Discretionary Time: A New Measure of Freedom

This page intentionally left blank

Discretionary Time

A healthy work–life balance has become increasingly important to people trying to cope with the pressures of contemporary society. This trend highlights the fallacy of assessing well-being in terms of finance alone; how much time we have matters just as much as how much money. The authors of this book have developed a novel way to measure ‘discretionary time’: time which is free to spend as one pleases. Exploring data from the US, Australia, Germany, France, Sweden and Finland, they show that temporal autonomy varies substantially across different countries and under different living conditions. By calibrating how much control people have over their time, and how much they could have under alternative welfare, gender or household arrangements, this book offers a new perspective for comparative cross-national enquiries into the temporal aspects of human welfare. ROBERT E. GOODIN

is Distinguished Professor of Philosophy and Social & Political Theory in the Research School of Social Sciences at the Australian National University.

J A M E S M A H M U D R I C E is an ARC Research Associate in the Research School of Social Sciences at the Australian National University. ANTTI PARPO

is Administrator of Somero Social & Health Services,

Finland. is an ARC Research Associate in the Research School of Social Sciences at the Australian National University.

LINA ERIKSSON

Discretionary Time A New Measure of Freedom

ROBERT E. GOODIN JAMES MAHMUD RICE ANTTI PARPO LINA ERIKSSON

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521882989 © Robert E. Goodin, James Mahmud Rice, Antti Parpo and Lina Eriksson 2008 This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2008

ISBN-13 978-0-511-38832-3

eBook (NetLibrary)

ISBN-13

978-0-521-88298-9

hardback

ISBN-13

978-0-521-70951-4

paperback

Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

For purposes of both social accounting and behavior modeling, a uniform ‘currency’ in which concepts can be structured and behavioral parameters estimated is of enormous value. Historically, the only such science with such a currency has been economics, where money has served as a measuring rod by which a large number of decisions can be understood, evaluated and aggregated. (Juster 1985a, pp. 19–20) Real economy – savings – consists in the saving of working time . . . Economising, therefore, does not mean the giving up of pleasure, but the development of power and productive capacity, and thus both the capacity for and the means of enjoyment . . . To economise on labour time means to increase the amount of free time, i.e., time for the complete development of the individual. (Marx 1858/1972, p. 148)

Contents

List of figures

page xi

List of tables

xiv

Preface

xvii

Part I

Introduction

1

1

Time and money 1.1 Time matters 1.2 Measuring rods: time and money 1.3 Two surveys, six countries 1.4 Welfare and gender regimes

3 3 7 19 24

2

Discretionary time and temporal autonomy 2.1 The value of temporal autonomy 2.2 Operationalizing discretionary time 2.3 Validating the measure

27 27 34 54

3

The distribution of discretionary time 3.1 Large variation 3.2 Common patterns 3.3 Country differences

61 61 65 65

Part II

67

Time pressure

4

Time pressure: a new problem? 4.1 Time pressure in a broad historical perspective 4.2 Contemporary sources of additional time stress 4.3 Better off but busier

69 70 73 77

5

Time pressure: a new measure 5.1 Conceptualizing time pressure 5.2 Magnitude and sources of time pressure

81 83 86

vii

viii

Contents

5.3 Distribution of time pressure among subgroups of the population 5.4 The best and the worst 5.5 Another example: burden-sharing in male-breadwinner families 6

89 93 95

Is it really an illusion? 6.1 Blaming the victims 6.2 Choice and the quality of options 6.3 Temporal neutrality

99 101 109 111

Part III

113

Welfare regimes matter

7

How welfare regimes differ 7.1 Defining ‘welfare’ 7.2 Welfare policy: a potted history 7.3 Three welfare regimes 7.4 Standard ways of classifying countries 7.5 Other dimensions of welfare: child care, for example

115 115 117 122 125 128

8

A temporal perspective on welfare regimes 8.1 Welfare measures: money and time 8.2 Differing state impacts on temporal welfare: the big picture 8.3 Differing state impacts on parents 8.4 Differing state impacts by household types 8.5 State impacts on subgroups of regime-specific concern

131 132

Welfare regimes and temporal autonomy

149

Part IV

151

9

Gender regimes matter

133 137 141 143

10

How gender regimes differ 10.1 Capturing gender: the challenge 10.2 Two modes of maternalism 10.3 Abstracting models of gender regimes 10.4 Classifying countries: some standard indicators 10.5 Taking lone mothers into account

153 155 157 164 169 171

11

A temporal perspective on gender regimes 11.1 Gender regimes: the big picture 11.2 Gendered impacts on parents

177 178 182

Contents

12

13

14

15

16

ix

11.3 Gender-regime impact on mothers in different household types

185

Gender regimes and temporal autonomy

192

Part V

197

Household regimes matter

How household regimes differ 13.1 Alternative household rules: a broad overview 13.2 Breadwinner rules 13.3 Conventional Dual-earner rule 13.4 Egalitarian rules 13.5 Withdrawal (Divorce) rules The difference that household rules make 14.1 Preliminary methodological remarks 14.2 Effects of alternative household rules: an overview 14.3 The impact of alternative household rules on gender equality 14.4 Alternative household rules and the paternity penalty 14.5 Alternative household rules and custodial versus non-custodial divorced parents

199 201 203 209 210 213 224 224 228 234 235 236

The difference that states make 15.1 How differences are made: policy instruments and social norms 15.2 Temporal consequences of changing household types 15.3 Temporal consequences of changing household rules 15.4 The major difference states make

239

245 252

Alternative household rules and temporal autonomy 16.1 Private choice matters 16.2 Public environment matters

254 254 256

239 241

x

Contents

Part VI 17

Conclusions

Conclusions 17.1 Major findings 17.2 So what? 17.3 Implications concerning public policy

259 261 261 263 267

Appendix 1: Methodology

271

Appendix 2: Data

326

Bibliography

426

Index

454

Figures

1.1 A temporal measure of tax impact: Tax Freedom Day 2005 page 15 1.2 Poverty rates, money and time 17 3.1 Mean discretionary time, overall and by gender 62 3.2 Mean discretionary time, by household type 64 5.1 Magnitude of time-pressure illusion nationwide 87 5.2 Components of time-pressure illusion nationwide 88 5.3 Magnitude of time-pressure illusion, by gender and parental status 89 5.4 Magnitude of time-pressure illusion by household type 92 5.5 The difference between households with the most and least time pressure 93 5.6 Decomposition of the time-pressure illusion of childless dual-earners 96 5.7 Time pressure on parents in traditional male-breadwinner households 97 6.1 Mean wage rates of prime-aged wage-earners with high, medium and low time-pressure illusion (as percentage of national mean) 107 8.1 Pre- and post-government discretionary time, whole population 135 8.2 State impact (decomposed) on discretionary time, whole population 137 8.3 State impact on discretionary time, by parental status 139 8.4 State impact (decomposed) on discretionary time of parents 141 8.5 State impact on discretionary time, by household type 142 8.6 State impact on discretionary time of single-earner couples with children, by earner status 147 11.1 Pre- and post-government discretionary time, by gender 179 xi

xii

List of figures

11.2 State impact on gender gap in discretionary time 11.3 State impact on discretionary time, by gender and parental status 11.4 State impact (decomposed) on mothers’ discretionary time 11.5 State impact on mothers’ discretionary time, by household type 11.6 State impact (decomposed) on coupled mothers’ discretionary time, by employment status 11.7 State impact (decomposed) on lone mothers’ discretionary time 11.8 State impact (decomposed) on working mothers’ discretionary time, lone mothers versus partnered mothers 13.1 Changes in discretionary time through the life cycle, US versus Sweden (first approximation) 13.2 Proportion of wives with higher wage rates than their husbands 14.1 Effect of alternative household rules on average household discretionary time 14.2 Effect of alternative household rules on discretionary time, by gender 14.3 Gender gap in discretionary time under alternative household rules 14.4 Paternity penalty under alternative household rules 14.5 Gap in discretionary time between custodial and non-custodial parents under alternative divorce rules 15.1 Temporal consequences of shifting from single Atomistic individuals to Dual-earner couples without children 15.2 Temporal consequences of shift from Conventional Dual-earner couples without children to Conventional Dual-earner couples with children 15.3 Temporal consequences of shift from Conventional Dual-earner household with children to Gendered Divorce with children 15.4 Temporal consequences of shift from Conventional Dual-earner household with children to MaleBreadwinner household with children

180 182 184 185 187 189

190 201 207 229 233 234 236

237 242

243

244

246

List of figures

15.5 Temporal consequences of shift from Male Breadwinner with children to Most-efficient Breadwinner with children 15.6 Temporal consequences of shift from Conventional Dual-earner with children to Equal Temporal Contribution with children 15.7 Temporal consequences of shift from Gendered Divorce with children to Financially Egalitarian Divorce with children 15.8 Temporal consequences of shift from Financially Egalitarian Divorce with children to Strictly Egalitarian Divorce with children

xiii

247

249

250

251

Tables

1.1 Income and time-use surveys used page 23 2.1 Percentages of the population doing less than minimally necessary in three realms 38 2.2 Subjective time pressure as a function of spare and discretionary time 56 2.3 Satisfaction with life as a whole, as a function of time and money 58 7.1 Welfare regime characteristics 124 7.2 Classifying countries into welfare regimes 125 10.1 Gender-relevant differences across countries 170 10.2 Fundamental divides in gender regimes with respect to mothers 176 12.1 Fundamental divides in gender regimes with respect to mothers (revised) 195 A1.1 Years and national currencies 272 A1.2 Necessary time in personal care (hours per week) 275 A1.3 Half of median household ‘equivalent’ actual time in ‘cross-nationally comparable’ unpaid household labour (equivalent hours per week) 277 A1.4 Hourly costs of child care per child (national currency per hour per child) 279 A1.5 The poverty line (equivalent national currency per year) 279 A1.6 Expected alimony and child support received for households that receive alimony or child support (national currency per year) 281 A1.7 Households’ expected ‘taxes-and-transfers’ (national currency per year) 283 A1.8 Mean actual time in travel to/from work during workdays (hours per day) 292 A2.1 Discretionary time, US 327 A2.2 Spare time, US 332 xiv

List of tables

A2.3 The state impact on discretionary time and the time-pressure illusion, US A2.4 Discretionary time (post-taxes, post-transfers, post-childcare-support) under alternative household rules, US A2.5 Discretionary time, Australia A2.6 Spare time, Australia A2.7 The state impact on discretionary time and the time-pressure illusion, Australia A2.8 Discretionary time (post-taxes, post-transfers, post-child-care-support) under alternative household rules, Australia A2.9 Discretionary time, Germany A2.10 Spare time, Germany A2.11 The state impact on discretionary time and the time-pressure illusion, Germany A2.12 Discretionary time (post-taxes, post-transfers, post-child-care-support) under alternative household rules, Germany A2.13 Discretionary time, France A2.14 Spare time, France A2.15 The state impact on discretionary time and the time-pressure illusion, France A2.16 Discretionary time (post-taxes, post-transfers, post-child-care-support) under alternative household rules, France A2.17 Discretionary time, Sweden A2.18 Spare time, Sweden A2.19 The state impact on discretionary time and the time-pressure illusion, Sweden A2.20 Discretionary time (post-taxes, post-transfers, post-child-care-support) under alternative household rules, Sweden A2.21 Discretionary time, Finland A2.22 Spare time, Finland A2.23 The state impact on discretionary time and the time-pressure illusion, Finland A2.24 Discretionary time (post-taxes, post-transfers, post-child-care-support) under alternative household rules, Finland

xv

335 338 344 349 352

355 361 366 369

372 378 382 385

388 394 398 401

404 410 414 417

420

Preface

One thing leads to another, in scholarship as in life. The project reported here grows out of Goodin’s earlier collaboration with Bruce Headey, Ruud Muffels and Henk-Jan Dirven on The Real Worlds of Welfare Capitalism. That book used a very different sort of data from ongoing ‘panels’ to examine the real impact of each of the three main types of welfare regimes on various standard indicators of social concern: poverty, equality, efficiency, social stability, social integration. In addition to all those standard indicators, we thought that we ought also try to assess their impact on people’s ‘autonomy’, somehow construed. We floundered searching for a good measure. Eventually we hit upon one that seemed particularly telling: a time-and-money measure of ‘combined resource autonomy’, representing the proportion of the population earning at least a poverty-level income and spending no more than the internationally agreed maximum of 40 hours a week in paid labour to do so. The proportion of people failing that standard might be said to be in ‘time-or-money poverty’. In countries representing corporatist and social-democratic welfare regimes, that proportion was only a shade higher than the proportion in money poverty alone. But in the liberal US welfare regime, where money poverty was already twice as high as in the other two regimes, ‘time-and-money poverty’ was double that again – a whopping 35 per cent.1 We were mightily impressed by that finding. So was Nobel Laureate Robert Solow, in his long and appreciative review of our work in the New York Review of Books.2 Clearly, it was something that merited much further investigation. Alas, there was no way of pursuing the matter further within the confines of standard data sets focusing on

1

Goodin et al. 1999, pp. 276, 312.

2

Solow 2000.

xvii

xviii

Preface

income alone. A different approach, employing specifically time-use data, was clearly required. Hence the present project. Goodin mapped out the basic conceptual strategy for calculating ‘discretionary time’ in a paper for a May 1998 ‘Workshop on Social Policy and Political Theory’ convened by Stein Ringen for the Maison des Sciences de l’Homme in Paris. Invaluable early advice was received at that point from Abram de Swaan, Bernd Marin, Ulrich Mu¨ckenberger, Claus Offe, Einar Overbye, Stein Ringen and Philippe Van Parijs. Later we benefited from the hospitality of the Institute for Social and Economic Research at the University of Essex and received excellent advice from Kimberly Fisher and Jay Gershuny on how the Multinational Time Use data set might best be used in our project. Along the way, we benefited from the continuing advice of several of ANU’s distinguished Adjunct Professors – Nancy Folbre, Bob Haveman, Claus Offe, David Soskice and Bobbi Wolfe – during their recurring visits to RSSS over many years. Extraordinarily helpful comments on near-final drafts of the book as a whole have come from Nancy Folbre, Paul ‘t Hart, Kieran Healy and Sandy Jencks. Our initial attempts at implementing that strategy empirically came in a pair of preliminary, exploratory papers. One was co-authored with Michael Bittman and Peter Saunders, the other with Olli Kangas. While various aspects of our methodology have shifted since, and the findings reported here supplant those earlier ones in certain ways, we remain greatly indebted to those early collaborators for help getting us started. Versions of these arguments have been presented at various conferences and seminars: to the annual conference of RC19 of the International Sociological Association, meeting in Orviedo, Spain; to the ‘Time Use and Economic Well-Being’ conference of the Levy Economics Institute of Bard College, Annandale-on-Hudson, New York; to the International Association for Time Use Research meetings in Halifax, Nova Scotia, and Copenhagen; and to seminars at the University of Bergen, the Norwegian Business School, the Stockholm Institute for Future Studies, the University of Turku, Social Policy Research Centre, UNSW, Sydney and ANU. For discussions then and later, we are grateful to Sara Arber, Tony Atkinson, Christine Benesch, Geoff Brennan, Frank Castles, Kenny Easwaran, Marc Fleurbaey, Marzia Fontana, Bruno Frey, Paul Frijters, Jay Gershuny, Diane Gibson, Bruce Headey, Karl Hinrichs, Charlotte Koren, Andrew Leigh, Bernard Manin, Julie McMillan, Sue Mendus, Jane Millar,

Preface

xix

David Miller, Dennis Mueller, Claus Offe, Joakim Palme, Axel West Pedersen, Thomas Pogge, Alf Erling Risa, Richard Rose, Tim Smeeding, Cass Sunstein, David Tait, Bertil Tungodden, Philippe Van Parijs and Robert van der Veen. We are grateful to them all, and have only ourselves to blame for shortcomings that remain. Above all, for permission to use their data we are grateful to the Australian Bureau of Statistics, the Multinational Time Use Study, the German Institute for Economic Research, the Luxembourg Income Study, Statistics Finland and Statistics Sweden. Our work on the ‘Discretionary Time Project’ has been carried out with the financial support of Australian Research Council Discovery Grant DP0450406, for which we are also most grateful. Finally, we are grateful for the permission of the editors, publishers and our other co-authors of the following articles to rework some material from them for this book: Lina Eriksson, James Mahmud Rice and Robert E. Goodin, Temporal aspects of life satisfaction, Social Indicators Research 80 (3) (Feb. 2007), 511–33 # Springer Netherlands. Robert E. Goodin, Antti Parpo and Olli Kangas, The temporal welfare state: the case of Finland, Journal of Social Policy 33 (4) (Oct. 2004), 531–52 # Cambridge University Press. Robert E. Goodin, James Mahmud Rice, Michael Bittman and Peter Saunders, The time-pressure illusion: discretionary time vs free time, Social Indicators Research 73 (1) (Aug. 2005), 43–70 # Springer Netherlands. James Mahmud Rice, Robert E. Goodin and Antti Parpo, The temporal welfare state: a crossnational comparison, Journal of Public Policy 26 (3) (Sept.–Dec. 2006), 195–228 # Cambridge University Press.

PART I

Introduction

1

Time and money

1.1 Time matters Recent theorizing about politics has been characterized by a quest for an appropriate ‘currency of egalitarian justice’.1 Time has some very special properties that combine to make it a particularly apt candidate for that status.2

 Time is inherently egalitarian. Everyone has just 24 hours in a day. Some people may value time more than others.3 Still, an hour is the same for everyone, everywhere. That makes it a natural metric for social comparison. Time is inherently scarce.4 No one has more than 24 hours in a day.  Some people’s projects are more time-consuming than others’, and some people’s lives last longer than others’. Still, virtually everyone agrees that more time would be better.5 That makes time a resource that is always scarce relative to demand.

1 2

3

4 5

Sen 1980; Cohen 1989; Kymlicka 2001. At least among people in normal circumstances, by which we mean able-bodied, prime-aged people who are not involuntarily unemployed. Theories of justice often have to be heavily adapted to accommodate the disabled (Brock 1995; Dworkin 1981; 2000; Nussbaum 2006, chs. 2–3; Silvers et al. 1998; Stark 2007; tenBroek and Matson 1966); the disabled may need more time to perform the same tasks, and in justice they should get it. Others such as the young, the old and the involuntarily unemployed might suffer the opposite problem – too much time and too little to do (Jahoda et al. 1933/1971) – and the currency of time might not be the most relevant way of specifying what, in justice, they most need. There is a large literature on the ‘social meaning of time’ exploring those differences, which we will largely be eliding in this book; cf. e.g. Adam (2004). Indeed, it is the ultimate scarce resource (Zeckhauser 1973). The unemployed, for whom time drags, wish not for ‘less time’ but rather for ‘more to do’ in it. As noted above, however, our study is confined to able-bodied prime-aged people in employed households.

