interpreting the willingness to pay, we must therefore always bear in mind that this is a relative consideration: we expect to find the greatest willingness to pay among people who are sicker and have a higher income, as health is more important to them and income less important. In the next chapter, we will examine whether we actually find this pattern in our MEQIN dataset.

172

23 How Do We Measure Well-Being?

Equivalent Income How can we now use this information about willingness to pay to create a measure of individual well-being? Let us depart from a new (simpler) example, which is described in Table 23.3. Felix earns 2000 euros per month and Felicia earns 1500 euros. However, they both enjoy perfect health, are completely happy, have many friends and are completely satisfied with their job. It would therefore seem logical to say that the difference between their individual levels of well-being is determined solely by their income differences. Now, let us suppose that Felix gets sick. His well-being will certainly be affected as a result. How much the illness will reduce his well-being depends on his willingness to pay: for example, if he is willing to sacrifice 700 euros to stay healthy, we can regard this amount as a monetary measure of his loss of well-being. We can then calculate his new level of well-being by deducting this willingness to pay from his observed monetary income: this yields 2000 − 700 = 1300 euros, and we can see that healthy Felicia is now achieving a higher level of well-being. This example can be generalised. We can then use “equivalent income” as an overarching concept of well-being. In general, this equivalent income is established by deducting from the actual income the loss of well-being that is incurred by not achieving the best possible situation for the other, non-monetary, aspects of life.3 In the previous example, Dirk’s equivalent income is therefore 500 euros. Anne’s will be higher, because her choices have demonstrated that better health, a restful job and good friends are worth less than 5500 euros to her. When it comes to Benny and Chantal, we will need to determine their willingness to pay based on the hypothetical questions we described earlier. If Benny answers that he would be willing to sacrifice over 600 euros for good health and enjoyable work, it means that he is suffering greatly as a result of his illness. His equivalent income would then be lower than Dirk’s, despite the fact that he has a higher income. Equivalent income is expressed in monetary terms, with the advantage that we can easily use it to perform calculations. However, this obviously does not mean that the non-monetary aspects of life are neglected. On the contrary, the examples have made it clear that ranking people on the basis of their income can generate very different results from ranking them on the basis of their equivalent income. When Felix is ill he achieves a lower equivalent income than Felicia, despite having a higher income. In a realistic situation, Benny’s equivalent income will also be lower than Dirk’s. This will only not be the case if Benny attaches very little importance to the non-monetary aspects of his life. In this case, this would be exactly the result we wanted to achieve: after all, the main objective of the technique is the ability to take into account people’s different views on what they consider important in life.

Interested readers can find more information about equivalent income in the book by Marc Fleurbaey and Didier Blanchet (2013).

3

Equivalent Income

173

Table 23.3 Life situations of Felix and Felicia Income Health Work situation Social integration Happiness

Felix

Felicia

2000 euros Extremely healthy Ideal job Lots of friends Happy and enjoys life

1500 euros Extremely healthy Ideal job Lots of friends Happy and enjoys life

Finally, let us warn against possible terminological confusion. In the previous chapters (such as Chaps. 4 and 5), we often applied a procedure, whereby income was corrected for family size through the use of an equivalence scale. The equivalised incomes obtained by means of this procedure only measure material welfare and do not take the non-material circumstances of the family members into account. The equivalent income introduced in this chapter is a broader concept of well-being. Confusion between the two concepts must therefore be avoided at all costs.

Who Has the Lowest Levels of Well-Being?