3

4

Introduction

 Time is a necessary input into anything that one cares to do or to become.6 Some people make better use of their time than others, getting more done in the same amount of time. Still, everyone needs some time to do or become anything. That makes time a universal good. Those facts combine to make ‘time’ a particularly apt currency for egalitarian justice.7 What we ought to be concerned with, more precisely, is the just distribution of control over the resource of ‘time’. When we say that someone ‘has more time’ than someone else, we do not mean that she has literally a twenty-fifth hour in her day. Rather, we mean to say that she has fewer constraints and more choices in how she can choose to spend her time. She has more ‘autonomous control’ over her own time. ‘Temporal autonomy’ is a matter of having ‘discretionary control’ over your time.8 We offer more elaborate definitions and more precise operationalizations in chapter 2 below. The basic idea, however, can be simply stated. There is a ‘realm of necessity’, in which there are certain things you simply have to do.9

 You have to satisfy bodily necessities: you have to spend at least a minimal amount of time eating, sleeping and otherwise taking care of your body. 6

7

8

9

As modern economists put it, time and capital are the two fundamental inputs into production functions for ‘well-being’ (Becker 1965; Juster and Stafford 1985a, pp. 2–4; Dow and Juster 1985). Within Marxian economics, ‘direct labour time . . . is the determinant factor in the production of wealth’ (Marx 1973, p. 704, emphasis added; see further Postone 1978). Existing indicators of well-being already include some things with a temporal dimension, such as ‘life expectancy’ (Sen 1999, ch. 4; World Bank 2005, pp. 211–23; WHO 2005, pp. 149 ff.). Our aim here is to temporalize such indicators in a more thoroughgoing fashion. As Campbell et al. (1976, p. 349) say in their pathbreaking study of The Quality of American Life: ‘if ‘‘time is money’’, then it is worth bringing time itself within our circle of personal resources. Of course all mortals are endowed with the same 24-hour day in which to live, but the proportion of that day available as a discretionary resource varies widely according to age, occupational status, and the like.’ Unfortunately, their own brief discussion of those issues goes on to conflate ‘discretionary time’ with ‘spare time’ (which we distinguish sharply in sec. 2.2.5 and part II below) and with ‘leisure time’ (which is different yet again, being that subset of spare time in which people are actually engaged in ‘subjectively gratifying activity’: Andorka 1987, p. 151). The phrase is Marx’s (1858/1972, p. 145), but this particular elaboration of it is our own.

Time and money

5

 You have to satisfy financial necessities: you have to spend at least a minimal amount of time securing the cash that you need to purchase the things you need from the cash economy.  You have to satisfy household necessities: you have to spend at least a minimal amount of time cooking, cleaning, taking care of the kids and otherwise keeping your household functioning. Exactly how much time you have to spend in each of those sorts of activities is something we will be calibrating over the course of this book. For now, let us just agree that it is necessary to spend at least a certain amount of time in those necessary activities of daily life.10 The time beyond that necessary to attend to necessary functions is yours to use as you please. That is what we will call ‘discretionary time’. That is how much ‘temporal autonomy’ you possess.11 Note that ‘necessity’ here relates to a social standard, not a natural one. It is not literally impossible for people to do less in each of these dimensions; indeed, as we shall see in chapter 2, around a tenth of people do so in each of these three dimensions on the operationalizations we have chosen. Then again, it is perfectly proper that we do not set our poverty line so low that literally no one in the country falls below it (convenient though politicians might find that). The poverty line demarcating what is ‘necessary’, in money and all these other dimensions as well, represents not a threshold below which it is physically or logically impossible to fall. Just as people sometimes fall below the threshold of financial necessity and are in poverty, people sometimes fall below the threshold of household necessity (and are prosecuted for child neglect) or below the threshold of bodily necessity (and are sleep-deprived). The poverty line represents instead a threshold below which it is socially unacceptable to let people involuntarily fall.12 People falling below those thresholds are simply not satisfying crucial preconditions for participating fully in the life of their society.13 10

11

12

13

This is the time that is ‘socially necessary for labour to reproduce itself’, as Marx (1867/1906, vol. I, p. 208; cf. 1858/1972, p. 144) put it. Marx (1858/1972, pp. 144–9) similarly emphasizes the importance of what he alternately called ‘free’ or ‘disposable time’, operationalized differently to our way. When, as occasionally happens, a person falls below those thresholds through his or her own genuinely voluntary choices, we often (but not invariably) tend to think that there is nothing that socially ought be done to alter that. Townsend 1979.

6

Introduction

When saying that it is socially ‘necessary’ for people to spend a certain amount of time in a certain activity, we mean merely that socially it is only to be expected that they should do so, and that they are not socially criticizable insofar as they do. People are responsible for what they do with their discretionary time, in a way they are not for spending the minimum amount of time that is socially necessary on necessary activities of daily life. That is the respect in which we deem of the one a ‘choice’ and the other a ‘necessity’. The amount of discretionary time left over, after deducting the strictly minimal amount of time devoted to those necessary activities, is greater for some people than others. It is greater in some sorts of households than others. And it is greater in some sorts of countries than others. Here we undertake a cross-national examination of how discretionary time and temporal autonomy vary across the three classic welfare-gender regimes: the social-democratic welfare regime and female-friendly gender regime exemplified in our study by Sweden and Finland; the corporatist welfare regime and traditionalist gender regime exemplified by Germany and France; and the liberal welfare regime and individualist gender regime exemplified by the US and Australia. In the next chapter, we describe and defend our operationalization of temporal autonomy through the notion of ‘discretionary time’. In part II, we show that that notion is distinct from, and as a measure of temporal autonomy it is superior to, more familiar notions of ‘spare time’ (or ‘free time’ or ‘leisure time’). People typically work far longer hours in paid labour than they would strictly need to do purely in order to escape poverty. It is perfectly reasonable that they should do so. When they do, however, they come out looking time-poor on those more familiar notions. But that cannot be right. By definition, ‘avoiding poverty’ defines the limits of strict necessity. Insofar as people work longer hours than strictly necessary for that, purely by their own choice (because they prefer a higher income than the minimum necessary), that should be seen as an exercise of their temporal autonomy, not a constraint upon it. Welfare economists equate ‘welfare’ with ‘being in a chosen position’. By that standard, people who work longer hours by choice rather than necessity should be regarded as having more welfare, not less. Our notion of ‘discretionary time’ tracks those intuitions well, whereas ordinary measures of time pressure couched in terms of ‘spare time’ and its cognates do not.

Time and money

7

In parts III and IV of the book, we look across our six countries to try to surmise what differences the different sorts of existing social policies there actually make to people’s temporal autonomy. Part III concerns the impact of different welfare regimes on people’s temporal autonomy, and part IV concerns the impact of different gender regimes. In part V of the book, we try to surmise how people’s temporal autonomy might differ depending on how their household is organized. We explore the effects of different rules for governing one’s household through a series of counterfactual thought experiments, projecting the same population first into one household regime (households run on one sort of rule), and then into another. We also show how different states’ differing social policies exacerbate or alleviate the differences that different household regimes make to people’s temporal autonomy. How exactly these three regimes interrelate with one another is an open question meriting much further investigation, using a variety of other techniques. Our speculation is that state policies operating on and through choices among ‘household regimes’ constitute microfoundations, certainly of the ‘gender regimes’ and even of certain aspects of the ‘welfare regimes’. That is to say, we speculate that a different mix of household-sharing rules would be found in different gender and welfare regimes; and that those differences are traceable, in turn, to the different styles of each of those gender and welfare regimes. But different data and different styles of analysis would be required to test those speculations. For the purposes of the present work, we will simply treat ‘welfare regimes’, ‘gender regimes’ and ‘household regimes’ as three distinct influences on people’s temporal autonomy. To foreshadow, our findings will be that the most important influences on people’s temporal autonomy are life-cycle changes: marrying, having children, divorcing. The main impacts of welfare states, gender regimes and household regimes on people’s temporal autonomy come by making those life-cycle changes more or less costly, in time terms. Those impacts are often substantial, and can vary substantially across the different regime types.

1.2 Measuring rods: time and money The sources of human satisfaction and dissatisfaction are many and varied, within any given individual’s own life and across individuals. Economists try to render all those multiple contributions commensurable

8

Introduction

through ‘the measuring rod of money’.14 Economists value things that are literally bought and sold directly at their sale price. But many things of value to us are not bought and sold.15 Economists bravely persist in trying to bring them ‘indirectly into relation with the measuring rod of money’ through ingenious techniques of ‘shadow pricing’. There are various familiar problems with economistic attempts to bring all human values – even just everything we comfortably regard as ‘goods and services’16 – under the measuring rod of money. We rehearse some of them briefly in section 1.2.1 below, by reference to the particular problem of valuing the unpaid household labour involved in what economists call ‘home production’. That is a productive contribution of a relatively straightforward sort.17 Economists in the OECD and in various national statistical offices agree it ought to be included somehow in the National Accounts statistics. The difficulties they encounter in bringing even that under the measuring rod of money is a harbinger of even greater troubles to come in applying their measuring rod yet further afield. The measuring rod of money is not the only one possible. A ‘measuring rod of time’ could serve as an alternative metric.18 In many ways it is a more natural metric. As we have already observed, everyone has exactly 24 hours in a day. If our aim is to render things commensurable in welfare terms, then looking at how much time they cost people (or how much time people are willing to devote to them) might be one very good way to do so.19 14 15

16

17

18

19

Pigou 1932, p. 11. And even for those that are, the sale price understates the value of the thing to the buyer, who obviously must value it more than he or she paid for it. This ‘consumer’s surplus’ is uncounted by the measuring rod of money. There are many things we resist thinking about in these ways altogether, of course (Douglas and Isherwood 1979; Zelizer 1985; 1994; Radin 1996). Cf. Folbre (1991) on how the notion of ‘the unproductive housewife’ was constructed in the nineteenth century. Just such a metric lies at the heart of Marxian economics; the value of a thing derives from the amount of labour time invested in producing it. Different though it is in many other respects, our exercise might be seen as being in broadly the same ‘temporal’ spirit. As that phrasing reveals, we shall here be employing what some call a ‘commodified’ as well as a ‘quantified’ notion of time (Thompson 1967, p. 61; Adam 2004, p. 126). That places us more in the tradition of Benjamin Franklin’s Poor Richard’s Almanac (‘he that is prodigal of his hours is, in effect, a squanderer of money’) than Wordsworth’s The Prelude and its complaints about

Time and money

9

From one angle, this might seem a highly novel suggestion. From various other angles, however, it is a very familiar one.

 Think of criminal sentencing. We are quite accustomed to calibrating the badness of criminal acts in temporal terms, ‘jail time’. In certain jurisdictions, even non-custodial criminal penalties are expressed in temporal terms, with a person’s fine explicitly being expressed in terms of a certain number of days’ wages (section 1.2.2).  Think of anti-tax rhetoric. Right-wing organizations talk about how much time it takes to earn enough to pay your tax bill for the year. They acidly observe that you spend longer working for the government than you do paying off your mortgage. Translating tax dollars into temporal terms in these ways is a powerful, and familiar, rhetorical trope (section 1.2.3).  Think, finally, of ‘time poverty’. We all know about the ‘working poor’, people whose wages are so low that they cannot escape poverty even working full-time. There are also people who manage to avoid being ‘money poor’ only by making themselves ‘time poor’, working terribly long hours often in multiple jobs. Looking at poverty in a joint time-and-money framework reveals important differences in how different socio-economic regimes impact on people’s welfare (section 1.2.4). Each of those examples will be elaborated further below. Taken together, those familiar ways of talking and thinking should serve to remind us that there often are both temporal and monetary metrics that can be used in assessing social arrangements of concern to us. They should further serve to remind us that the ‘measuring rod of time’ can be superior to the ‘measuring rod of money’ for many purposes. In the next chapter we will suggest another way of systematically combining both time and money under a temporal metric – one that in our view best captures the notion of ‘temporal autonomy’ and welfare associated with it.

the ‘skilful usury of time’ (quoted in Thompson 1967, pp. 89, 97). Sociologists rightly point to the narrowness of that ‘taken-for-granted feature of our lives’, and point to a wide variety of richer ways of understanding time (Adam 2004, pp. 125–8; see similarly Sorokin and Merton 1937; Nowotny 1994; cf. Gershuny and Sullivan 1998, pp. 70–2). Those lie outside the present work, however.

10

Introduction

1.2.1 Factoring home production into the National Accounts The National Accounts are supposed to be a comprehensive measure of all economic activity within the country. What they actually measure are cash flows. Anything that is not bought and sold for money is not automatically counted in ordinary National Accounts statistics. Any goods exchanged in barter, in the formal or informal economy, are typically left out. So too is anything produced for one’s own consumption. So too are the goods and services produced and consumed within the household. All these things are typically largely missing from National Accounts. But it is quite wrong to suggest that they are of no value, even in a narrowly economic sense.20 ‘All of these activities are productive in an economic sense’, as the UN Statistics Office is the first to concede in its 1993 update of the System of National Accounts.21 Furthermore, it would be quite wrong to imagine that the magnitudes involved are small. One early calculation suggested that, if all married men divorced their wives and hired them back as housekeepers, National Income would double.22 Economists are anxious over the vast amounts of unpaid household labour that does not, but really ought to, get counted in National Accounts statistics.23 The OECD has been actively developing ways 20

21

22

23

‘Counting for nothing’ women’s work in the household, in the acerbic title of Marilyn Waring’s (1988) powerful critique. UN Statistics Division 1993, sec. 1.21. As a report from the OECD Statistics Directorate expands: ‘A household is an institutional unit which is responsible for, and manages, the production of goods and services. In the production process it uses its labour, capital and market goods. In this sense, household production can be compared to production in the market’ (Varjonen and Hamunen 1999, para. 7). Clark 1958. Preliminary results of the ‘satellite accounts’ projects described below confirm that non-SNA production may well be over 60 per cent of all ‘extended private consumption’ even in rich, developed countries like Finland and Germany (Goldschmidt-Clermont and Pagnossin-Aligisakis 1999, pp. 524–5). The UN Statistics Division (1993, sec. 1.21) worries that we have much firmer evidence on money flows, and only much less reliable estimates of non-monetary flows; and it would be a mistake to let the less reliable data swamp the more reliable. If calibrating in the metric of money, they might be right. We suggest instead calibrating in the metric of time, however, where we have almost as good information on time in unpaid household labour as on time in paid labour. (But only ‘almost’, because we have no good way of cross-checking purely survey-based information on unpaid labour, whereas survey-based information on paid labour can be cross-checked against employers’ reports; see e.g. Bersharov et al. 2006).

Time and money

11

of adjusting the National Accounts accordingly.24 Various countries are busily developing systems of ‘satellite accounts’, in which time-use surveys are used to generate estimates of how much time people spend in ‘unpaid household labour’, a monetary value is then assigned to that, and that value is then added to a household’s money income to get a ‘full income’ measure.25 The difficulties in ascribing a monetary value to things that are not actually bought and sold for money are legion, however.26 For just one example of particular relevance to the concerns of this book, there are two standard options for assigning a monetary value to such goods and services. One is an ‘opportunity cost method’. That values the time spent in unpaid household labour at the rate of pay that that person could (given her human capital characteristics) command in the paid labour market. The other is a ‘replacement cost method’. That values unpaid household labour at the price the household would have to pay to hire in someone else to perform the same services. Those two methods can come dramatically adrift in the values that they assign to the same activity.27 Imagine a corporate lawyer who spends 5 hours a week cooking and cleaning her house – activities that she could in principle contract out to hired help. On the ‘opportunity cost method’, her 5 hours a week of unpaid household labour would be valued at her hourly wage rate net of tax (say, $200 per billable hour, or $1,000 in total). On the ‘replacement cost method’, those 5 hours a week would be valued only at what it would cost to get the cheapest cook-housekeeper to perform the same services ($20 an hour, say, or $100 in total).28 24

25

26 27

28

Beckerman 1978; Goldschmidt-Clermont and Pagnossin-Aligisakis 1995; OECD 1995. Varjonen and Hamunen 1999; Holloway et al. 2002. See similarly work by economists outside government: Becker 1965; Garfinkel and Haveman 1977; 1978; Eisner 1988; Saunders et al. 1994; Ironmonger 1996; Smeeding 1997; Smeeding and Marchand 2003. Self 1975. They do so whenever supply and demand prices are not equal, and there is a ‘surplus’ that the consumer or producer can then appropriate. Within the ‘replacement cost method’, there are further choices to be made as regards which wage rates to use: of specialized workers in market enterprises (e.g. cooks); of specialized workers paid to work in the home (e.g. nurses or cleaners); or generalist workers who are ‘polyvalent’ substitutes for all the various activities in the home. These wage rates differ, sometimes dramatically, introducing yet further indeterminacy in the estimates (Varjonen and Hamunen 1999, paras. 28–30).

12

Introduction

There is something to be said for and against each way of performing that calculation.29 But that is actually bad news for the economic attempt to put a shadow price on unpaid household labour. It means that there is no clear, determinate solution.30 Note well the source of that problem: it comes from trying to force ‘time in unpaid household labour’ into the metric of money. That immediately suggests an obvious alternative. If the ‘measuring rod of money’ proves problematic, perhaps we ought to try to perform our welfare assessments – which is in the end what the National Accounts statistics are (or anyway aspire to be) – in a metric of time instead.31 In the metric of time, the housework performed by the lawyer is ‘valued’ at 5 hours a week. That is how long she spends doing it; that is how valuable it is to her for her to do it; and that is the end of the story. Using the ‘measuring rod of time’, there is no attempt to cash out that time in terms of money, so the problem of choosing between the ‘opportunity cost’ and ‘replacement cost’ methods of putting a cash value on time simply do not arise.

1.2.2 Custodial sentences and day fines Thinking in temporal terms comes quite naturally to us in connection with criminal sentencing. Convicted criminals are sentenced to a certain number of days, months or years in prison. ‘Time in the slammer’ is the way judges calibrate the seriousness of offences. ‘Serving their time’ is how convicted criminals ‘pay their debt to society’.32 In many places, time-based thinking is being extended to non-custodial criminal sentencing as well. It has long been common for criminal activities to be punishable by fines rather than incarceration. Traditionally these fines have taken the form of ‘tariffs’, either flat-rate 29

30

31 32

Compare, for example: Becker 1976, ch. 11; 1981; Goldschmidt-Clermont (1993); Varjonen and Hamunen 1999, paras. 5, 16–33; Abraham and Mackie 2005, pp. 28–34. Indeed, if (and how can they not?) individuals have preferences over alternative uses of their time, and they derive differential utility from how they use their time as well as the household goods and services produced by them, then it is impossible to determine a shadow price of the produced good independently of preferences for goods and time use (Pollak and Wachter 1975; Dow and Juster 1985, p. 401). For a pioneering suggestion as to how that might be done, see Gershuny (2005b). It was not always so. Prior to the rise of the prison system, convicted criminals were more typically branded, mutilated or executed (Foucault 1977).