24

In the MEQIN survey, we included questions that were intended to measure the equivalent income of Belgians (hence, our choice of the acronym MEQIN that stands for “Measuring Equivalent Incomes”). In order to do this, we need to measure the willingness to pay for various non-material aspects of life. The questions regarding the willingness to pay for these aspects are difficult for respondents and were therefore formulated carefully. In this chapter, we will explain this question more specifically with regard to the willingness to pay for perfect health. Even after careful formulation, this remains a difficult task for the respondents and is also a more experimental exercise from a scientific point of view than the results presented in the previous chapters. Chapter 4 has already described the difference between income and consumption (measured by expenditure). Income reflects how much material welfare can be acquired. Consumption, on the other hand, is a measure of actual material welfare at a given time. With a view to measuring well-being at a given time, we regard consumption as more informative. In this chapter, we use a measure of material welfare that we call personal consumption. In order to calculate this measure, we use detailed questions about private consumption (discussed in Chaps. 10 and 15) as the starting point. As this private consumption can vary greatly between the various family members, this also allows us to take the unequal distribution of material welfare within the family into account. In Chap. 15, however, we saw that in families with more than one person, some of the goods are consumed simultaneously by more than one family member: they might travel together or go to a restaurant together, for example. Other examples include heating, water and electricity. The housing itself consumed the bulk of these public consumer goods. We did not include this in the personal consumption, because we wanted to ascertain the importance that people attach to decent housing and also to non-material aspects of the housing—such as the living environment and relationships with neighbours—in

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_24

175

176

24

Who Has the Lowest Levels of Well-Being?

relation to material consumption. For the other goods which are consumed jointly, we assumed an equal distribution of the effective consumption within the family.1 From this personal consumption, we then deduct the willingness to pay for the non-material aspects of life in order to calculate equivalent incomes. In this chapter, we are therefore not strictly speaking calculating equivalent incomes but equivalent personal consumption. However, because it reads more easily and in order to remain consistent with the standard terminology in the literature, we continue to refer to equivalent incomes. We start with health as a non-material aspect of life and then add housing, ending the chapter by examining the willingness to pay for an ideal job.2

How Do We Measure Willingness to Pay? We will first examine the aspect of health. Here, we explain the survey procedure in broad terms without going into too much detail. The procedure consists of three steps. In the first step, we showed the respondents a visual summary of their own answers for personal consumption and for the five health dimensions that we discussed in Chap. 6. In the second step, we asked the respondents if they would find a new imaginary situation with the same personal consumption and perfect health a better situation over the next 12 months. We presented this new imaginary situation in the same visual way as their own situation in the first step. Since, by definition, no respondent can have better than perfect health, the logical answer to this question is positive. Finally, in the third step, we asked about the monthly personal consumption that, together with perfect health, would lead to a situation that the respondent considers as good as his or her actual situation. This question was as follows: “In this new imaginary situation, you would have perfect health for the next 12 months. To be equivalent to your current situation, you will need to reduce your monthly consumption. How much would you reduce your monthly consumption by over the next 12 months in order to consider your new situation with perfect health to be equivalent to your current situation?”

Of course, it is perfectly possible to answer this question by saying that you are unwilling to sacrifice any consumption. People will do so if they are already completely satisfied with their current health or if they feel that their consumption is already so low that it couldn’t possibly be reduced any further. However, some 1

This is undoubtedly a heroic assumption. However, with no prior information about the preferences of individual family members with respect to these goods and the extent to which the goods are effectively consumed together, there is no other obvious option. Given that the main component of public consumption (housing) does not form part of personal consumption, the impact of this simplifying assumption may not be particularly great anyway. 2 As the questions about quality of housing and living environment were only asked to the reference persons, the data in this chapter are based on answers from the reference persons alone.

How Do We Measure Willingness to Pay?

177

people refused to answer this question because they found it too difficult (even after an additional explanation). These “protest responses” (about a hundred) were removed for this analysis. The distribution of the other answers is shown in Fig. 24.1. The height of the bars indicates the relative number of people who specified the values on the horizontal axis. Around half have a willingness to pay for perfect health that is equal to zero. On the other hand, 4.3% of the respondents are willing to pay more than 400 euros per month. The average willingness to pay is 85 euros per month. If we look only at those who responded positively, the average is 169 euros per month. Since we measure health on a scale from 0 to 100 and also know for each of our respondents how much their actual state of health differs from perfect health, we can then convert these amounts into a willingness to pay for a single unit on the health scale: this yields 9 euros per unit for the respondents who stated a positive willingness to pay. Someone who had difficulty lifting or carrying his groceries, but can now do so, has progressed two points on the health scale: on average, people are therefore willing to sacrifice 18 euros per month (or 216 euros per year) for this. In order to be free of a chronic condition that restricts them in their daily lives (thus progressing 20 points on the scale), people are willing to sacrifice an average of 180 euros per month (2160 euros per year). We are certainly not talking about small amounts here.3 Some of the phenomena discussed in the previous chapter are illustrated in Fig. 24.2. For the construction of the left-hand panel, all the respondents were ranked by health from low to high and then divided into ten equal groups (deciles). The figure shows the average willingness to pay per decile. It appears that the willingness to pay increases significantly as people get sicker, i.e. when the distance between their current state of health and perfect health increases. In the right-hand panel, we do the same for personal consumption. When people reach a higher level of personal consumption, they are also willing to sacrifice more for better health. As described in the previous chapter, this can easily be explained by the fact that people with a higher level of personal consumption are in a better position to afford the same loss of consumption. Figure 24.2 naturally conceals major differences in individual willingness to pay. People with the same health and the same income can nevertheless attach very different importance to better health. It is precisely the intention of this technique to ensure that these differences in opinion about what constitutes a good life are reflected in the measure of individual well-being. In this context, it is interesting to note that the personality characteristics, as shown in Chap. 22 to affect individual life satisfaction, have a smaller impact on the willingness to pay. In the MEQIN survey, we also applied the same procedure to establish willingness to pay for other aspects of life such as housing and ideal job characteristics. We found that the average willingness to pay for trouble-free housing is 47 euros a 3

Moreover, we only asked about the willingness to pay for a better quality of health, not about the willingness to pay to live one year longer. Other studies (e.g. by Ryen and Svensson 2015) show that this latter willingness to pay is much higher.

178

24

Who Has the Lowest Levels of Well-Being?

60%

Percentage of respondents

50% 40% 30% 20% 10% 0%

Euros per month

150 100 50 0 1

2

3

4

5

6

7

Deciles for health

8

9

10

Willingness to pay for perfect health (in €/month)

Willingness to pay for perfect health (in €/month)

Fig. 24.1 Distribution of the willingness to pay for perfect health

200 150 100 50 0 1

2

3

4

5

6

7

8

9 10

Deciles for personal consumpƟon

Fig. 24.2 Willingness to pay for perfect health, per decile for current health status (left) and personal consumption (right)

month, lower than the figure for perfect health. This finding could mean two things: either that Belgians are already in high-quality housing or that they care less about their housing than their health. Both explanations are plausible. The first explanation is in line with what we found in previous chapters. After all, in Chap. 11, we saw that Belgian housing is of decent average quality and in Chap. 22 that this is indeed reflected in (average) high levels of satisfaction with housing quality. Figure 24.3 once again shows that willingness to pay decreases as housing quality improves and increases in line with personal consumption, as for health as shown in Fig. 24.2. We also asked these people how much personal consumption they would be willing to sacrifice in exchange for perfect health and trouble-free housing. The average willingness to pay was about 102 euros. Although this value is higher than the willingness to pay for health alone (i.e. 85 euros), we find that the willingness to pay for the additional improvement in housing is lower than what these people were

179

100 80 60 40 20 0 1

2

3

4

5

6

7

8

Deciles for housing quality

9

10

Willingness to pay for perfect housing (in €/month)

Willingness to pay for perfect housing (in €/month)

How Do We Measure Willingness to Pay?

80 60 40 20 0 1

2

3

4

5

6

7

8

9 10

Deciles for personal consumpon

Fig. 24.3 Willingness to pay for trouble-free housing, per decile for housing (left) and personal consumption (right)

already willing to pay for health plus what they are willing to pay for housing alone. This finding is also logical: if the respondents have already reduced their personal consumption for one aspect (health), they have less scope to reduce their consumption still further for other aspects. Finally, we also asked the people who were employed at the time of the survey how much they would be willing to pay for what they considered an ideal job.4 We can see that 41% of workers would be willing to give up part of their personal consumption in exchange for an ideal job. The average is 57 euros per month. These are surprisingly large sums: an improvement in job quality certainly seems desirable for a substantial proportion of the population. Even though the correlation is less pronounced than for the other aspects, we note once again in Fig. 24.4 that willingness to pay largely decreases as the current job draws closer to the ideal and increases in line with personal consumption.