Time and money

13

or within a moderately narrow band, fixed ‘without regard to the financial circumstances of individual offenders’. That makes fines highly unsatisfactory as a mechanism of deterrence or punishment. ‘When tariffs are set at low levels, the fines have little punitive or deterrent effect on more affluent offenders. When they are set . . . higher . . . collecting the fine amount from poor defendants is difficult or impossible, and, in many cases, these defendants are eventually given jail sentences.’33 One way around that problem would be simply to broaden the range of possible fines and to instruct judges to fix a fine with regard to the offender’s economic circumstances as well as (obviously) the severity of the offence. Some countries have gone that route. Beginning in Scandinavia at the start of the twentieth century, however, other countries have developed a more elaborate system of ‘day fines’.34 What such fines actually take from people is of course money rather than days, in any literal sense. But those monetary fines are fixed in a two-step process. The first step calibrates the ‘seriousness’ of the offence by fixing how many ‘days’ the offender should suffer. The second step then determines how much money the offender should be fined to accomplish that, given the offender’s wage rate.35 The rates and rationales of day fines differ from one country to another. In Germany, such fines ‘were introduced mainly as an alternative to short-term imprisonment’.36 There, the day fine ‘is calculated as, in effect, the cost of a day of freedom. The [German] day fine is, in theory, the offender’s net income for a day without deduction for family maintenance. The daily net income the income the offender would have forfeited had he been imprisoned is the amount he must pay.’37 The Swedish system ‘is far less onerous for the offender, being

33 34

35 36

37

Mahoney et al. 1996, p. 1. See similarly Hillsman 2002, p. 2. English-language sources tend to say ‘the concept of day fines was first introduced in Sweden in the 1920s and quickly incorporated into the penal codes of other Scandinavian countries’ (Mahoney et al. 1996, p. 3). The Swedish parliamentary committee established to revise the 1927 bill (eventually enacted, in revised form, in 1931) took advice, however, from Finland where the day-fine system was introduced in 1921 (Sweden 1931, p. 16). For a fuller account of this history see Eriksson and Goodin (2007). Mahoney et al. 1996, p. 1. And very successfully so: ‘Between 1968 and 1976, the number of prison sentences with terms shorter than 6 months dropped from more than 110,000 to aproximately 10,000 – a 90 percent decrease’ (Mahoney et al. 1996, pp. 17, 3). Morris and Tonry 1990, p. 143. See also Hillsman 2002, p. 3.

14

Introduction

roughly . . . one fourth of total daily income’.38 In terms of its rationale, ‘The Swedish system is in effect a system for depriving offenders of the pleasures of life, those expenditures that are made after basic living expenses are met,’ for a set number of days.39 Once fines are tied in this way to the number of days offenders must work to pay them off, experience from across a number of West European jurisdictions suggests that ‘fines become more widely usable as a stand-alone sanction across a broader array of offences of varying level of severity. This is because the fines can be used both consistently and equitably, and because they are enforceable in the community without high rates of default resulting from indigency.’40 Thus, just as we have long been accustomed to custodial sentences being calibrated in terms of time, there is a strong case for calibrating non-custodial sentences similarly. In criminal justice, time really does seem to be the most useful metric.

1.2.3 Tax as ‘time working for the government’ Ronald Reagan once quipped that ‘a taxpayer is someone who works for the federal government but who doesn’t have to take a civil service examination’.41 Reagan might have added that that is not a job from which one can readily resign: the coercive apparatus of the state will be used, often in extraordinary ways, to compel payment of taxes. For such reasons, right-wingers tend to liken taxes to ‘forced labour’ and talk of ‘tax slavery’.42 Celebrating the Reagan legacy, the House Majority Whip revisited Reagan’s claim that ‘Americans are over-taxed, over-regulated and over-governed’ and restated it in time terms: ‘It is not right that most people spend half the year working for the government.’43 Representative DeLay was alluding here to an evocative time-based trope developed by right-wing think tanks for use in their anti-tax campaigns. ‘Tax Freedom Day’, as they call it, is that day in the year when the average citizen stops working for the government (i.e., has earned enough income to pay his or her tax bill) and begins working for 38 39 40 42

Mahoney et al. 1996, p. 17. Morris and Tonry 1990, p. 144. See also Hillsman 2002, p. 3. Hillsman 2002, p. 3. 41 Reagan n.d. Nozick 1974, p. 169; Machan 2000; Wollstein 2005. 43 DeLay 1998.

Time and money

15

days into the year to pay tax bill

350 300 250 5 July

200

29 July

150 17 April 100 50 0

US

France

Sweden

Figure 1.1 A temporal measure of tax impact: Tax Freedom Day 2005 Source: Butler 2005.

him- or herself.44 That date comes earlier or later in the year, depending purely on how high taxes are. Over-time comparisons suggest that Americans were free of their tax obligations by 22 January in 1900, but not until 31 March in 1950 and 3 May in 2000. Across the American states in 2005, people spent 32 more days working for the government in the highest-taxing state (Connecticut) than in the lowest-taxing one (Alaska).45 Cross-national comparisons are even more dramatic, as Figure 1.1 reveals. How much is too much of the year to spend ‘working for the government’? In helping us try to think through that question, these right-wing organizations offer a further set of time-based comparisons. They calculate in similar fashion the number of days people spend working to pay other major expenses each year, as compared to the number of days they spend working to pay their tax bill. In the US in 2005, the average American spent 107 days working for the government, which was almost as much as ‘housing and household operation’ (65 days) and ‘food and clothing’ (44 days) combined.46 And that does sound awfully high.

44 45 46

Tax Foundation 2005; Adam Smith Institute 2005. Tax Foundation 2005, based on Dubay et al. 2005, pp. 2–4, 11. Tax Foundation 2005, based on Dubay et al. 2005, p. 7.

16

Introduction

Of course, the ‘Tax Freedom Day’ methodology constitutes a very partial (as well as, obviously, a very partisan) measure of tax burdens. It takes account of the taxes people on average pay; but it takes no account of the benefits that people receive from government, or the time it would take them to earn wages equivalent to those public transfers that they receive. Tax Freedom Day is calculated as if all the money collected by the tax office just vanished into thin air. In chapter 2, we suggest procedures for calculating the temporal impact of state taxing-and-spending in a more even-handed way. The point to notice for now is simply that it is familiar to think of taxes in temporal terms, thanks to this rhetorical trope of anti-tax right-wingers.

1.2.4 Poverty, time and money Societies are commonly, and rightly, judged by how they treat their most vulnerable members.47 The proportion of people living in poverty is one important measure of that. Those proportions vary surprisingly greatly across countries. In the last decade of the twentieth century, just under 10% of people in the OECD lived in poverty, on average, and poverty rates in Australia and Germany, for example were very close to that. But in Sweden that figure was under 4%, whereas in the US it was over 16%.48 We know, however, that those poverty statistics do not tell the whole story. One important fact that they elide is how many hours (indeed, how many jobs) some people have to work to avoid poverty. We know that average hours at work also vary substantially across countries.49 We also know stories of how those on the cusp of poverty are ‘nickle and dimed to death’, stories of what long hours they often work at what low wages to keep their heads above water.50 If a country keeps people out of poverty only by making them work inhumanely long hours, that too would be a pretty poor performance. We can get some sense of the magnitude of that problem through the indicator of ‘time-or-money poverty’ alluded to in our 47 49

50

Goodin 1985. 48 OECD 2005b, Data Chart EQ1.1. Increasingly so, according to the conventional wisdom, as the US lags behind the rest of the OECD in working-hour reduction (Alessina et al. 2005; see further Jacobs and Gerson 2004). Ehrenreich 2002. As the joke circulating in the middle of the Clinton boom has it: ‘President Clinton has created millions of new jobs’; ‘I know, I have three of them.’

Time and money

17

51.32

60

0

US

Australia

money poverty rate

Germany France

5.41 9.02

6.42 10.85

34.37 10.54

17.67 all countries

7.99

10

10.13

20

12.74

36.77

30.06

per cent

40 30

38.03

50

Sweden Finland

time-or-money poverty rate

Figure 1.2 Poverty rates, money and time Source: LIS

preface.51 The ‘money poverty’ part of that measure is just the ordinary ‘poverty rate’: the proportion of people whose income is below the poverty line as it is standardly operationalized.52 As an indicator of ‘time poverty’ – anyway in respect of having ‘too little time left over after time in paid labour’ – we can borrow the international convention setting the maximum working hours at 40 hours a week.53 Anyone who works longer than that in paid labour will be said to be experiencing ‘time poverty’, for the purposes of this preliminary exercise. (Time in paid labour is, of course, only part of what we will be considering in subsequent chapters.) The ‘time-or-money poverty rate’ is simply the proportion of the population that is either living on a lessthan-poverty-level income or is working more than 40 hours a week in paid labour. Recalculating on that basis reveals how big a difference it can make to take time into account in our thinking about poverty. Figure 1.2 51

52

53

Goodin et al. 1999, pp. 225–35, 311–13; ‘time-money poverty’ is the complement of ‘combined resource autonomy’ reported there. As having less than half the median equivalent income of people in the country concerned; for further details see sec. 2.2.2 below. The ‘Convention Concerning the Reduction of Hours of Work to Forty a Week’, concluded in 1935, came into force in 1957 (ILO 1996).

18

Introduction

displays the relevant statistics for the countries that figure in our study, for the years to which our study pertains.54 In all of the six countries under study, the combined ‘time-or-money poverty rate’ is higher than that for money alone. It is only a little higher in the two Scandinavian countries. It is substantially higher in all the other countries. The combined ‘time-or-money poverty rate’ is more than double the poverty rate for money alone in every nonScandinavian country, and almost five times the ‘money poverty’ rate alone in a couple of them. Of course, to know what exactly to make of these findings we need to know more about the circumstances of the people who are working those extra-long hours. It would be one thing if they were mostly nearpoor and need to work those long hours to avoid poverty. In that case, these findings would constitute a serious indictment of the way a country treats its most vulnerable members, forcing them to work longer than international conventions dictate they should just to keep their heads above water financially. It would be quite another matter if those working long hours turn out to have comfortable incomes. They may or may not have a completely free hand in choosing exactly how many hours they worked, of course (we will return to this and related issues in subsequent chapters); but at least they would not be putting in long hours at the workplace merely to escape poverty. There are various ways in which we might further refine our timeor-poverty measures to try to address those issues.55 But another pair of thoughts encourages us to move in other directions. One is that, important though it is how societies treat their worst-off members, we ought not to judge societies purely on that basis. We ought to prefer an indicator that allows us to take some account of what life is like for the average person in any given country. A second, related thought is that it would be desirable to have an indicator not purely of how bad but

54

55

The ‘time-or-money poverty rates’ reported here are similar to those previously reported for an earlier period in the US (35.3 in 1990), but they are at substantial variance from those reported for Germany for a similar period (11.3 for 1992); the explanation for that difference is that that previous report included only persons in the former West Germany, whereas this report includes persons across all of unified Germany (Goodin et al. 1999, pp. 312, 107). By, for example, calculating what proportion of people in ‘time poverty’ have ‘near-poverty’ incomes, conventionally defined as having an equivalent income between 60 and 120 per cent of the poverty line.

Time and money

19

rather of how good life is for people – how much control they have over their time, rather than how little, on the measure we here employ – in various places. Important though it is to get everyone above the poverty threshold, it also matters how far above it they are (in our terms, how much ‘discretionary time’ they enjoy). Like the time-or-money measure of poverty, our ‘discretionary time’ incorporates temporal aspects of well-being. But whereas the time-or-money poverty indicator is just a matter of what proportion of the population get over a pair of poverty thresholds, ‘discretionary time’ is a scalar measure. The more discretionary time you have, the greater your ‘temporal autonomy’. Furthermore, our ‘discretionary time’ measure does so in a way that is sensitive to the situation of people on average in a society and what it is strictly necessary for him or her to do (and in unpaid household labour and personal care, as well as spending enough time in paid labour to earn at least a poverty-level income).

1.3 Two surveys, six countries Since our ‘discretionary time’ indicator combines considerations of both time and money, its empirical operationalization requires data of a temporal as well as a monetary sort. Unfortunately, these two are rarely found in sufficient detail in the same survey. Income surveys rarely include any time-use questions.56 And the information contained in time-use surveys on income is often highly rudimentary at best, in ways that will be explained below. Thus, in our empirical work (except in the case of Finland, more of which below) we have been forced to merge information from two different kinds of surveys, one of income and the other of time use. We rely upon surveys in the Multinational Time Use Study (MTUS) in constructing our time-use variables, and upon the Luxembourg Income Study (LIS) in constructing our income variables. LIS and MTUS are both major international projects devoted to gathering, harmonizing and standardizing income and time-use surveys 56

A conspicuous exception is the German Socio-Economic Panel, upon which the analysis reported in sec. 2.3.2 below is based. But even that was essentially an income survey, with a time-use module far too rudimentary to conduct full analyses of the sort reported elsewhere in this book.

20

Introduction

respectively across a wide range of countries and years. Thanks to the heroic efforts of those teams over many years, we can undertake these comparisons in some substantial confidence that the data upon which we are relying are as genuinely cross-nationally comparable as they can be made to be.57 The Multinational Time Use Study was initiated by Alexander Szalai in 1965 and is now based at the Centre for Time Use Research at Oxford University under the leadership of Jonathan Gershuny.58 This archive holds time-use surveys from twenty-four countries conducted over forty years.59 The MTUS team has ‘harmonized’ those surveys to render them as cross-nationally comparable as possible given the original research frames. The most recent data at the time of writing are contained in their ‘World 5.5.2’ files; our analyses are based on material in that file, and in the earlier ‘World 5.5.1’ files. The Luxembourg Income Study is an ongoing project, begun in 1983 under the joint sponsorship of the government of the Grand Duchy of Luxembourg and the Centre for Population, Poverty and Policy Studies. It aims to collect cross-nationally comparable data on those topics and related ones, in approximately five-yearly ‘waves’. Thirty countries are now represented in the LIS files.60 We employ LIS data from the year nearest that of the MTUS time-use survey, for each country under study. Basically, our procedure is to derive certain ‘necessary time’ variables from the MTUS files (in ways described in section 2.2 below) and to merge those variables onto the LIS files.61 Those amended LIS files then become the focus of all our further analyses.

57

58 59

60

61

The harmonization in LIS is perhaps more thoroughgoing than in MTUS. As explained in chapter 2, we rely on LIS more heavily than MTUS in our own analyses: we extract a few key variables from MTUS files, write them onto LIS files and then conduct the rest of our analyses on the LIS files thus supplemented. See Szalai et al. (1972) and Gershuny (2000, pp. 270–88) for fuller descriptions. Details of countries covered and variables included for each country in each year can be found on their website www.timeuse.org/mtus/. Details of countries covered and variables included for each country in each year are found on their website www.lisproject.org. Specifically, we wrote onto each individual’s file in LIS our estimates of ‘necessary time in personal care’ and ‘necessary time in unpaid household labour’ for a person with the LIS individual’s characteristics. We derived ‘necessary time in paid labour’ from variables in LIS files, adding estimates of ‘necessary time in travel to work’ derived from MTUS to that.

Time and money

21

Income surveys are relatively familiar. They ask people how much they earn, how long hours they work – and the same of all other adults in their households. They also ask how much the household as a whole receives as income from capital investments and how much the household as a whole pays in taxes and receives in transfers from the state. Time-use surveys are, perhaps, less familiar. The precise survey methodologies vary somewhat.62 But typically, these are diary-based exercises involving 5,000 or more people, often with everyone in the household aged 15 years and older filling in a diary, typically for two days. Care is taken to ensure equal representation of each day of the week and each quarter of the year (with weights to correct for any unrepresentativeness63). In their diaries, respondents are asked to record what they were doing, indicating beginning and ending points of each activity within 5 minutes.64 If they were doing more than one thing at the same time, respondents are asked to specify their ‘main activity’ and ‘what else they were doing’; here we confine our attention to ‘main activity’ codes alone.65 Respondents are typically asked to describe their activities in their own words. Those are then coded by 62

63

64

65

But less than might be imagined, thanks to the pioneering work of early time-use researchers (Szalai et al. 1972; Harvey et al. 1984); and further efforts at standardization and harmonization are underway, under the influence of Eurostat (2000a, b) and the UN (2005). For a survey of alternative methods, see Gershuny (2000, pp. 249–69). Even such apparently dramatic differences as between diary and recall methodologies seem to make less difference than might be imagined to overall results (Robinson 1985), although there is evidence that they now diverge more than they once did particularly in respect of certain items such as overtime in paid labour (Niemi 1993; Robinson and Bostrom 1994). Those weights typically pertain to the diary days in each survey. When analysing larger aggregates (ranging from the person up to the household) we take the harmonic mean of the weights for the units of analysis being aggregated, as recommended by ABS (1993, p. 36). The ‘harmonic mean’ is the reciprocal of the arithmetic mean of the reciprocals of a given set of positive numbers: thus, for example, the harmonic mean of 2, 3 and 4 is 12/13. For sample pages from the Australian survey, see ABS (1993) and Stinson (1999). On the ‘diary’ versus ‘recall’ methodologies, see Robinson (1985). Lest it be thought that keeping a diary itself takes time, and does so disproportionately from some activities rather than others, evidence from ‘buzzer’ cross-checks seems to suggest that there is no systematic bias introduced in the proportions of people’s days spent on various activities as reported in diary-based exercises (Robinson and Godbey 1997, pp. 61–7). As is traditional in time-use analyses – although analysis of secondary activities has been used to show the ‘intensity’ of time use, particularly by mothers of young children (Gershuny and Sullivan 1998, pp. 75–9; Bittman and Wajcman 2000).