Equivalent Incomes for Health and Housing The amounts for willingness to pay for non-material aspects of life can now be used to calculate equivalent incomes, the measure of individual well-being that we introduced in the previous chapter. This is the value of the personal consumption that remains after we have deducted the willingness to pay for the non-material aspects of life. We ranked the respondents according to their equivalent income and divided them into ten equal groups (deciles). In Table 24.1, we examine the characteristics of the people belonging to the lowest decile, i.e. the group of people with the lowest levels of well-being. For this table, we calculate two equivalent incomes (shown in the second and third columns). We first deduct only the willingness to pay for perfect health, then the willingness to pay for both perfect health and trouble-free housing. For the purposes of comparison, the characteristics of the people in the 4

In Chap. 9, we explained how we measured the characteristics of this ideal job.

24 100 80 60 40 20 0 1

2

3

4

5

6

7

8

9

Deciles for job characteriscs

10

Who Has the Lowest Levels of Well-Being? Willingness to pay for perfect job (in €/month)

Willingness to pay for perfect job (in €/month)

180

120 100 80 60 40 20 0 1

2

3

4

5

6

7

8

9 10

Deciles for personal consumpon

Fig. 24.4 Willingness to pay for a perfect job, per decile for job quality (left) and personal consumption (right)

lowest decile for personal consumption (in the first column) and life satisfaction (in the fourth column) are also shown. We start by comparing the first two columns, i.e. only looking at personal consumption as a measure of well-being and then deducting the loss of well-being caused by health problems. We note that the people with the lowest levels of well-being in accordance with equivalent income have a higher personal consumption but also a lower health index than when we examine personal consumption alone. This is logical: sick people and people with a high willingness to pay for their health have a lower equivalent income and thus a greater chance of ending up in the lowest decile. We can also see that fewer migrants belong to the group with the lowest levels of well-being. This finding is in line with what we saw for migrants from Eastern European countries in Chaps. 6 and 7: these groups have better health. In the third column, we deduct the loss of well-being caused by health problems and low housing quality. The difference between the second and third columns is minor and could be expected as a result of the low additional willingness to pay for housing. Nonetheless, we find fewer pensioners among the lowest decile when we introduce housing to the measure of well-being. In this finding, too, there is an echo of previous results from Chap. 19. If we now look at the last column, showing the characteristics of people with low levels of life satisfaction, we see a different story. As we saw in previous chapters, life satisfaction is not just determined by the circumstances of the respondents but also by their expectations and aspirations. This is also evident in these results. After all, we can see that people with low levels of life satisfaction are richer, less healthy, less often in a relationship and often highly educated. These findings are consistent with what we could expect on the basis of Chaps. 21 and 22. In Chap. 21, we noted the major importance of health, relationships and income as determinants of life satisfaction. At the same time, we already observed in Chap. 22 that highly educated and richer respondents have higher expectations and aspirations. We also indicated that a significant number of relatively wealthy people specified low levels of life satisfaction.

Equivalent Incomes for Health and Housing

181

Table 24.1 Who are the 10% of Belgians with the lowest levels of well-being? The 10% of Belgians with the lowest levels of well-being according to …

Personal consumption Health Housing Female In a relationship Migrant Highly educated Unemployed Pensioners

Personal consumption

Equivalent income (health)

Equivalent income (health and housing)

Life satisfaction

240.37 euros

329.80 euros

345.92 euros

68.41 79.77 62% 59%

64.17 79.90 57% 58%

64.20 79.65 57% 61%

714.12 euros 52.00 75.99 48% 42%

25% 14%

20% 16%

19% 18%

17% 18%

19% 33%

16% 34%

16% 31%

12% 26%

Table 24.1 illustrates the relevance of the individual measure of well-being that is chosen when devising social policy. In order to elaborate such a policy, it is necessary to identify the people with the lowest levels of well-being so that they can be prioritised. The choice of the individual measure of well-being is therefore not merely an academic exercise. A one-sided focus on material aspects such as personal consumption will lead to the relative neglect of people who are struggling with the non-material aspects of life. A unilateral focus on life satisfaction, on the other hand, will (unduly) favour subjective factors: the people who are least satisfied are not the ones in the worst circumstances. Even though its implementation requires further refinement, a measure such as equivalent income is a promising new approach when it comes to identifying the people who are worst off in our society.