22

Introduction

researchers into various major categories, which we further collapse into four broad categories: ‘time in paid labour’, ‘time in unpaid household labour’, ‘time in personal care’ and ‘spare time’.66 Respondents are also typically interviewed to obtain background information on their age and sex, employment, education, income and so on. While there are a great many surveys on file with both MTUS and LIS, the data requirements imposed by our operationalization of ‘discretionary time’ (section 2.2 below) prove to be surprisingly demanding. Although there are plans to enhance them in future, at the time of writing the income data in MTUS files are uniformly inadequate for our purposes.67 That is why we are forced to merge MTUS and LIS data. And even LIS data surprisingly often lack one or more of the elements required for our calculations.68 Although there are a great many more income surveys in LIS and time-use surveys in MTUS, only one survey from each of six countries both provides all the variables needed for our analyses and corresponds to a near-contemporaneous survey in the other data set.69 Table 1.1 lists the countries and surveys concerned, together with the years to which the data pertain, which are mostly in the 1990s. 66 67

68

69

See Appendix 1 for details. Earlier versions of the MTUS files (as in Gershuny 2000, appendix 2) report incomes only in very large (quartile) brackets. Respondents’ actual incomes in actual national currency units are reported only beginning with the World 5.5.2 version (and even there, for some but not all countries). Even in those later MTUS files there is no information on taxes paid; they report ‘total’ (i.e. gross) income from employment and self-employment. Transfer income from the state (‘income from benefits or other public assistance’) is scheduled to become available in the MTUS World 6.0.1 file, not yet released at the time of writing (and again, once released it will contain such data for some but not all countries). Finally, it is sometimes hard to estimate a wage rate from the information in MTUS for the same reasons as with some of the LIS files, as discussed in the next footnote. To calculate ‘necessary time in paid labour’ in ways described in sec. 2.2.1 below, we need to know the wage rate of each employed adult in the household. Typically LIS files report respondents’ annual gross salary or wages. To derive an hourly wage rate from that, we need to know various other things: the number of hours a respondent worked per week; and the number of weeks per year s/he was in full-time employment, the number in part-time employment and the number unemployed. Several countries’ LIS files were missing one or another of these variables, thus rendering them unusable for our purposes. In the cases of Sweden and the US, the MTUS data do not contain quite all we need. Those are such crucial cases, as exemplars of their respective types of welfare regimes, that we employed the procedures described in sec. 2.2.3 to make good that lack.

Time and money

23

Table 1.1. Income and time-use surveys used Years to which data relate Source Australia 1990 Survey of Income and Housing Costs and Amenities Time Use Survey Australia 1992

1989

LIS

1992

MTUS

Finland Income tax records merged onto Finnish 1999 Time Use Survey 1999/2000 Finnish Time Use Survey 1999/2000 1999 France Household Budget Survey 1995 Time Use Survey

Statistics Finland Statistics Finland

1994 1998/9

LIS MTUS

1994 1991/2

LIS MTUS

Sweden Income Distribution Survey Time Use Survey 1990/1

1992 1990/1

LIS MTUS

United States Current Population Survey American Time Use Survey 2003

2000 2003

LIS MTUS

Germany German Socio-Economic Panel 1991/2 Time Budget Survey of the Federal Republic of Germany

Source: LIS and MTUS.

As indicated in Table 1.1, in the case of Finland we use neither LIS nor MTUS data. Instead, Statistics Finland has provided us with a version of the 1999–2000 Finnish Time Use Survey, upon which the MTUS files are based.70 (We recoded time-use variables in that file following MTUS conventions.) Using the national ‘person number’, Statistics Finland has merged onto that file information from individuals’ 1999

70

For details, see Niemi and Pa¨a¨kko¨nen (2001) and Va¨isa¨nen (2002).

24

Introduction

tax returns concerning their income, taxes paid and public transfers received. For Finland uniquely, therefore, we have full income and time-use data on the same individuals, which provides an invaluable cross-check on our procedures for merging data from two separate surveys in other countries.

1.4 Welfare and gender regimes Our selection of countries to study is primarily dictated by data availability. Fortuitously, we have been able to secure two cases for each of the three principal types of welfare and gender regimes. Welfare-state researchers typically talk of the ‘three worlds of welfare capitalism’.71 Of course there are important variants within each of the three major clusters; of course there are important cases that do not fit within any of them; of course no country fits any of the ideal types perfectly.72 Still, the basic features of the main ‘three worlds’ are settled, fixed points for orienting comparative welfare-state studies. And while the corresponding gender and family regimes overlap those welfare regimes only imperfectly, there are some general patterns that do nonetheless stand out.73 The liberal regime, represented by the United States and Australia in our study, is a residualist welfare regime. The main mechanism for promoting people’s welfare, in a liberal regime, is the capitalist economic market. There, the state is classically relegated to a ‘nightwatchman’ role, safeguarding the conditions of free exchange and fair competition and correcting market failures. Poor relief in a liberal regime is traditionally a matter of charity, initially a religious duty that has now been assumed by the state. Categorical benefits, sometimes of a moderately generous sort, go to those (such as the old, the young, the disabled) who are excused from paid labour as a matter of public policy. Otherwise, however, liberal welfare benefits are traditionally targeted tightly on the poor and only the poor, and they are traditionally paid at a rate only barely adequate to alleviate the worst of 71

72 73

Esping-Andersen 1990. See similarly: Titmuss 1974; Castles 1998; Goodin et al. 1999; Goodin and Mitchell 2000, vol. II. Castles and Mitchell 1992. Compare, e.g.: Lewis 1992; Gauthier 1996; Sainsbury 1996; Esping-Andersen 1999, ch. 4; O’Connor, Orloff and Shaver 1999; Korpi 2000.

Time and money

25

their distress, for fear of creating disincentives to people engaging in paid labour. Liberalism’s approach to family and gender is similarly dominated by the stark individualism lying at that ideology’s core. Of course, liberal regimes might intervene with anti-discrimination legislation; but they do so purely to prevent anti-competitive practices in the labour market, in ways that are assiduously gender-blind. And while the poverty of lone mothers might be addressed as a matter of ‘poor relief’, liberals basically regard the family as falling decisively on the ‘private’ side of the public–private dichotomy, unfit as the subject for any substantial public intervention. The corporatist regime, represented by Germany and France in our study, is a ‘conservative’ welfare regime.74 Society is seen as a cooperative venture, with various groups (labour and capital, men and women, etc.) each having their distinctive role to play. The task of corporatist social policy is to underwrite social cohesion and social stability. Social benefits are typically earnings-related and hence status-preserving. Fiscally, corporatist regimes tend to engage in a substantial amount of ‘churning’, giving back to people in benefits roughly what they take from them in taxes.75 The family, and traditional gender roles within it, have historically been lynchpins of corporatist social thinking. The male’s role in a classically corporatist society is that of breadwinner for the family as a whole; the female’s role is that of homemaker. Marriage and childbearing are strongly encouraged. Lone motherhood, in particular, is strongly discouraged. The social-democratic regime, represented by Sweden and Finland in our study, is a highly egalitarian welfare regime.76 Characterized by class politics and socialist economics, social-democratic regimes strive towards social equality in a multitude of ways.77 One is through

74 76

77

van Kersbergen 1995. 75 Palda 1997. Finns prefer to label the regime ‘Scandinavian’ or ‘Nordic’, since the Social Democratic Party was not actually the one to introduce those arrangements there (Kangas and Palme 1996). Still, the policy style is unmistakably of the same sort (Esping-Andersen 1990, p. 74; Abrahamson 1999), and the ‘social democratic’ label is firmly entrenched in the cross-national literature. So, with apologies to Finnish sensitivities, we follow that established practice in so labelling this regime type. Korpi 1983.

26

Introduction

macroeconomic management, promoting high levels of employment and earnings. Another is through redistributive taxes and generous public benefits, characteristically of a universal, flat-rate sort. The egalitarianism of social democrats extends to their approach to gender relations as well. They strive to bring women into the paid labour force fully on a par with men, with public employment as one major mechanism. Partly in furtherance of that, social democrats typically provide a generous system of public care for children of preschool age. Social-democratic family policy is strongly oriented towards the interests of ‘the next generation’, and provides myriad forms of support to children and their carers out of a mix of egalitarian and pro-natalist concerns. Some features are commonly present in welfare states across all three regimes. Each, for its different reasons, strives to provide some kind of social safety net for those who are especially disadvantaged. This is seen by liberals as a question of ‘poor relief’, by corporatists as a means of promoting ‘social cohesion’ and by social democrats as an instrument of ‘social equality’. Whichever the rationale, redistributing towards the bottom is a common feature across all three welfare regimes. Another feature is a concern with freedom and autonomy – although once again the meanings vary. The freedom promoted by liberals is the ‘negative liberty’ of free markets: freedom from purposive intervention by particular others in one’s affairs. What liberals view as ‘freedom to choose’, socialists view as ‘freedom to lose’.78 Social democrats promote not ‘freedom from’ but rather ‘freedom to’, by providing people with the resources that would allow them to implement their choices. Corporatists see freedom in more Hegelian terms, where people are freed to realize their true nature as fundamentally social beings. We will explore the temporal consequences of alternative welfare and gender regimes in parts III and IV respectively. First, however, we must say more to justify and elaborate the measure of ‘discretionary time’ that will serve as our instrument of assessment in those later parts of the book. It is to this task that we now turn.

78

In the titles of books by Friedman and Friedman (1980) and Roemer (1988) respectively.

2

Discretionary time and temporal autonomy

This chapter is devoted to methodological preliminaries. In it, we argue philosophically why ‘autonomy’ in general matters, and why ‘temporal autonomy’ in particular matters as an aspect of that (section 2.1). We go on to explain technically how we measure ‘discretionary time’, which is our operational indicator of ‘temporal autonomy’ (section 2.2). We end by validating that empirical indicator sociologically, demonstrating that having ‘discretionary time’ of that sort is a significant influence on the extent to which people both subjectively ‘feel rushed’ and are ‘happy with their lives as a whole’ (section 2.3).

2.1 The value of temporal autonomy Our particular concern in this book is with ‘temporal autonomy’, and with ‘discretionary time’ as a manifestation of that. The reason that temporal autonomy matters is that autonomy itself matters. Thus it behooves us to step back briefly from temporal issues to consider that broader moral notion, before delving into the peculiarities of the present ‘temporal’ application.

2.1.1 The concept of autonomy Philosophers point to many kinds of freedom.1 ‘Autonomy’ refers to one’s capacity to form principles of one’s own and to act upon them. To be autonomous is ‘to decide for oneself what one shall do’.2 1 2

Berlin 1958; Raz 1986; Pettit 2001; Skinner 2001. Feinberg 1980, p. 19. Raz (1986, pp. 369, 370, 381) expands: ‘The ruling idea behind the ideal of personal autonomy is that people should make their own lives . . . The idea of personal autonomy is the vision of people controlling, to some degree, their own destiny, fashioning it through successive decisions throughout their lives . . . An autonomous agent’s well-being consists in the successful pursuit of self-chosen goals and relationships . . . [T]he conditions of autonomy include both opportunity and the ability to use it.’

27

28

Introduction

Philosophically, that is arguably the most important kind of freedom, being as it is a necessary condition for leading a moral life at all. Despite being a (perhaps ‘the’) pre-eminent moral good, autonomous control over one’s life is not the only moral good there is. Sometimes it can be maximized, or fully exercised, only at some substantial cost to other morally important goods. In such circumstances, you ought not to pursue autonomy utterly to the hilt but instead ought to sacrifice some measure of autonomy in order to secure significant gains in those other dimensions.3 Even where it is being traded off in that way for other morally important goals, autonomy nonetheless always remains a major moral consideration that weighs heavily in those calculations. Freedom and autonomy might of course be good from a philosophical point of view whether or not ordinary people actually value those things themselves. Evidence abounds, however, that having control over one’s life is not only desirable from an abstract philosophical point of view but is also actually desired by most people in the real world.4 Having a sense of control over your life makes you happier and more satisfied with life.5 It makes you less depressed;6 it even makes you physically healthier.7 This is true across a range of sites: people are more satisfied with their jobs the more control they have in them;8 elders are more satisfied with their nursing homes the more control they are allowed in them;9 punters are even more satisfied with lottery tickets if they get to choose their numbers.10 Conversely, people express dissatisfaction when they are not (or feel they are not) in control of their lives. ‘Disaffected democrats’ complain they are not in control of their government and that their government is not really in control of events either.11 ‘Overworked Americans’ 3 4

5 6 7 8 9

10

Dworkin 1982. Maybe you can have ‘too much control’ as well as ‘too little’; that is the burden of the ‘escape from freedom’ hypothesis (Fromm 1941). But the evidence glossed below suggests that, within the ordinary range within which ordinary lives are ordinarily lived, ‘more control’ is indeed subjectively desired by most people. Perlmuter and Monty 1977; Veenhoven 1984; 1999; Peterson 1999. Abramson et al. 1989. Marmot et al. 1991; Pulkkinen et al. 1998; Marmot 2004. Karasek 1979; Warr 1990; Van der Doef and Maes 1999. Langer and Rodin 1976; Lai and McDonald 1995; Duncan-Myers and Huebner 2000. Nichols et al. 1994. 11 Pharr and Putnam 2000.

Discretionary time and temporal autonomy

29

complain of being in a ‘time bind’, not only because they are busy but also because they are not in control of how busy they are.12 Philosophers writing on ‘autonomy’ tend to fixate on the first part of the formula we offered at the outset, the part about a ‘capacity to form principles of one’s own’. Philosophers are particularly preoccupied with the quality of the will. The will of the autonomous agent ‘gives laws to itself’, in Kant’s famous phrase; that of heteronomous agents merely takes orders from others.13 Giving law to yourself is undeniably important for the autonomous agent. But as with literal lawmakers so too with the metaphorical lawgivers of Kant’s Kingdom of Ends: it does no good to make a law if there is no way that that law can be implemented.14 That in turn underscores the importance, for the practical exercise of autonomous agency, of the second half of our formulation of autonomy – having not only the ‘capacity to form principles of one’s own’ but also ‘the capacity to act upon them’. That lesson was learnt painfully by emancipated slaves in the American South. Emancipation from chattel slavery was only ‘one kind of freedom’. Given their freedom but nothing else – no money, no land, no tools, no mules, no assets of any kind – they were forced into the position of tenant farmers, who had to borrow so heavily to plant their first crop that they were forever in debt peonage to the local store.15 As Raz says, ‘Autonomy cannot be achieved by a person whose every action and thought must be bent to the task of survival, a person who will die if ever he puts a foot wrong.’16 Of course, people would not count as autonomous if the principles upon which they act were not their own, no matter how great a 12

13 14

15

16

Or at least that is what the results in sec. 2.3.1 below suggest is true in Australia and Finland. Kant 1785; Dworkin 1988; Darwall 2006. Recall in this connection President Andrew Jackson’s riposte to a Supreme Court decision he disliked: ‘Well, [Chief Justice] John Marshall has made his decision, now let him enforce it’ (Warren 1926, vol. I, p. 759). Ransom and Sutch 1977, p. 198. See similarly Stinchcombe (1995) on how slavery was reimposed in some (but not other) islands of the Caribbean postemancipation. The day-fine system discussed in sec. 1.2.2 above is of course akin to debt peonage. Raz 1986, p. 379. Rawls (1971, sec. 32) distinguishes ‘liberty’ from the ‘worth of liberty’, but that distinction would seem to underscore our point here rather than rebutting it: why would anyone ever want to champion worthless liberty? (Goodin 1988, p. 308).

30

Introduction

capacity they might have to work ‘their’ will on the world. So the first part of our characterization is lexically prior to the second. You can only count as autonomous if you have sufficient capacity to form principles of your own.17 Once you are across that threshold, however, how much autonomy you have is then primarily a function of the second criterion: how much capacity you have to act upon them. People with a capacity to form principles of their own but no capacity to act upon them (like Stoics’ slaves18) might be said to be autonomous but only in a minimal way. People with both a capacity to form principles of their own and a great capacity to act upon them would be said to be highly autonomous. How autonomous any given person is, once over the threshold of having adequate capacity to form principles of his or her own, depends just on how much capacity he or she has to act on principles of his or her choosing. It is a variant on this scalar notion of autonomy that is represented in our notion of ‘discretionary time’ as a measure of temporal autonomy. People can have more or less discretionary time, and hence more or less temporal autonomy, precisely in this sense that they can have more or less capacity to act in ways of their choosing. One antonym of ‘autonomy’ is ‘necessity’. People (or processes, come to that) that depend upon and are controlled by another are not ‘autonomous’. Someone who acts out of necessity has no choice. That just follows from the definition of ‘necessity’. And that fact makes ‘necessity’ the antonym of ‘autonomy’, understood as we do as the capacity to choose principles of one’s own and act upon them.

2.1.2 Temporal autonomy ‘Temporal autonomy’ is a subspecies of that more general notion of ‘autonomy’. It is simply a matter of having control over how one chooses to use one’s own time. As with the notion of autonomy in general, so too with temporal autonomy: the capacity to make choices and act upon them (in the temporal case, over how to spend one’s time) lies at the heart of the notion. 17

18

For simplicity, the present discussion treats that as a threshold notion; the same points could however be made, more cumbersomely, treating it as a scalar notion. Nussbaum 1994.

Discretionary time and temporal autonomy

31

As in the case of autonomy in general, so too (all the more so) in the case of temporal autonomy: that is not the only moral good there is. Indeed, temporal autonomy is not the only morally important aspect of autonomy that there is. Here as before, you ought not necessarily to pursue temporal autonomy to the hilt, if in so doing you would end up sacrificing other things (perhaps even other aspects of autonomy more generally) that are equally or more morally important.19 As people who are long-term unemployed are painfully aware, having lots of free time is useless (or frustratingly worse) if you lack many of the other resources that are required to make good use of it.20 While you can thus have ‘too much’ free time as well as ‘too little’, across the broad range of cases (certainly across those with which we will be concerned in this book, which for methodological reasons specifically excludes households where everyone is unemployed, retired or otherwise not in paid labour21) more ‘temporal autonomy’ is always better, other things being equal.