Conclusion

25

We had two objectives in mind when writing this book. Our first objective was to map the individual well-being of the Belgian population. With the help of the Belgian Science Policy Office (BELSPO), we therefore conducted a large-scale survey in 2016 under the name MEQIN. 3404 adults in 2098 randomly selected families across Belgium participated in this study. This innovative survey contains a wide range of information about various dimensions of life (such as health, individual expenditure and time use or living environment) for these individuals. This allowed us to paint a multidimensional picture of the well-being of Belgians. As we asked all the adult members of the selected families about their individual expenditure and time use, we also gained a unique insight into the distribution of well-being and power within these families. Another innovative feature of our research is the fact that we established the relative importance attached by these individuals to various dimensions of life. This information gave us new insights into the recipe for the good life, as well as the diversity in this area. In this conclusion, we will delve deeper into the second objective of this book. Not only did we wish to map individual well-being, but also to test how useful such an exercise can be when devising and evaluating social policy. We wondered how we could use all the available information about the well-being of Belgians to identify their main needs and shape policy. We can draw four general lessons from this exercise.

Just Looking at Averages Can Be Misleading A first general lesson we can learn from this book is that averages can be extremely misleading. We are convinced that those who are worst off in society deserve to be prioritised in social policy. These people are hidden by averages. One example of this was found in our comparison of the expenditure and time use of men and women in Chap. 16. On average, Belgium appears to have a low level of gender © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9_25

183

184

25

Conclusion

inequality. However, when we look beyond this average, we can see that inequalities between partners often nonetheless occur. Some men (and women) have very high individual expenditure, while their partners have very low individual expenditure. This latter group is hidden behind the average. It is precisely this hidden group with low expenditure and little leisure that forms an important target group for social policy. It is therefore important to look beyond averages or other simple statistics and to try to map the distribution of well-being as a whole. To achieve this, we need an operational measure of well-being. But how can we measure well-being?

Income Is Not a Good Measure of Well-being We do not think it is a good idea to use just a monetary benchmark, such as income or assets, to measure well-being and establish the priorities of social policy. The first part of this book contains numerous examples that also demonstrate the existence of major needs and inequalities in non-monetary dimensions such as health, jobs and the quality of housing and living environment. By determining the priorities of social policy on the basis of a monetary benchmark alone, we are closing our eyes—as it were—to all these non-monetary inequalities. Moreover, Belgians consider these non-monetary benchmarks important for their well-being. Health, in particular, appears to play a major role for many of the Belgians. A high income clearly does not guarantee a good life. However, we have seen in several chapters of this book that too low an income often constitutes an obstacle to a good life. The monetary indicators such as income poverty or income inequality that we examined in Chaps. 4 and 5 are certainly not useless, but could best be supplemented with other indicators in order to avoid a one-sided picture of well-being and the distribution of well-being.

Happiness Is Not a Good Measure of Well-being Recently, many observers and policy-makers have been attracted to the idea of using subjective measures of well-being such as happiness and life satisfaction to gauge well-being and determine policy priorities. Indeed, subjective measures of well-being do have several advantages: they are relatively easy to collect and people also take non-monetary dimensions into account when answering questions about happiness and life satisfaction. Nonetheless, we do not think it is a good idea to use subjective measures of well-being to determine policy priorities, for two reasons. First of all, subjective measures of well-being are often relatively insensitive to the objective situation in which people find themselves. After all, people seem to have a remarkable ability to adapt to their objective situation. In Chap. 21, for