19

20

21

Where the effective exercise of choice is costly in terms of time and attention that can be devoted to other pursuits, for example, it is not necessarily the case that ‘more choice is better than less’ (Dworkin 1982). Jahoda et al. 1933/1971. Whether what the unemployed have is ‘temporal autonomy’, strictly speaking, is an open question. In terms of the distinctions we will be making in our operationalizations below, they have lots of ‘spare time’ (time not in paid labour, among other things). But how much ‘discretionary time’ – our measure of ‘temporal autonomy’ – people have is a function of how long it takes them to secure an income at the poverty line. And if the unemployed have no way of securing an income at that level, then the logic of that analysis would by extension lead us to conclude either that they have no ‘discretionary time’ or else that the concept is meaningless as applied to those people. Our methodology for operationalizing ‘temporal autonomy’ by calculating ‘discretionary time’, described in sec. 2.2 below, requires us to know how long it takes people in the household to secure at least a poverty-line income. For people in paid labour, their wage rate tells us that. For people not in paid labour, we would need to find some other way of reckoning how time translates into money: for the unemployed, perhaps, how long they have to spend satisfying the requirements to qualify for unemployment benefits, and how much money they get in such benefits (and whether by spending more time they could somehow get higher benefits or more money, and how much, in other ways). Standard timeuse surveys do not allow us to differentiate people’s interactions with government in nearly so fine-grained ways as that. Notice this further fact about all people not in paid labour whose income stream (property income, or private transfers, or retirement pensions, or in principle even unemployment benefits) is above the poverty line without their having to spend any time at all to secure that outcome: such people’s ‘necessary time in securing income’ (the

32

Introduction

If autonomous moral agents are supposed to choose for themselves what they will be and do, they need – centrally among all the other things they might need – time to use in ways of their own choosing, in order so to do or become. Sovereignty over one’s time, or anyway some appreciable portion of it, is as central to leading an autonomous life as is sovereignty over one’s body.22 It may even be more so. Stoics advised slaves that even if one’s body were in chains, one’s mind could wander free; but that is true only if they can find the time free from their masters’ control to do so. Chattel slavery is not the only kind of slavery there is. Debt peonage is another, as emancipated slaves in the American South discovered to their cost. So too is ‘wage slavery’, as Marx dubbed the employment relationship.23 Beneath that provocative phrase lies a point that is familiar from even the most standard neoclassical accounts of that relationship. What you do in entering the employ of another is to ‘contract your time’ to him or her, agreeing to do (within certain limits) whatever he or she requires of you during that period.24 How much autonomous control you have over how much time to contract to an employer in that way is an open question: Marx thought very little, and that may well have been true in his day; in our analyses below, we find people have substantial autonomy in that matter. But the point on which we are agreed is that, insofar as you (largely) have to do what you are told when in paid labour, you lack (full) autonomy during any period you have during which it is strictly necessary for you to engage in paid labour. Marx’s contemporary John Stuart Mill described wives of that era as being ‘subjugated’ to their husbands.25 That too is dramatic language. It was an apt enough characterization of life in England before passage of the Married Women’s Property Act, when the legal personality of

22

23

24

generalization of our ‘necessary time in paid labour’ variable) is zero, and their ‘discretionary time’ as we calculate it below would be determined purely by the other two components of that calculation (‘necessary time in unpaid household labour’ and ‘necessary time in personal care’). We couch our discussion in terms of autonomy, but it could of course be recast in the converse terms of social power: how long you have to spend waiting on others, and how long you can make others wait on you, is an effective measure of your social power relative to them (Schwartz 1975). Marx 1875/1978, p. 535. The libertarian notion of ‘time working for the government’ (sec. 1.2.3 above) is related to this. ˚ s 1978; 1982; ABS 1998a, b. 25 Mill 1869. Simon 1951; A

Discretionary time and temporal autonomy

33

the wife was literally subsumed within that of her husband.26 But even in a liberated era when wives are no longer expected to be ‘obedient’ to their husbands, there is some time in the household that they cannot truly call ‘their own’. There are certain household tasks that the household as a whole (nowadays, both the husband and wife, albeit in unequal measure) simply have no choice but to perform. Some of them are imposed by force of law. (Parents will be prosecuted for ‘child neglect’ if they do not meet some minimal standards in taking care of their young children.) Others relate to social expectations enforced by serious social pressure of a less formal sort. (Keeping gardens minimally tidy or shared staircases minimally clean, for example.) But this time too is ‘committed’, rather than under one’s own control.27 Perhaps only a Stoic or someone who is severely disabled would describe oneself as a ‘slave to one’s own body’. But in an important sense of course we all are that, too. Its demands exercise unavoidable control over how we spend much of our time. We can choose when and where to eat and sleep; but we have no real choice whether or not to eat or sleep. Again, a certain amount of time must be spent bowing to bodily necessities, if we are to continue living at all. The sense in which something akin to ‘slavery’ is involved across all these realms (of paid labour, unpaid household labour and personal care) is precisely that ‘temporal autonomy’ is clearly lacking. Even if you work for an enlightened firm rather than a literal slave-driver, you have no real discretionary control over whether or not you spend however long it takes to get at least a poverty-level income.28 Even if your household is a model of modern egalitarian family values rather than a Victorian tyranny, you have no real discretionary control over whether or not you spend as long as it takes to keep your household 26 27

28

Pateman 1988b. In the standard time-use terminology which (while unfortunate in one crucial ˚ s 1978; way we discuss in a footnote below) is apt on this particular point. See A 1982; ABS 1998a, b. In this book we will be focusing on time spent in paid labour, but there are obviously other ways to secure an income than that. Our point here would simply be that all of them – panhandling, standing in the dole queue, keeping your sugar daddy satisfied – all take time. Unfortunately, due to the limitations of the data on which we will here be relying, we simply cannot surmise how much time they take; hence our focus upon paid labour as a way of securing income.

34

Introduction

functioning to some socially acceptable standard.29 Even if your body is fit rather than demandingly disabled, you have no real discretionary control over whether or not you spend as long as it takes to keep it functioning similarly.

2.2 Operationalizing discretionary time As with any complex piece of empirical research, there are a great many fine-grained decisions concerning coding and methodology embedded in our study. Appendix 1 describes our procedures in much more detail – hopefully sufficient detail to enable replication of our results from data sets publicly available through MTUS and LIS.30 In the text that follows, we offer a much broader overview of our method for operationalizing the notion of ‘discretionary time’. Readers who are not interested in even that much detail can get the basic idea by reading this introduction to section 2.2, and then skipping the more detailed discussions of components of ‘discretionary time’ in subsequent subsections. ‘Discretionary time’, as we conceive it, is simply the time over which you have autonomous control. A crucial antonym of ‘autonomy’, as we have said, is ‘necessity’. So the time over which you have autonomous control will here be said to be time the use of which is not dictated by the ‘necessities of life’. Here we identify three realms of ‘necessity’, when it comes to time use. One is economic: at least in market societies of the sort we are here discussing, everyone needs to access enough income (usually through paid labour) to meet the minimum requirements embodied in the ‘poverty line’. A second is social: everyone needs to spend enough time in ‘unpaid household labour’ (shopping and cooking, cleaning the house, minding the children if one has any, etc.) to keep one’s

29

30

How long that is depends, obviously, on who else might be available to tend to the household chores for you. For understandable reasons, Statistics Finland is reluctant to allow unrestricted use of the data set onto which they have merged information from people’s tax files. Similar results to those reported here for Finland can be reproduced applying our methodology to the Finnish files in MTUS and LIS, although in calculating a ‘wage rate’ you need to make certain assumptions about ‘hours worked’ since that information is absent from LIS files for Finland for that period.

Discretionary time and temporal autonomy

35

household ticking over.31 A third is biological: everyone needs to spend a certain amount of time in ‘personal care’ (eating, sleeping, bathing, etc.) to keep one’s physical body ticking over.32 ‘Discretionary time’ as we operationalize it is simply the amount of time you have left over, once you have done what is strictly necessary in each of those three realms. A few people have ‘special needs’, biologically. But by and large, the time every adult strictly needs (as opposed to takes!) to eat, sleep, bathe and so on is about the same. How much time people need to spend tending the house varies wildly, depending principally on their household structure: having children (particularly children under 5 years old) increases that dramatically, and the more the more children there are. And how much time people need to spend in paid labour to secure at least a poverty-level income also varies wildly, depending on their wage rate. Our measures take those latter two sources of variation systematically into account, by indexing each individual’s ‘necessary time in paid labour’ to his or her wage rate and each individual’s ‘necessary time in unpaid household labour’ to his or her household structure.33 But we attribute the same ‘necessary time in personal care’ to everyone in the country.

31

32

33

Note that our measure of ‘unpaid household labour’ is ‘broad’, in the sense that it includes ‘child care, shopping and odd jobs’ as well as what some time-use researchers call the ‘core domestic work’ of cooking and cleaning (Gershuny and Robinson 1988; Gershuny 2000, ch. 5; Gershuny and Sullivan 2003; Sullivan and Gershuny 2001). Those may be interesting subsets of unpaid household labour to examine for certain purposes. But for any broader purposes, defining unpaid household labour in such a way as to exclude child care is just plain daft. According to a terminological convention common among time-use researchers (Robinson and Converse 1972; A˚s 1978; 1982; ABS 1998a, b), the label ‘necessary time’ is applied to personal care alone. Time in paid labour is labelled ‘contracted’ and time in unpaid household labour ‘committed’. That terminology is seriously misleading, however. Calling only the first ‘necessary’ implies that it is not necessary to spend any time in the other two sorts of activities, which is obviously absurd. Both are also indexed to the country in which one lives. Note that ‘time in paid labour’ and ‘time in unpaid household labour’ are fungible: you can either spend time cooking and cleaning and child-minding for yourself, or you can spend time earning money to pay someone else to do these things for you. Hence any given person can compensate for spending less time than we deem necessary in one of those realms by spending more than necessary in the other. Our calculations of ‘discretionary time’ take into account both these factors at once, however, and thus incorporate that trade-off.

36

Introduction

People almost invariably prefer to do more than is strictly necessary. People can meet their minimal needs with a poverty-line income, but few people would choose to live like that if they could avoid it. Similarly, people generally prefer more sleep than the minimum that is strictly necessary, biologically; and they prefer tidier homes than the bare minimum will be tolerated, socially. What we here will call ‘spare time’ is what time-use researchers typically – but misleadingly34 – call ‘free time’. ‘Spare time’ is defined as the amount of time you have left over after all the time you actually devote to paid labour, unpaid household labour and personal care. Doing the minimum necessary in all those realms amounts to bowing to necessity. Socially, you have no real choice in those matters. Doing more than the minimum necessary in those realms, however, is indeed a matter of choice. The extra time you spend on such activities, above and beyond what is strictly necessary, is an exercise of your autonomy, whereas doing what is strictly necessary is a constraint on your autonomy. Because people generally spend more time than strictly necessary in each of these realms, ‘spare time’ is generally less (typically much less; often very much less) than ‘discretionary time’. The gap between the two varies dramatically across household types and somewhat across countries. This gap constitutes the ‘time-pressure illusion’ that will be discussed in part II of this book.

2.2.1 Necessary time: some general methodological considerations That is a thumbnail sketch of the basic ideas underlying the methodology of this book. In sections 2.2.2 through 2.2.5 below, we will explain more fully our approach to each of the key components of these calculations. Before doing so, however, let us make a few general methodological remarks. The first concerns our general strategy in specifying ‘necessity’ across all these realms. ‘Necessity’, by definition, refers to bare social essentials. 34

True, this is time ‘free of’ paid labour, unpaid household labour and personal care. But insofar as people spend more time than strictly necessary on these activities, they are also ‘free to’ engage in them or not (i.e., not compelled by necessity to do so), for many of the hours they actually spend in them.

Discretionary time and temporal autonomy

37

Estimates of what is necessary should therefore be minimalist. Accordingly, wherever we are forced to choose among otherwise equally credible ways of specifying what is ‘necessary’, we always opt for the method yielding the lowest estimate of ‘necessity’. One way of assessing the plausibility of our estimates of what is necessary in each of these realms is to check what proportion of the population does ‘less than necessary’, according to the standards we propose. It would be an odd notion of ‘necessity’, indeed, if no one in the country managed to do what was deemed necessary. It would be almost as odd to define ‘necessity’ in such a way that no one in the country failed it. Setting the poverty line equal to the lowest amount of income enjoyed by anyone in the country would be defining poverty out of existence, purely by fiat; however convenient it might be for political leaders to be able to claim to have thus eradicated poverty, that is just not a credible representation of social reality. In setting a standard of what is ‘necessary’, we should do so in such a way that a modest but nonetheless non-negligible proportion of the population falls below it. Table 2.1 shows the proportion of the population falling below the standards of necessity elaborated in each of the next three sections, first averaging across all six countries and then for each of the six countries separately. The first column indicates the proportion of people living in households in poverty, as defined below (section 2.2.2). That represents the proportion of people living in households that are spending less time than we deem to be minimally necessary in paid labour. The second column indicates the proportion of people living in households where less time than we deem to be minimally necessary is being spent in unpaid household labour. The third column indicates the proportion of people who are spending less time than we deem to be minimally necessary in personal care.35 As we see in Table 2.1, the proportions of people in all countries doing less than we deem minimally necessary are, indeed, in the ‘modest but non-negligible’ range – just as we would desire for a plausible estimate of what is ‘minimally necessary’. Averaging across all six countries, the proportion of the population doing less than we deem 35

The first calculation is based on LIS data, and pertains to the whole population. The latter two calculations are based on MTUS data, and owing to limitations of those samples pertain to people aged 20 to 65 only.

38

Introduction

Table 2.1. Percentages of the population doing less than minimally necessary in three realms Paid labour

Unpaid household labour

Personal care

Average across all countries

10.13

9.90

9.92

US Australia Germany France Sweden Finland

17.67 12.74 7.99 10.54 6.42 5.41

9.65 9.19 9.83 12.88 4.89 12.93

12.04 8.38 8.79 11.06 8.27 10.94

‘minimally necessary’ is almost exactly 10 per cent in each of these dimensions. Looking at the statistics country-by-country, there is a little more variation than that. But more than three-quarters of values are in the range 10  3; and none of the values is more than twice or (much) less than half the six-country averages. All of that reassures us that our standards of necessity are plausible ones. A further methodological consideration to note at the outset is that our standards of what is necessary are all ‘relative’ rather than ‘absolute’ standards. That is to say, what is necessary for you to have or to do is deemed to be a function of what other people in your country have or do, in all three realms of ‘necessity’. Income poverty is conventionally defined in relative terms, as half the median equivalent income in one’s country (section 2.2.2 below).36 Our method of estimating ‘necessary time in paid labour’ is based on that way of measuring poverty, and our methods of estimating ‘necessary time in unpaid household labour’ and ‘necessary time in personal care’ are modelled on that. All those 36

That is not completely uncontroversial. But trying to estimate an ‘absolute’ standard of poverty by specifying a ‘basket of goods’ representing the ‘bare necessities’ that everyone must have is even more controversial, within any given country (cf. Deeming 2005) and all the more so across countries. Perhaps some people might be more tempted by a ‘basket of goods’ approach to certain aspects of personal care like sleep (Schor 1992, p. 11). But even there, ‘necessary time’ is not a physiological given that is invariant across time and place: both folk wisdom and learned disquisitions have varied wildly over the past several centuries on precisely that score (Ekirch 2001; 2005).

Discretionary time and temporal autonomy

39

measures of ‘necessary time’ are therefore relative to how much time others in one’s own country are spending in those activities.37 We think this is as it should be: bathing and changing your shirt only once a week was socially acceptable in the days before most people had indoor plumbing and washing machines, but ceased being so thereafter.38 What it is necessary to do, and how much time is necessary to do it, is socially relative in that way. Another set of general methodological remarks concerns the different subsets of the samples employed at various stages in our analyses. Following standard practice among poverty researchers, we base calculations of necessary time in all three realms (paid labour, unpaid household labour and personal care) on the entire sample, i.e., everyone in the survey.39 Thereafter, we confine our subsequent analyses to a restricted subset of that sample: households comprised of one adult or two married (or partnered) adults and their children (if any), where the adult(s) are of prime working age (25 to 54), and with at least one adult in paid labour. All calculations of ‘discretionary time’ and ‘spare time’ are based on that more restricted subsample. That is only partly because the requirements of our methodology require us to restrict our attention to that particular subsample.40 The larger point is, however, a substantive one: our primary concern in this book is with work–life trade-offs of the sort experienced primarily by prime-aged earners, so it is right from a substantive as well as methodological point 37

38 39

40

In principle those could be relativized more specifically to one’s reference group defined in terms of age, social class, geographical location (urban/rural) or whatever. The precedent of poverty calculations suggests relativizing to the country as a whole instead. Schor 1992, p. 89; Strasser 1982. In the income surveys in LIS, there are no age restrictions of any consequence. In the time-use surveys in MTUS there typically are, and they vary a little from survey to survey; people aged between 20 and 65 years of age is the most extensive group of people represented in all six MTUS surveys we will be using here, so that constitutes the ‘entire sample’ for calculations based on time-use data. At certain steps in the necessary time calculations we use more restricted samples. For example, in calculating ‘necessary travel time to work’, for the number of days we reckon someone has to go to work, we looked only at people reporting some hours in paid labour. We need a wage rate to calculate ‘necessary time in paid labour’, so we need for at least one of the adults actually to be in paid labour and hence have a wage rate. We omit households with ‘extra’ (i.e., more than two) adults, because it is uncertain how to apportion ‘necessary time in unpaid household labour’ among adults in such households (a lodger is one thing, a live-in granny quite another).

40

Introduction

of view that we confine our attention to that subset of the population for present purposes.41 Finally, the results reported in the following chapters are calibrated in hours per week. The number given for any group (whether whole countries or particular household types within them) is, following standard practice among time-use researchers, the mean across all people in that group.42 In the complete results contained in Appendix 2, country-specific means are presented for each of the following household types:

     

single adult, no children dual-earner couple, no children one-earner couple, no children single adult with children dual-earner couple with children one-earner couple with children

The most interesting results typically involve some aggregation of household types (and, indeed, countries), weighting appropriately. In subsequent chapters, our results will most typically be presented in those more aggregated forms. Full breakdowns of all of our results, country-by-country and household-type-by-household-type, can be found in Appendix 2.

2.2.2 Necessary time in paid labour Discretionary time, as we here operationalize it, is the residual that remains after deducting necessary time in three realms: paid labour; unpaid household labour; and personal care. Strictly speaking, what is necessary is merely securing income sufficient to escape poverty. 41

42

Of course today’s pensioners and unemployed typically were formerly employed, and their current state benefits are in part a function of contributions to the pension scheme they made during their working lives. So, conceptually, we can think of those state benefits as amounting to a (typically, compulsory) sacrifice of some discretionary time for more later. But those across-time transfers are beyond the scope of the data sources available to us for this study. The median might arguably be a better representation of what life is like for the average individual in a country, as poverty researchers have long recognized. But the convention among time-use researchers is otherwise. In any case it is convenient for certain analytical purposes later in the book to have a measure that is additively decomposable, which the mean is and the median is not.