Happiness Is Not a Good Measure of Well-being

185

example, we saw that there was no noticeable difference between the average levels of life satisfaction for men and women. Based on this finding, there appears to be no reason to pay particular attention to the situation of women in the policy. Yet in Chap. 20, we noted that women are considerably more likely than men to suffer from cumulative deprivation in multiple dimensions of life. This is also because disproportionately more women than men are single parents, and we also saw that single parents achieve lower scores for various life dimensions. Secondly, subjective measures of well-being are also influenced by expectations and aspirations. A remarkable example was given in Chap. 22. In the MEQIN survey, we asked respondents to state their life satisfaction in various imaginary situations. This showed that people with a higher level of education reported a lower level of life satisfaction for exactly the same imaginary situation. More highly educated people may have higher expectations, and therefore, report lower life satisfaction in the same imaginary situation. Using life satisfaction to determine the priorities of social policy would therefore benefit people with a higher level of education (often associated with higher incomes). This conclusion seems difficult to defend from an ethical viewpoint.

Well-being Is Best Measured in a Multidimensional Way In this book, we have argued that the information from various dimensions of life must be combined if we are to measure well-being in an appealing way. Income and happiness can certainly form two of these dimensions of life, but other dimensions such as health, social interactions, jobs and the quality of housing and living environment are also important. A first step towards a more comprehensive picture of well-being could be to examine all these dimensions of life separately. We did so in the first part of this book, and it revealed some interesting insights. However, we have argued in this book that an analysis of well-being can (and must) go further by examining various dimensions of life together. An initial reason for this was given in Chap. 20. A remarkably high percentage of Belgians suffer from cumulative deprivation: not only do they have low incomes, they also find themselves in the group with the worst health and housing. A separate analysis of each individual dimension of life would never be able to detect cumulative deprivation. Consequently, policy-makers would be unable to prioritise the individuals who achieve extremely low scores for different dimensions at the same time. A second reason is that not all life dimensions are equally important. Health proved more important to many Belgians than housing quality, for example. In order to take into account what Belgians themselves consider important for a good life, a multidimensional measure of individual well-being is required. This measure of well-being can then lend greater weight to life dimensions that individuals consider more important.

186

25

Conclusion

In Chaps. 23 and 24 of this book, we stated that well-being can be measured in a way that takes into account what people themselves consider important in life. To this end, we looked at “equivalent income” as a new measure of well-being. This measure corrects an individual’s income or material welfare for the situation in other dimensions, such as health or housing quality. An appealing feature of this measure is that the correction depends on the importance that individuals themselves attach to these other dimensions for their well-being. For those who consider housing more important than good health, a worse housing situation will carry greater weight than worse health in the income correction. This allows researchers and policy-makers to avoid paternalism and imposing a specific vision of a good life on multidimensional comparisons of well-being. In Chap. 24, we saw that it is mainly unhealthy and poor people who are worst off according to this equivalent income measure. We believe that they deserve to be prioritised in social policy. To implement equivalent incomes, a comprehensive dataset is needed and the method of calculating this new criterion can certainly be further refined. Nonetheless, at the end of this book, we are optimistic about the possibility of social policy being based on a precise and rigorous measure of individual well-being. We therefore hope that this book will give rise to a scientifically substantiated debate on a subject that affects every individual in society.

References

Almond, D. (2006) Is the 1918 influenza pandemic over? Long-term effects of in utero influenza exposure in the post-1940 U.S. population. Journal of Political Economy, 114(4), 672–712. Atkinson, A. B., B. Cantillon, E. Marlier, & B. Nolan. 2002. Social Indicators: The EU and Social Exclusion. Oxford University Press. Becker, G. (1973). A theory of marriage: Part I. Journal of Political Economy,81, 813–846. Becker, G. (1974). A theory of marriage: Part II. Journal of Political Economy,82, S11–S26. Bruno, M.-A., Bernheim, J., Ledoux, D., Pellas, F., Demertzi, A., & Laureys, S. (2011). A survey on self-assessed well-being in a cohort of chronic locked-in syndrome patients: Happy majority, miserable minority. British Medical Journal . https://doi.org/10.1136/bmjopen-2010000039 Case, A., & Deaton, A. (2015). Rising morbidity and mortality in midlife among white non-Hispanic Americans in the 21st century. PNAS,112(49), 15078–15083. Cherchye, L., Demuynck, T., De Rock, B., & Vermeulen, F. (2017). Household consumption when the marriage is stable. American Economic Review,107, 1507–1534. Chiappori, P.-A. (1988). Rational household labor supply. Econometrica,56, 63–89. Chiappori, P.-A., Fortin, B., & Lacroix, G. (2002). Marriage market, divorce legislation and household labor supply. Journal of Political Economy,110, 37–72. Diener, E. (1994). Assessing subjective well-being: Progress and opportunities. Social Indicators Research,31(2), 103–157. Easterlin, R. (1974). Does economic growth improve the human lot? Some empirical evidence. In P. David & M. Reder (Eds.), Nations and households in economic growth: Essays in honor of Moses Abramovitz (pp. 89–125). New York: Academic Press. Fleurbaey, M., & Blanchet, D. (2013). Beyond GDP. Measuring welfare and assessing sustainability. Oxford: Oxford University Press. Gosling, S. D., Rentfrow, P. J., & Swann, W.B., Jr. (2003). A very brief measure of the big five personality domains. Journal of Research in Personality,37, 504–528. Idler, E., & Benyamini, Y. (1997). Self-rated health and mortality: A review of 27 community studies. Journal of Health and Social Behavior,38, 21–37. Kahneman, D., Krueger, A., Schkade, D., Schwarz, N., & Stone, A. (2004). A survey method for characterizing daily life experience: the day reconstruction method. Science306(3), 1776–1780. Kahneman, D., Wakker, P., & Sarin, R. (1997). Back to Bentham? Explorations of experienced utility. Quarterly Journal of Economics,112, 375–406. Layard, R. (2005). Happiness: Lessons from a new science. London: Allan Lane. Myria. Migratie in cijfers en in rechten 2017 [Migration in figures and rights 2017]. Web page. Available at https://www.myria.be/files/MIGRA2017_NL_AS.pdf. O’Donnell, O., Van Doorslaer, E., & Van Ourti, T. (2015). Health and inequality. In A. Atkinson & F. Bourguignon (Eds.), Handbook of income distribution (Vol. 2B, pp. 1420–1533). New York: Elsevier. OECD. (2015). In it together: why less inequality benefits all. Paris: OECD. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2020 B. Capéau et al., Well-being in Belgium, Economic Studies in Inequality, Social Exclusion and Well-Being, https://doi.org/10.1007/978-3-030-58509-9

187

188

References

Pen, J. (1971). Income distribution. New York: Praeger Publishers. Piketty, T. (2013). Le capîtal au XXI° siècle [Capital in the twenty-first century]. Paris: Le Seuil. Ryen, L., & Svensson, M. (2015). The willingness to pay for a quality adjusted life year: A review of the empirical literature. Health Economics,24, 1289–1301. Schokkaert, E., Steel, J., & Van de Voorde, C. (2017). Out-of-pocket payments and unmet need of health care. Applied Health Economics and Health Policy,15(5), 545–555. Sen, A. K. (1985). Commodities and Capabilities. Amsterdam and Oxford: North-Holland. Sen, A. K. (2009), The idea of justice, Allen-Lane. Stiglitz, J., Sen, A., & Fitoussi, J.-P. (2009). Report by the commission on the measurement of economic performance and social progress. Paris. Swinnen, L. (2012). Burn-out. Leuven: Davidsfonds. Schwartz, C. R., & Mare, R. D. (2005). Trends in educational assortative marriage from 1940 to 2003. Demography,42(4), 621–646. Van Doorslaer, E., C. Masseria, & OECD Health Equity Research Group. 2004. Income-related inequality in the use of medical care in 21 OECD countries. Towards high-performing health systems: Policy studies. Chapter 3. Paris: OECD. Vandenbroucke, F., & Vinck, J. (2015). Child poverty risks in Belgium, Wallonia and Flanders: Accounting for a worrying performance. Belgisch Tijdschrift Voor Sociale Zekerheid [Belgian Journal of Social Security],57(1), 51–98. Wilkinson, R., & Pickett, K. (2010). The spirit level: Why equality is better for everyone. London/New York: Penguin Books.