Discretionary time and temporal autonomy

41

Working in paid labour is not the only way of doing that. Still, it is the principal one in advanced industrial societies of the sort we are here discussing. In any case, it is the only one we can examine, given the data we shall be using. In analysing ‘necessary time in paid labour’, we can piggyback on a large and well-established literature concerning how much income is minimally necessary – the poverty line. That has been the subject of long-standing discussions, straddling both the academic and policy communities. There are of course still hold-outs academically as well as politically.43 Still, we have now come as close to consensus as will ever be found on any matter of social importance. In specifying our measure of ‘necessary time in paid labour’, we simply adopt and adapt that broadly agreed method of measuring poverty.44 When in subsequent sections we specify measures of ‘necessary time in unpaid household labour’ and ‘necessary time in personal care’, we attempt to mimic that method as closely as is practicable. How much time it is necessary for you to spend in paid labour depends upon three things: how much income you need; how much of that you get from other (non-labour) sources; and what your hourly wage rate is. The strictly minimum amount of total income you need is defined by the poverty line. Following our methodological precept of setting standards of ‘necessity’ as low as they can plausibly be set, we adhere to the older and more conservative European Union methodology of setting the poverty line at ‘50 per cent of the median equivalent income 43

44

The US is conspicuous in this regard: it has an official ‘poverty line’ that is set in a different and much disputed way, for curious historical reasons (Orshansky 1969; Ruggles 1990; Citro and Michael 1995). Wherever there are means-tested social benefits, as there conspicuously are in Australia, the means test in effect specifies a poverty line; note, however, that those means tests often vary from programme to programme, and are typically set on some basis other than that described in the text below. In the other countries under study, there is either no officially recognized poverty line or it is specified in broadly the way we describe below (although as noted there, post-Lisbon in 2001 the EU poverty line was set at 60 rather than 50 per cent of median equivalent income). The most accessible version is Atkinson (1998); see further Atkinson et al. (1995). At least that methodology is broadly agreed for the sorts of developed countries under discussion here. Something different may well make more sense for developing countries (Deaton 1997, ch. 3). Sen’s (1999) ‘capabilities’ standard, adopted by the World Bank (2005) and World Health Organization (2005), is one such attempt.

42

Introduction

across all people in the country’.45 A person’s ‘equivalent income’ is that person’s actual income, adjusted for the size of that person’s household so as to take account of the effects of economies of scale; the ‘equivalence scale’ here used is the square root of the number of people in the household.46 To calculate an individual’s equivalent income we divide his or her household’s total income by that equivalence scale, and assign equal ‘equivalent income’ to every member of the household. As is conventional, we calculate the poverty line on the basis of ‘net disposable income’ (what in the terminology of part III might be thought of as ‘post-government income’). That is usually defined as income net of taxes and government transfer payments. In addition to direct cash payments, governments also provide non-cash subsidies that can substantially affect people’s real net disposable incomes. Those subsidies should ideally be taken into account in ‘net disposable income’; in practice, however, it is difficult to do so systematically given data constraints, so those factors are typically just left out of the calculations.47 Here we deviate from that practice by taking into account government subsidies of child-care costs, in ways explained below. To achieve a poverty-level income, a household needs enough total income to give each of its members an equivalent income equal to the poverty line. How much total income the household as a whole needs is therefore the poverty line multiplied by the equivalence scale described above. That income can of course come from any of several sources: not just as income from paid labour but also as income from property ownership, occupational pensions, government benefits, child support and alimony, or gifts from family or friends or charities. How much income the household needs to raise via paid labour equals the household’s total income need, less the income it receives from those other sources. Some of those other income streams (property income, occupational 45

46

47

Latterly, the EU has shifted from talking about ‘poverty’ to ‘low income’, defined as 60 per cent of median national equivalized income (EC Social Protection Committee 2001). Various different ‘equivalence scales’ have been proposed over the years, but the differences between them turn out not to make much difference to overall national poverty rates, although they can make a difference to the composition of the poor within countries (Atkinson et al. 1995). Smeeding et al. 1993; Atkinson 1998, p. 31.

Discretionary time and temporal autonomy

43

pensions) are generally insensitive to changes in people’s other income. Others (government benefits, gifts, child support and alimony) may well depend on how much other income people receive. And, obviously, how much money people pay in taxes is also highly sensitive to changes in their income. In calculating ‘necessary time in paid labour’, we are asking in effect how much paid labour a person would have to do if he or she earned just a poverty-level income. (The best way to think about this is as people reducing their work hours one-household-at-a-time, rather than all-at-once.48) That is of course less – typically much less – than most people actually earn at present. Hence, while we can just take income from income-insensitive streams as given in the household’s actual income records, an adjustment must be made to income from sources that seem likely to be affected by people’s other income.49 To calculate the amount of money a household needs to earn in paid labour, we deduct that amount of money (together with the amount of 48

49

If all households did that simultaneously, all sorts of things would happen: the economy would collapse, with consequences for wage rates and hence ‘necessary time in paid labour’; government revenues would fall precipitously and expenditures would soar, forcing major changes in the government programmes under study; and so on. Thinking of households reducing their time in paid labour one-at-a-time, with all else held constant, avoids those complications that would be virtually impossible for us to model here. As our estimate of how much income a household would have at the poverty line from those other income-sensitive sources, we take the mean amount of each of those kinds of income for households whose actual incomes are around the poverty line (and who are included in the more restricted subsample, described above, on which our calculations of ‘discretionary time’ and ‘spare time’ are based). ‘Around the poverty line’ might more usually be defined as ‘having an equivalent income of 40 to 60 per cent of median equivalent income’; but we have to expand that range to ‘25 to 75 per cent’ to get enough cases, once we disaggregate household types that are likely to receive very different amounts of government benefits and gifts. In calculating that we distinguish: (1) households with no children; and, among those with one or more children, (2) ones with a single parent (who is also in paid labour), (3) two-earner parent couples and (4) one-earner parent couples. There are too few cases of childless households with incomes around the poverty line to allow further disaggregation of that category. Because there are too few cases of households around the poverty line receiving ‘child support and alimony’ (items not disaggregated in LIS files), we just take the mean of the amount received in child support and alimony by all households that are in receipt of such payments and who have equivalent incomes around the poverty line. In our ‘necessary time in paid labour’ calculations we attribute that amount to all (but, obviously, only) households who are actually in receipt of child support and alimony at present.

44

Introduction

income the household actually receives from income-insensitive sources) from the total amount of money the household needs to get to the poverty line. Similarly, we calculate taxes as the taxes that are actually paid by households whose incomes are around the poverty line, and we add that to the total amount of money the household needs to acquire to get to a poverty-level net disposable income. That represents the total amount of income from paid labour needed by the household as a whole. For households with a single earner, that will be the amount of money that that earner needs to raise through his or her paid labour. That earner’s ‘necessary time in paid labour’ is (modulo a couple of adjustments discussed below) simply that amount of paid-labour income the household needs, divided by that earner’s wage rate. For households with multiple earners, responsibility is shared. We assume that responsibility for earning enough for the household as a whole to meet the poverty line is shared among adults in the same proportion as they presently contribute to the household’s total actual earnings.50 Each partner’s ‘necessary time in paid labour’ is then just the amount of money represented by his or her proportional share of the household’s total paid-labour income needs, divided by his or her wage rate. Working just long enough to earn a poverty-level income usually would involve people working far less hours than they actually do at present. In calculating ‘necessary time in paid labour’ on the basis of people’s present wage rates, we are implicitly assuming that a person’s wage rate does not vary depending on how many hours he or she works. That is a common enough assumption in much economic modelling.51 But there are also good economic reasons for doubting it.52 In earlier studies, therefore, we tried estimating an ordinary leastsquares regression predicting wage rates on the basis of hours worked 50

51

52

Even if children of the household bring earnings into the household, we assume they should be assigned no responsibility for doing what is necessary merely to get the income of the household as a whole up to the poverty line. In times past, children even as young as 6 years old made significant productive contributions within the household (Petty 1899, pp. 307–8; Folbre 2006, ch. 7), but that has not been generally expected of children in advanced industrial societies for several generations. See e.g.: Garfinkel and Haveman 1977; 1978; Haveman 1993; Haveman and Buron 1993; Saunders et al. 1994; Haveman and Bershadker 1998; 2001. See, e.g., Lundberg 1985.

Discretionary time and temporal autonomy

45

and a range of human-capital and other characteristics; and we used that equation to estimate the wage rate that a person would actually have, if working only the number of hours we deem ‘necessary’.53 Those adjustments turned out not to make much difference within the range relevant to our study, however. Hence we have omitted that complication in the analyses reported in this book. We have, however, performed one plausibility check on our implicit assumption that people can reduce their work hours while retaining their present wage rate. We have confirmed that it is possible (in the sense that some people are actually doing it) for people to receive high wage rates and still work no longer than we deem ‘necessary’.54 That is evidence that there are at least some such jobs out there to be had, from the individual’s point of view. From the larger societal point of view, that suggests that it is possible for all such jobs to be arranged in timeflexible ways – which is an issue to which we shall return in our conclusions. Of course that does not prove that any given individual could stay in the same job and reduce his or her hours of work at will. People may, and presumably often would, have to change jobs to do that. But what that evidence seems to suggest is that, whatever the person’s present wage rate, there is some job out there that pays that wage rate for the generally substantially lower number of hours that our ‘necessary time in paid labour’ calculations envisage that person working. There may or may not be many such jobs.55 There may or may not be enough such 53 54

55

For details see Goodin et al. (2005, pp. 55–6) and associated footnotes. Specifically, we look at people in the top third of the distribution of wage rates among earners within our more restricted sample. We examine the proportion of them who actually engage in paid labour for no longer than mean necessary time in paid labour for people in that top third of the wage-rate distribution, excluding for these purposes travel time to and from work. The proportions of people with ‘high wage rates and low hours in paid labour’ in the six countries are as follow: US 3.12%; Australia 7.45%; Germany 5.10%; France 2.06%; Sweden 3.41%; Finland 0.95%. These are not large percentages, to be sure, and Finland’s is particularly small: but remember, the question before us is simply whether there exist at least some such jobs (so households could rearrange their affairs in this way one-at-a-time). Looking just at what jobs are presently being done for how many hours at what wage rates cannot tell us how many there could be. How many hours people are presently working is a function partly of what their employers demand and partly of what they themselves prefer; and not many people voluntarily opt to work so few hours as to leave themselves near poverty.

46

Introduction

jobs to permit everyone to reduce their hours in paid labour simultaneously, and still find a job paying the same wage rate as they presently enjoy. We will make some public-policy recommendations along those lines in our concluding chapter. But for now just remember the nature of the analytic exercise upon which we are here embarked: for now, we are merely asking what would happen if households reduced their hours worked one-household-at-a-time, rather than all-at-once. There are two further adjustments that we do make in calculating ‘necessary time in paid labour’. One concerns travel time to work; the other concerns the costs of child care for people who need it in order to engage in paid labour. The child-care adjustment obviously only affects households with a pre-school child and with no stay-at-home adult to look after it. In households of that sort, however, it is a necessary precondition of parents’ employment that they make arrangements for someone to care for their pre-school child or children for as long as they need to be at work.56 Sometimes family or friends do so for free. But most typically working parents have to pay to put their children into commercial care, if public child care is not available. That should be seen as a ‘necessary expense of paid labour’, just as surely as the uniforms that workers can typically claim as necessary expenses on their tax returns. ‘Child-care expenses’ do not appear consistently in the LIS files nor in any other systematic cross-national collection that we have been able to discover. We have therefore been forced to estimate them in very indirect ways. We begin by estimating what it costs to provide child care, which is what we presume parents would have to pay for it were there no state subsidies.57 We assume that parents with poverty-level incomes will be charged virtually nothing for enrolling their children in publicly supported early childhood education and child care, where such programmes are available. The proportion of children using, or 56

57

We assume (conservatively, per sec. 2.2.1) that school-aged children can either take care of themselves after school or go home with friends or otherwise be costlessly minded. This assumption might be more problematic in some countries – especially perhaps Germany, where in the period under study the school day typically ended at noon (Gornick and Meyers 2003, pp. 230–1). We take that to be equivalent to OECD (1993; 1996; 1997a; 2003) reports of hourly expenditure on pre-primary education per student in public and private institutions for each country. That yields an estimate that is usually close to the unsubsidized market costs of child care, at least for the handful of countries for which we have been able to obtain (highly unsystematic) evidence on that.

Discretionary time and temporal autonomy

47

entitled to use, such programmes determines the extent to which it is necessary for parents to pay the full costs of providing child care.58 Having thus estimated a cost-per-hour of child care, net of government subsidization of those costs, we then multiply that by the household’s number of children of the relevant age and by the number of hours that the parent needs to work in paid labour, as calculated above.59 We then calculate how much extra time the parent would need to work in paid labour in order to cover those necessary costs of engaging in paid labour, and we add that extra time to the parent’s ‘necessary time in paid labour’ as estimated above.60 In order to engage in paid labour, one also needs to get oneself to work.61 That takes time, too. We calculate ‘necessary travel time to work’ as the mean amount of time a workday it takes to travel to and from work, times as many days as the above calculation suggests the person needs to be at work.62 We add that, too, to our estimate of

58

59

60

61

62

The extent of its availability can be determined from Gornick et al. (1997) and Gornick and Meyers (2003) which tell us: (a) whether or not (and for what age groups of children) there is a legal entitlement to publicly supported early education and care of children; and (b) the proportion of pre-school-aged children in a country that is enrolled in publicly supported early childhood education and care programmes. We assume a zero cost to poverty-line parents in countries where there is a legal entitlement to publicly supported child care. In other countries, we take the cost of subsidized child care to be, on average, the cost of unsubsidized care times one minus the proportion of pre-school-aged children enrolled in publicly subsidized child care (whose parents we again assume pay nothing). Or in the case of two-earner couples, for as long as the partner with the lesser amount of ‘necessary time in paid labour’ needs to be at work. Among dualearners, responsibility for child-care costs is apportioned in the same way as is responsibility for income described above. Insofar as governments subsidize child care through cash benefits to parents or tax subsidies, those will of course have already been picked up in our calculation of government taxes and transfers to parents with incomes around the poverty line, as described above. Engaging in paid labour at or from home is not common in the countries under study. That is to say, we take the mean amount of ‘actual travel time to work’, among all those in the country who report being at work on the day in question. We calculate how many days people need to be at work on the assumption that they can bundle all their ‘necessary time in paid labour’ into standard 8-hour workdays. That may not always be true, of course; but that procedure gives us the lowest possible estimate of ‘necessary travel time to work’, which is preferable for reasons given in sec. 2.2.1.

48

Introduction

‘necessary time in paid labour’, thus yielding the final estimate that will be used in the analysis in this book. Households in which people spend less time than we thus deem ‘minimally necessary’ in paid labour will be in poverty. Table 2.1 has reported the proportion of the population living in households in poverty, for each country and averaging across all six countries under study. Given the broad academic consensus on this methodology for calculating poverty, those proportions serve as our ‘benchmarks’ for assessing the appropriateness of the standards of what ought to be deemed ‘minimally necessary’ in the other two realms. Of course, different countries have different proportions of people in poverty: we ought not to expect uniformity in that respect. But if our standards of what counts as ‘minimally necessary’ in all these realms are equally demanding, we ought to expect a comparable proportion of people in each country (and especially averaging across all countries) to fall below each of those three standards of necessity.

2.2.3 Necessary time in unpaid household labour Our method of calculating ‘necessary time in unpaid household labour’ is closely analogous to the procedures described above for calculating the poverty line for money. We begin by calculating, from time-use files, the amount of time actually spent in unpaid household labour by all the members of a person’s household combined.63 As with cash costs so too with the costs of time in unpaid labour, there are economies of scale that come from living together in larger 63

Where time-use files contain reports from all members of the household, this is straightforward. However, some surveys – including ones in two countries crucial to our study, Sweden and the US – collect data on the time use of only one individual. There, to get total household time in unpaid household labour for households with two adults, we calculate the mean of unpaid household labour for spouses by using the subgroups: gender, household type (dual-earner, oneearner couple, no-earner couple), earner status (no earner or earner), household size. Total unpaid household labour of household is then calculated by summing the hours of adults, one of whom is in the data file and the other of whose hours come from the estimated hours of spouses in households of that person’s type. To cross-check this procedure, we have applied it to countries where we have timeuse information from everyone in the household; and we find that this procedure yields estimates of necessary time in unpaid labour that are very similar to those yielded by our standard procedure.

Discretionary time and temporal autonomy

49

households (it takes longer, but not twice as long, to cook and clean up for four people compared to two). We use the same ‘equivalence scale’ as for income – the square root of the number of people in the household – to render equivalent unpaid household labour across households of differing sizes. We divide the household’s actual total unpaid household labour time by that equivalence scale, and assign that to each adult in the household as his or her ‘equivalent unpaid household labour time’. We then calculate a ‘poverty line for unpaid household labour’ in the same way as the money poverty line: as ‘half the median equivalent actual unpaid household labour time across all people in the country’. In each country, we calculate such ‘poverty lines for unpaid household labour’ separately for four different groups of people, whom we expect to have very differing needs in relation to unpaid household labour, in particular child care. First we distinguish households with pre-school-aged children from those without. We further distinguish households with a stay-at-home adult from those with all adults in paid labour. Once again, the total amount of time the household as a whole needs to devote to unpaid household labour is the poverty line for households of that type, times the equivalence scale. In households with only one adult, we take that to be that adult’s ‘necessary time in unpaid household labour’. (While children might help with the housework, we assume here as with cash income that meeting the household’s bare minimum needs is the responsibility of adults alone.64) Among couples, we apportion responsibility between the two partners for the household’s total necessary time in unpaid household labour according to the proportion of the household’s total actual unpaid labour that each of those partners presently does. 64

In times past, children’s work – both in paid labour and unpaid household labour – was crucial to household survival, and it remains so in much of the Third World. But in the sorts of countries under discussion in this book, it is not. There, the ‘job’ of the child lies in ‘getting an education’ and, more generally, building up human capital (Miller 2005). Children in lone-parent families certainly do more unpaid household labour than in two-parent families, and often they do ‘necessary tasks’ (like cooking dinner). But in assigning ‘necessary time’ for purposes of our ‘discretionary time’ calculations, we do so on the basis of responsibilities. And the responsibility for feeding the family is surely the parent’s, whosoever happens to do it on any given day: one of the children, or the school barbeque, or whoever.

50

Introduction

That method of calculating ‘necessary time in unpaid household labour’ has obvious procedural affinities with calculations of the money poverty line. The best way to determine whether applying those procedures yields substantively similar results in those two very different realms is to examine the proportion of people doing less unpaid household labour than we thus deem ‘minimally necessary’. If the proportion doing less than ‘minimally necessary’ in unpaid household labour by this standard is about the same (in each country and especially averaging across all countries) as the proportion doing less than ‘minimally necessary’ in paid labour to escape income poverty, then we would conclude that our standards of what is ‘minimally necessary’ are pretty comparable across those two realms. As we have seen in Table 2.1, that is indeed broadly the case. In four out of our six countries, the first two columns are very similar indeed. The proportion of the population doing less than we deem ‘minimally necessary’ in paid labour is substantially higher than that doing less than ‘minimally necessary’ in unpaid household labour only in the US; and the converse is the case only in Finland. What those two cases are probably picking up is simply the fact that where incomes are higher people can afford to buy in more substitutes for necessary unpaid household labour that they would otherwise have had to do for themselves (take-away rather than home-cooked meals; hiring a cleaner or a child-minder rather than doing those things themselves). Conversely, where incomes are lower people do more of the necessary unpaid household labour themselves.

2.2.4 Necessary time in personal care Estimating the amount of time people need to spend in personal care in exactly the same manner would yield substantively implausible results. The median amount of time people actually spend in personal care in the countries under examination is around 70 hours a week; setting a ‘poverty line for personal care’ at half that would imply that people only strictly need around 5 hours a day to sleep, eat, groom, etc. That would be ridiculously low.65 65

By way of diagnostics: The reason the half-median strategy does not work with personal care seems to be that most items of ‘personal care’ are highly inelastic (often on the order of 0.2). So people with a lot of time to spare do not spend a lot

Discretionary time and temporal autonomy

51

As the original Multinational Comparative Time-Budget Research Project remarked, ‘while there is no doubt that human beings could tend to these basic needs a bit more briefly if the situation required, the time normally given to these activities in everyday life probably does not exceed by any very large margin the minimum necessary . . .’.66 Hence, in estimating ‘necessary time in personal care’, we set each person’s ‘necessary time in personal care’ equal to four-fifths (rather than half) of the median amount of time people in their country actually spend in personal care. That procedure yields results that have surface plausibility. In most countries ‘necessary time in personal care’, estimated that way, is around 8 hours a day. That sounds like just time enough for a short night’s sleep, a quick shower and meals on the run. It would all be very rushed; but remember, estimates of what is the ‘minimum necessary’ should be low. As we have seen in Table 2.1, the proportion of people that fall below this ‘minimum necessary’ standard of personal care varies somewhat from country to country. But within any given country – and certainly averaging across all six countries – the proportion of the population that falls below the ‘minimum necessary’ standard in respect of personal care is broadly similar to the proportion that falls below the ‘minimum necessary’ standard in respect of paid labour and unpaid household labour. Thus we conclude that this way of operationalizing what is ‘minimally necessary’ in the realm of personal care sets about as tough a standard there as is set in the other two realms.

2.2.5 Discretionary time, pre- and post-government, and spare time The standard time-use method of calculating ‘spare time’ (or ‘free time’, as it is misleadingly called in that literature67) treats it as the

66

67

more time in personal care, and people with little spare time do not spend much less time in personal care, either (Harvey 1984, pp. 124–8). Robinson et al. 1972, p. 129. As evidence of this, note that the coefficient of variation across the population for ‘time in personal care’ is only around 0.20 in most countries, whereas that for the other two categories of time use is four (in the case of unpaid household labour) or five (for paid labour) times that (Gershuny 2000, p. 143). We explained previously, but it bears reiterating: Spare time is ‘free time’, in the sense it is time free of paid labour, unpaid household labour and personal care. But if we are really interested in how much time you have free, we should add to

52

Introduction

time left over, out of 168 hours per week, after all the time actually spent in paid labour, unpaid household labour and personal care has been deducted. Thus: spare time ¼ 168 hours per week minus actual time in paid labour minus actual time in unpaid household labour minus actual time in personal care We will discuss ‘spare time’, and how it compares to our measure of ‘discretionary time’, in part II. Our measure of ‘discretionary time’ is structurally similar to that standard measure of spare time. It is likewise what remains after taking account of the same three categories of time use. The difference is that ‘spare (free) time’ is calculated on the basis of how much time a person actually spends in each of those categories of time use, whereas ‘discretionary time’ is calculated on the basis of how much time a person strictly needs to spend in each of them. Thus, discretionary time ¼ 168 hours per week minus necessary time in paid labour minus necessary time in unpaid household labour minus necessary time in personal care Note that, while ‘spare time’ can never be less than zero, it is logically possible (and occasionally happens) that ‘discretionary time’ is negative. That occurs whenever people’s wage rates are so low that, in order to get a poverty-line income, they would have to work more hours than there are in the day (after taking into account their necessary time in personal care and unpaid household labour).68 The procedures described in sections 2.2.2 through 2.2.4 above relate to how we reckoned necessary time post-government. That is to say, we there describe the calculation of how much time would be necessary in each of three sorts of activities, after taking into account

68

that – as our ‘discretionary time’ concept does – time you are free to choose whether or not to engage in those activities. These are even more dramatic cases than Haveman and Bershadker’s (1998; 2001) ones of people who cannot be ‘self-reliant’, in the sense that they cannot achieve a poverty-line income even if they were to work full-time.

Discretionary time and temporal autonomy

53

the effects of government taxes, transfer payments and child-care subsidies. Post-government discretionary time figures form the basis of our discussion in parts II and V. In parts III and IV, however, we will be trying to reckon the impact of various welfare and gender regimes’ taxes, transfers and child-care subsidies on people’s discretionary time. To do that, we need some measure of how much discretionary time people would have had in the absence of those government interventions (which we call ‘pre-government discretionary time’). Of course, we cannot possibly predict what the world would be like without any government at all. We cannot even predict what the macroeconomy (and people’s wage rates and hence ‘necessary time in paid labour’) would be like if all taxes and government expenditures disappeared. The best we can do, within the scope of the present project, is merely to assume that those charges and payments disappear and nothing else changes.69 One way to think about this is to imagine government taxes, transfer payments and child-care subsidies being removed for only one household at a time. In calculating pre-government discretionary time for the people within any given household, we assume that that household pays no taxes and receives no public-transfer payments. We recalculate necessary time in paid labour on that basis. We also assume that there is no government-subsidized child care, so the cost of child care for parents in need of it thus rises to the full private-market cost of child care; and we adjust ‘necessary time in paid labour’ accordingly, for working parents in need of child care. In countries where households with a stay-at-home adult access state-provided child care, we adjust such households’ necessary time in unpaid household labour upward to reflect the extra time they would have to spend in child care in the absence of that state-provided service.70 69

70

In this we are also ignoring adaptations that households themselves would make to that changed environment, except insofar as those adaptations are dictated by ‘necessities’ as we here define them. We assume that households with no stay-at-home adult get first call on places in publicly subsidized child care. Based on that assumption, we estimate for different age groups of children the proportion of children of stay-at-home parents who are in publicly subsidized child care as the proportion of children in the relevant age group who are in publicly subsidized child care minus the proportion of children of that age that have no stay-at-home adult, all divided by one minus the proportion of children of that age that have no stay-at-home adult.

54

Introduction

2.3 Validating the measure The principal justification for using this notion of ‘discretionary time’ in empirical studies of social life is that it answers well to our key normative concerns with freedom and autonomy, sketched in section 2.1. In this section, we shall strive to demonstrate that this indicator of ‘discretionary time’ really does actually track two things that theoretically it should in the empirical world. Specifically, we shall show that ‘discretionary time’ (over and above ‘spare time’) is a good predictor of people’s subjective sense of suffering ‘time pressure’ (section 2.3.1) and, indeed, of people’s satisfaction with life as a whole (section 2.3.2). Through those two demonstrations, we empirically validate our notion of ‘discretionary time’.

2.3.1 Subjective time pressure Let us begin, most simply, with the question of what makes people subjectively ‘feel rushed’ or ‘pressed for time’. One reason for them feeling that way might be that they objectively are short of time: there just are not very many hours a day left, net of all the time that people actually spend on the ‘necessities of life’ (paid labour; keeping house; eating and sleeping). That obviously must be a large part of the story. But if ‘control matters’, sociologically and psychologically (as we argue it ought philosophically), then the extent to which people feel subjectively ‘time-pressured’ would also be expected to be a function of how much control they have over how much time to devote to all those activities. If control matters to perceptions of ‘time pressure’ in these ways, then people who put in long hours at work, but who do so by their own choosing, ought not to report themselves as feeling ‘pressed for time’ in the same way as those who put in similarly long hours without any choice in the matter. Time-use surveys sometimes contain questions that allow us to pick these issues apart. Among the countries under examination here,

We increase the necessary time in unpaid household labour of households with a stay-at-home adult and children in the relevant age group by an amount equal to this proportion times an estimation of the average hours per week that publicly subsidized child care was provided for children in such child care.

Discretionary time and temporal autonomy

55

two have time-use surveys containing both questions about subjective time pressure and questions sufficient for us to estimate (after a fashion71) ‘discretionary’ and ‘spare time’ in ways described above. The Finnish Time Use Survey of 1999–2000 asked respondents how often they were ‘feeling rushed’.72 The Australian Time Use Survey of 1997 asked people how often they felt ‘rushed or pressed for time’.73 Remember what we are looking for here. We would not necessarily expect the coefficients to be exactly the same in these two countries. The phrasing of the questions and response categories differ (even beyond, obviously, one being in English and the other Finnish). Indeed, the concept of ‘time pressure’ and social conventions governing it might differ from country to country. Still, if temporal autonomy matters to people in the way we have been suggesting it ought, then our ‘discretionary time’ variable ought to have a significant impact on people’s subjective sense of being ‘pressed for time’, even controlling for how much ‘spare time’ they actually have. Table 2.2 reports the results of logistic regressions predicting how changes in how much ‘spare time’ respondents have and in how much ‘discretionary time’ they have would impact on the odds of respondents reporting themselves as ‘feeling rushed’. We put each of those two ‘time’ variables into the equation, first one-at-a-time (Models 1 and 2) and then both together (Model 3). Putting both variables into the

71

72

73

We can estimate ‘discretionary time’ exactly as described above for the Finnish data, which is the same as used in the rest of the book, but not for the 1997 Australian Time Use Survey. The 1992 Australian survey we use in the rest of the book lacks the ‘feel rushed’ question, which is why we must turn to the 1997 Australian data for the present exercise. We do not use that 1997 Australian data in the rest of the book, however, because people’s wage rates cannot reliably be calculated in the nearest LIS income survey for Australia (the 1994 survey reports respondents’ annual income and the number of hours worked per week but not the number of weeks they worked per year). The wage rate used in calculating ‘discretionary time’ in the analysis reported in Table 2.2 is based on even more rudimentary income and working hours information contained in the 1997 Australian Time Use Survey itself. Respondents were invited to answer ‘always’, ‘sometimes’ or ‘almost never’. For purposes of the logistic regressions reported in Table 2.2, we collapsed the first two responses. Respondents were invited to answer ‘always’, ‘often’, ‘sometimes’, ‘rarely’ or ‘never’. For purposes of the logistic regressions reported in Table 2.2, we collapsed the first three responses, as well as the last two.

56

Introduction

Table 2.2. Subjective time pressure as a function of spare and discretionary time Dependent variable: for Finland, ‘feeling rushed’ (1 ¼ ‘always’ or ‘sometimes’; 0 ¼ ‘almost never’); for Australia, ‘feels rushed or pressed for time’ (1 ¼ ‘always’, ‘often’ or ‘sometimes’; 0 ¼ ‘rarely’ or ‘never’) Australia 1997 Model 1

Model 2 Model 3

Finland 1999 Model 1 Model 2

Discretionary 0.983** 0.986* time Spare 0.983** 0.986** 0.992* time N 1950 1667 1667 943 Chi-square 22.231** 8.515** 20.255** 4.988*

0.982**

Model 3 0.983** 0.993*

943 943 10.522** 14.559**

Notes: (1) Weighted logistic regressions (2) Odds ratio effect (3) ** significant at the 0.01 level; * significant at the 0.05 level

equation simultaneously allows us to ascertain the effect of each one of those variables, controlling for the other. It is only to be expected that people’s subjective sense of time pressure should depend on how much time pressure they are objectively under, operationalized here by how little ‘spare time’ they actually have. In Table 2.2 (Models 1) we find that in both countries an increase in ‘spare time’ does indeed lead to a significant reduction in the odds of ‘feeling rushed’. But if our speculation about temporal autonomy is correct, so should an increase in ‘discretionary time’ – as indeed it does, in Table 2.2 (Models 2). Most importantly, our thesis predicts that the effect of ‘discretionary time’ should persist, even after the effect of ‘spare time’ has been taken into account. That is precisely what we see in Models 3 in Table 2.2. In both countries, the effect of ‘discretionary time’ on the odds of ‘feeling rushed’ is significant even controlling for the effect of ‘spare time’ in the same equation. Indeed, the effect of ‘discretionary time’ is about as great as the effect of actual ‘spare time’ itself on the sense of ‘feeling rushed’, when both those variables are taken into account in the same equation.

Discretionary time and temporal autonomy

57

2.3.2 Happiness and life satisfaction Psychologists have long studied people’s subjective well-being through their responses to a very simple ‘happiness’ question. The 1995 German Socio-Economic Panel Study, which forms the basis for the analysis we shall be discussing here, approached people with the question: ‘We would like to ask you about your satisfaction with your life in general . . . How happy are you at present with your life as a whole?’ People responded to that question by locating themselves on a scale ranging from zero (‘totally unhappy’) to ten (‘totally happy’). Responses to this very simple question turn out to track surprisingly much of what we would a priori expect ought to make a difference to the quality of people’s lives – being married, being healthy, being employed, being politically empowered and so on.74 In the arcane language of psychologists, those ‘satisfaction’ measures ‘seem to contain substantial amounts of valid variance’.75 Government advisors have begun recommending public policies on the basis of them.76 Even economists are now beginning to treat responses to this lifesatisfaction question as ‘proxies for ‘‘utility’’ ’ in their models of social behaviour.77 One survey that asks respondents about both ‘life satisfaction’ and ‘time use’ is the German Socio-Economic Panel Study. The time-use component of the GSOEP is rudimentary, compared to the time-use surveys discussed in the rest of this book.78 Still, the conjunction of the time-use component together with the happiness question makes GSOEP an invaluable resource for the present exercise. Table 2.3 reports the results of ordinary least-squares regressions predicting ‘satisfaction with life as a whole’ on the basis of two time-use variables, two economic variables and a range of socio-demographic controls. The two time-use variables are as before: ‘discretionary time’ and ‘spare time’. It has long been known that spare time correlates, in 74 75 76 77

78

Veenhoven 1984; 1999; Dow and Juster 1985, pp. 410–12. Diener 1984, p. 551; see also Diener et al. 1999; Kahneman et al. 1999. Donovan and Halpern 2002. Frey and Stutzer 2002b, p. 405; see also 2002a; Easterlin 2001; Layard 2005; Van Praag and Ferrer-i-Carbonell 2005; Blanchflower and Oswald 2006. It is based on whole-day recall of how much time respondents spend in a handful of pre-coded activities, pointedly excluding ‘time spent in personal care’, which we had to construct by exclusion. See Eriksson et al. 2007.

58

Introduction

Table 2.3. Satisfaction with life as a whole, as a function of time and money Dependent variable: satisfaction with life as a whole Germany 1995 Model 1 Discretionary time, log Spare time, log Household income, log Wage rate, log Sociodemographic characteristics included N R2

Model 2

Model 3

0.288*(2.06)

0.288*(2.03)

0.160**(5.87) 0.343**(6.49)

0.298**(5.49)

0.166**(5.54) 0.253**(4.76)

0.018(0.55) yes

0.003(0.09) yes

0.020(0.53) yes

11183 0.082

9583 0.071

9390 0.080

Notes: (1) Weighted linear regressions with robust standard errors (2) ** significant at the 0.01 level; * significant at the 0.05 level (3) t-values in brackets (4) The socio-demographic characteristics included in the regressions are: age, gender, citizenship, marital status, children living at home, education, employment status and household size. For details see Eriksson et al. (2007).

curvilinear fashion, with life satisfaction.79 The question we are here investigating is whether discretionary time also has an effect on life satisfaction, even controlling for spare time. The two economic variables are ‘total household income’ and ‘individual’s wage rate’, both of which might well be thought to affect life satisfaction (the former because ‘having more money at one’s disposal makes one happier’, the latter because ‘greater individual success in the competition of life makes one happier’).80 79 80

Campbell et al. 1976, pp. 356–7. Note that, while the combination of ‘hourly wage rate’ and ‘own earned income’ would dictate ‘hours worked’, there are other sources of ‘total household

Discretionary time and temporal autonomy

59

All four of those variables are logged, for the same reason that money is standardly logged in such regressions. Like money, both success in the competition of life and time (either spare or discretionary) presumably display ‘diminishing marginal utility’ with respect to life satisfaction. When you do not have much, an extra unit means a lot to you. When you have a lot, an extra unit does not mean nearly as much. On the economic side, what Table 2.3 tells us is that it is money rather than ‘success in the competition of life’ that matters to people’s life satisfaction. ‘Household income’ is significant in all these models. ‘Wage rate’ is significant in none of them.81 But our principal focus here is on the ‘time’ variables. Model 1 shows that ‘spare time’, taken on its own, matters to people’s life satisfaction – albeit only about half as much as ‘household income’. Model 2 shows that ‘discretionary time’, taken on its own, matters to people’s life satisfaction far more – about equally as much as ‘household income’.82 But again, Model 3 is the one of most real interest here. There, we enter both ‘spare time’ and ‘discretionary time’ as independent variables simultaneously. That allows us to assess the impact of ‘discretionary time’, controlling for the amount of ‘spare time’ a person actually has. The message of Model 3 is that ‘discretionary time’ – control over how much ‘spare time’ one has – remains significant as a predictor of life satisfaction, even controlling for the amount of ‘spare time’ one actually has.83 Indeed, looking at the magnitude of the coefficients, our best estimate is that any given proportionate change in ‘discretionary

81

82

83

income’ than your own earned income (e.g. your spouse’s earned income; your income from capital). And while ‘hours worked’ is one determinant of ‘spare time’ there are other determinants as well (i.e., hours in unpaid household labour and hours in personal care). So the intercorrelations among these variables are not so substantial as to invalidate the model. To foreshadow, ‘wage rate’ proves irrelevant; and simply dropping it from the model would make virtually no difference to our findings with respect to the other variables’ impact on ‘happiness’. As noted above, all the other results are virtually unchanged if we simply drop the ‘wage-rate’ variable from the models reported in Table 2.2. Although the standard error for the ‘discretionary time’ variable is also substantially greater, causing the t-statistic and significance to be lower. The GSOEP files also contain a question concerning ‘satisfaction with the amount of leisure time one has’. Performing a similar regression on that, ‘discretionary time’ and ‘spare time’ have about as much effect one as the other (the coefficients are just over 0.3 in each case), while neither of the economic variables is significant; that model explains R2 ¼ 0.173 of the variance in satisfaction with amount of leisure time. See Eriksson et al. 2007.

60

Introduction

time’ is two-thirds again more important than the same proportionate change in ‘spare time’ in generating life satisfaction; and that its effect is on a par with the effect of a similar proportionate change in ‘household income’ itself.

2.3.3 Discretionary time matters The upshot of these two exercises – one exploring the sense of subjective time pressure, the other exploring correlates of ‘life satisfaction’ – is that ‘discretionary time’ as here operationalized is not just of philosophical interest. It also has strong sociological and psychological resonances. Our measure of ‘discretionary time’ is dictated in the first instance by theoretical considerations of necessity and autonomy. But judging from the results of these last two empirical cross-checks, it seems that those notions do indeed track something powerfully important in the political sociology of the real world, as well.

3

The distribution of discretionary time

Subsequent chapters will analyse the results of those sorts of calculations in detail, from many different angles. This chapter is offered purely as an introductory sampler, showing the basic sorts of results we get when applying the methodologies just discussed to the MTUS and LIS data sets for the six countries under study. The aim of the present chapter is threefold. In it, our hope is to show:

 There is enough variation – across individuals; across types of persons and households within countries; and across countries themselves – in the amounts of people’s discretionary time for those differences to be interesting and important subjects for further discussion.  Although the magnitudes differ from country to country, the basic patterns are sufficiently similar across all these countries to allow us to make certain generalizations without reference to any particular country, as we sometimes will be doing (in chapters 5, 14 and 15).  Although the basic patterns are similar across all the countries, the magnitudes differ sufficiently from one country to another to allow us to make certain inferences about the differential effects of countries’ varying welfare regimes and gender regimes (as we do, respectively, in chapters 7–9 and 10–12).

3.1 Large variation At the most general level, we might be concerned just with the average (mean) amount of discretionary time available to people in each of the six countries under discussion. Figure 3.1 reports both the overall average, country by country, and the average broken down by gender. Figure 3.2 breaks that down by household type, reporting the average (mean) amount of discretionary time available to people who are in 61

62

Introduction

85.22 85.88 84.49

83.80 85.47 82.21

Australia

76.08 78.88 73.42

US

81.08 81.34 80.79

80

78.18 78.91 77.45

90

80.34 81.31 79.42

100

Sweden

Finland

hours per week

70 60 50 40 30 20 10 0

Germany total

men

France women

Figure 3.1 Mean discretionary time, overall and by gender

childless dual-earner couples, to people who are in dual-earner couples with children and to people who are lone parents in paid labour. There are other household types that are also of interest, of course. Those will be discussed at length in subsequent chapters, alongside the ones shown in Figure 3.2. So too are there obvious and important distinctions to be made within Figure 3.1’s broad groupings of ‘men’ and ‘women’, as subsequent discussions will also show. For purposes of this preliminary introduction to the basic structure of our results, however, that handful of categories suffices. One final preliminary note of explanation: Figures 3.1 and 3.2, and indeed all subsequent such figures in the book, group countries according to regime types. Grouped together at the left are the US and Australia, representing liberal welfare regimes and individualistic market-oriented gender regimes. Grouped together at the right are Sweden and Finland, representing social-democratic welfare regimes and female-friendly gender regimes. Grouped in the middle are Germany and France, representing corporatist welfare regimes and traditional familialistic gender regimes. More will be said about the nature of those regimes, and how each of those countries fits those regime types, in chapters 7–12 below. Inspecting Figure 3.1, we see that the amount of discretionary time available to people varies from one country to another. The nationwide

The distribution of discretionary time

63

mean varies from just over 76 hours a week in France to over 85 hours a week in Sweden. Looked at one way – the way that is most visually compelling to look at it, given the modest differences relative to the overall length of the bars in Figure 3.1 – that might not be such a big difference. It amounts to saying that Swedes have one-eighth more discretionary time than the French, on average. One-eighth is not a huge number. That is what our eyes tell us, in just glancing at Figure 3.1. But looking at those differences in absolute terms, rather than as proportions of total discretionary time, a very different picture emerges. In absolute terms, what Figure 3.1 is telling us is that Swedes have over 9 hours more discretionary time each week than the French, on average. Think about it. That is like having Monday and an hour or two on Tuesday off work. Put that way, it is clearly a big difference. At this highly aggregated level of country-wide comparisons, differences in mean discretionary time track conventional welfare-gender regimes only imperfectly. True, average overall discretionary time is greatest in the two social-democratic female-friendly countries, Sweden and Finland. But at this level of aggregation, liberal-individualist US looks more like corporatist-traditionalist Germany than it does liberal-individualist Australia. And liberal-individualist Australia looks more like corporatist-traditionalist France than it does the liberalindividualist US. Regime clusters emerge clearly only once we begin disaggregating these results and looking at specific population subgroups of regime-specific concern, in chapters 8 and 11 below (although in gender terms important differences remain between Germany and France, both conventionally classified as corporatisttraditionalist regimes). The amount of discretionary time available to people varies not only across countries but also within countries. Consider first the variation according to gender. Figure 3.1 shows that women have consistently less discretionary time than men in all countries under discussion. How much less, exactly, varies from one country to another: from a high of 5.5 hours a week less in France to a low of just over half-an-hour a week less in Germany, with the gender gap elsewhere generally hovering around 2 hours a week. While the differences according to gender within countries are less than the differences in nationwide averages between countries, the gender gap in discretionary time is still important. Men having two hours more discretionary time a week than

64

Introduction 100

93.82

90 80 hours per week

91.62

85.89 87.74

81.19

70

81.94

77.97

84.50

76.95 69.92

60 50

95.15

91.46

89.07

62.01

59.55

61.71

59.69

51.40

40 30 20 10 0

US

Australia

Germany France

dual-earners without kids

dual-earners with kids

Sweden

Finland

lone parents

Figure 3.2 Mean discretionary time, by household type

women is rather like men not having to go into work until 11 a.m. Monday mornings, when women have to be there at 9 a.m. on the dot. The amount of discretionary time available to people varies even more substantially across household types. Across all our countries, childless dual-earners average more discretionary time than ones with children, as Figure 3.2 shows. How much discretionary time having children costs dual-earner couples varies – from over 11 to almost 13 hours a week in Australia and the US, to just over 7 hours a week in Finland and Sweden, hovering around 9 hours a week in the other two countries. In the terms we have previously been using, it is like parents having to put in an extra day or more at work each week. As Figure 3.2 also shows, lone parents (and remember that throughout this book, we are talking exclusively about lone parents who are actually in paid labour) have substantially less discretionary time, yet again. Being a lone parent, compared to being a parent with a partner who is also employed, costs people anything between 15 hours’ (in France and Australia) and over 29 hours’ (in the US) worth of discretionary time a week. That is like lone parents having to work 2, 3 or almost 4 extra days a week. Comparing household types that are most-advantaged to those that are least-advantaged in terms of discretionary time, the differences are even more dramatic. Lone parents have far less discretionary time than

The distribution of discretionary time

65

childless dual-earner couples, in all our countries – but, as Figure 3.2 shows, much more so in some countries than others. The most dramatic difference is in the US, where lone parents have fully 42 hours a week less discretionary time than childless dual-earner couples (think of that as their working an extra full-time job for more than 5 days a week). But even in France and Sweden, where the difference is smallest, lone parents still have around 25 hours a week less discretionary time than childless dual-earner couples (equivalent to having an extra job for 3 days a week). Clearly, people do have substantially differing amounts of discretionary time, depending on their country, their gender and their household circumstances. Certainly there is enough variation in this measure to be well worth further analysis.

3.2 Common patterns The second thing to notice from Figures 3.1 and 3.2 is that, while the magnitudes vary from country to country, the same basic patterns recur across all the countries. In Figure 3.1 we see that women consistently have less discretionary time than men, on average, in all six of the countries under study. In Figure 3.2 we see that dual-earner couples with children consistently have less discretionary time than ones without them, on average, across all our countries; and lone parents have still less, yet again. Thus, we can talk meaningfully across countries of, for example, the extra time-cost of having children or of being a lone parent. We can generalize about those patterns across countries not only discursively but also statistically. We can average across all these countries, when it suits our expository purposes to do so, confident that those averages will not be masking different basic patterns in different countries. Our analyses in chapters 5 and 14 take advantage of that fact.

3.3 Country differences While the basic patterns are indeed the same across all our countries, those patterns are more pronounced in some countries than in others. Thus, for example, while women on average have less discretionary time than men in all the countries under study, that difference is much less in Germany than in France (0.5 hours a week, compared to 5.5).

66

Introduction

While lone parents on average have less discretionary time than childless dual-earners in all our countries, that difference is much less in France and Sweden than in the US (around 25 hours a week, compared to 42). The basic pattern is the same. But the magnitudes clearly matter, also – particularly when the magnitudes are of that magnitude. We will explore the cross-national variation in those magnitudes to surmise the effects of the different policy regimes found in each country. Looking at each country’s performance in securing discretionary time for their citizens in chapters 7–12, we will find the familiar regime types reassuringly re-emerging. The US and Australia cluster together, in the way that standard discussions of their liberal welfare regimes and individualist-market-oriented gender regimes would suggest. Sweden and Finland cluster together, in the way that standard discussions of their social-democratic welfare regimes and female-friendly gender regimes would suggest. Germany and France cluster together in broadly the way that standard discussions of their corporatist welfare regimes would suggest (although when it comes to gender, they come apart rather than forming a single traditionalist-familialist regime). The principal advantage of assessing the performance of these regimes using the metric of discretionary time is – as we have said – that that enables us to calibrate regimes’ contributions to people’s autonomy in a way that truly is interpersonally comparable and transnationally meaningful.

PART II

Time pressure

4

Time pressure: a new problem?

Whatever the objective facts of the matter – and they are controversial at least at the margins1 – it is nonetheless clear that people’s subjective sense of time pressure has become increasingly urgent. Americans complain ever more stridently that they are ‘overworked’;2 the ‘leisure class’ that it is ‘harried’;3 working wives about their ‘second shift’ at home, and the ‘time bind’ that they suffer in consequence.4 President Clinton’s Council of Economic Advisers pronounced the ‘time crunch’ to be real, and recommended a suite of policies to ameliorate it for the sake of America’s families.5 In this chapter we will briefly survey this ongoing discussion concerning the pressure on people’s time, how it might have changed over the years and why. We are not particularly concerned to adjudicate those controversies here, however. Our principal focus in this book is not so much with history – distant or recent – as with the current state of affairs. In chapter 5 we will discuss how much time pressure different groups in society are suffering. We will go on there to discuss the extent to which that time pressure is inevitable or optional. Those analyses in chapter 5 thus constitute the empirical core of this part of the book, with the present chapter being merely intended to set the scene for what follows.

1

2

3 5

Compare, particularly, Schor (1992; 2000) with Robinson and Godbey (1999). See similarly: Juster and Stafford 1991; Bittman 1992; 1999a, b; Esping-Andersen 1999, ch. 4; Bittman and Wajcman 2000; Gershuny 2000; Sullivan and Gershuny 2001. Schor 1992; 2000. The proportion of Americans reporting that they ‘always feel rushed’ rose from 24 per cent in 1965 to 33 per cent thirty years later (Robinson and Godbey 1999, p. 232). Linder 1970. 4 Hochschild 1989; 1997; Baxter and Gibson 1990. US Council of Economic Advisers 1999.

69

70

Time pressure

4.1 Time pressure in a broad historical perspective Every generation thinks it is unique: uniquely cursed, or uniquely blest, or sometimes both at once. Is time pressure truly a uniquely contemporary phenomenon? One hackneyed answer is, ‘Quite the opposite!’ In earlier periods in human history, that argument would go, people had to work from dawn to dusk to meet their bare subsistence needs. By comparison with more primitive times, modern industrial society makes us better off, not merely in terms of consumer goods and creature comforts but also in terms of leisure time itself.6 Closer inspection of the way of life of primitive peoples belies that thought. The best we can tell, people in hunter-and-gatherer societies typically spent moderately short amounts of time meeting their subsistence needs. Anthropologists tell us that ‘if the Kapauku of Papua work one day, they do no labor on the next. !Kung Bushmen put in only two and a half days a week and six hours per day. In the Sandwich Islands of Hawaii, men work only four hours per day. And Australian aborigines have similar schedules.’7 Earlier civilizations, too, were considerably more leisured than our own, at least for the ruling classes. In the Roman republic, fully a third of the days were marked on the calendar as ‘unlawful for judicial and political business . . . Whatever the work schedules of slaves and women, leisure for the ruling classes, for administrative and professional men, was never again so abundant.’8 In agriculture, people historically worked dawn-to-dusk, but only at very particular times of the year. When harvesting and when planting, they worked those long hours; over the long winter months, however, there was far less to do. Averaged over the year, their total hours

6

7 8

In a table depicting ‘a developmental typology of time use’, for example, Gershuny (2000, p. 32) says ‘pre-modern societies’ involve ‘for the majority, long hours of subsistence work’, whereas in ‘modern industrial (or ‘‘post-industrial’’) societies’ there is ‘declining work time, providing opportunities for increasing consumption’. Note that this stylized characterization is at odds with his own estimates (p. 25). Schor 1992, p. 10. See similarly Gershuny 2000, pp. 51–2. Wilensky 1961, p. 33. The ruling classes remained especially privileged in this respect to almost our own day: ‘A ten-month year in the upper civil service in Britain was usual in 1800 and prevailed until World War II’ (p. 34).

Time pressure: a new problem?

71

resembled those of average American workers today.9 Likewise in urban areas in the middle ages: ‘while daily schedules of 12–16 hours are reported as early as the 13th century . . . the number of days off and the record of one- or two-hour lunch periods and half-hour breaks suggests that annual hours were little more than they are for a fully employed worker today’.10 In pre-industrial Europe in both urban and rural life, ‘the work pattern was one of alternative bouts of intense labour and of idleness, wherever men were in control of their own working lives’.11 Things changed with the commodification of labour time, first in agriculture and then most especially in early manufacture. When buying people’s time rather than commissioning a particular task, employers’ incentives suddenly shift towards demanding ever more output for every hour’s labour that they have purchased. Workers who are paid by the task can afford to work at their leisure, whereas workers paid by the hour are subject to the very different time-discipline imposed by their employers.12 The true horror stories come with the ‘Satanic mills’, the textile mills and engineering workshops of the early industrial revolution.13 For English manufacturing workers, the workday climbed over the course of a single century ‘from a 12-hour day with a two-hour rest in 1700, to a 14- to 18-hour day in 1800’. Truly, at that point, ‘the stage was set for humanitarian protest against the costs of industrialism’.14 Our image of the time pressure suffered by earlier generations is powerfully shaped by accounts – of novelists and poets, as much as of early social scientists15 – of those nineteenth-century horrors. Things

9

10

11 12

13 14 15

Schor 1992, p. 45; Young and Willmott 1973, pp. 126–7; cf. Wilensky 1961, p. 34. Wilensky 1961, p. 34; 2002, pp. 51–2; see similarly Young and Willmott 1973, pp. 125–6; Schor 1992, p. 45. Thompson 1967, p. 73. ‘Those who are employed experience a distinction between their employer’s time and their ‘‘own’’ time. And the employer must use the time of his labour, and see it is not wasted: not the task but the value of time when reduced to money is dominant. Time is now currency: it is not passed but spent’ (Thompson 1967, p. 61). Thompson 1967, p. 85. Wilensky 1961, p. 34; see similarly Schor 1992, p. 45. Blake 1804; Engels 1844; Dickens 1854.

72

Time pressure

have indeed improved for the vast majority of workers a very great deal since then. Still, it is important to recognize what a historically peculiar period that was. Certainly the burden of labour in paid employment lessened over the course of the twentieth century. But in the larger historical sweep, that only marked ‘a return to the work schedules of medieval guildsmen’.16 As is standard in discussions of ‘hours of work’, of course, the account we have just given is of ‘work in the narrowest sense’. It leaves out unpaid work in the home: shopping, cooking, cleaning, child care. Which is to say, of course, it leaves out the bulk of the work done by women for much of human history.17 Let one poem stand in lieu of any formal quantification of women’s work time in the home, over the ages. Published in 1739, it was penned by Mary Collier (‘now a Washerwoman, at Petersfield in Hampshire’) and entitled, The Woman’s Labour: An Epistle to Mr Stephen Duck; in Answer to his Late Poem, called The Thresher’s Labour: when we Home are come, Alas! we find our Work but just begun; So many Things for our Attendance call, Had we ten Hands, we could employ them all. Our Children put to Bed, with greatest Care We all Things for you coming Home prepare: You sup, and go to Bed without delay, And rest yourselves till the ensuing Day; While we, alas! but little Sleep can have, Because our froward Children cry and rave . . . In ev’ry Work [we] take our proper Share; And from the Time that Harvest doth begin Until the Corn be cut and carry’d in, Our Toil and Labour’s daily so extreme, That we have hardly ever Time to dream.18

16 17

18

Wilensky 1961, p. 35. Although it ought to be noted that, in primitive societies, provisioning for the ‘household’ is gender-differentiated, with men typically doing the quick work of ‘hunting’ and women doing the endless ‘gathering’ that provides the bulk of the ‘household’s’ calories. Quoted in Thompson 1967, p. 79.

Time pressure: a new problem?

73

4.2 Contemporary sources of additional time stress Whether working time, in either paid labour or unpaid household labour, increased or decreased over the final decades of the twentieth century is a matter of some controversy. That controversy arises in part because different methodologies yield different conclusions.19 It arises in larger part because different countries and different subgroups within a country have genuinely different experiences.20 Generalizing by averaging across whole populations can easily mask divergent patterns for different groups within those populations.

4.2.1 The demands of paid labour In terms of time in paid labour, there is increasing variation across countries. In the US average hours have held broadly constant for many years; and the same is true (among the countries under study here) in Australia, Finland and Sweden. In continental Europe (represented here by France and Germany) they declined by around 7 per cent over the last decade of the twentieth century. That was largely due to deliberate government policies designed to reduce the hours worked by each worker and thereby create jobs for more workers.21 The jobcreation aspect of those policies may not have been particularly successful, perhaps, but the leisure-creation aspect of them certainly has. In France and Germany employees now work only about 80 per cent as many hours a year as in the US.22 19

20

21 22

Cf. Schor 1992; Coleman and Pencavel 1993a, b; Robinson and Bostrom 1994; Robinson and Godbey 1997; Figart and Golden 1998; Jacobs and Gerson 2000; Bianchi et al. 2006. For a good summary of the basic issues involved, see Schor 2000. For example, Bittman and Rice (2002) find no marked change between 1974 and 1997 in the average amount of time spent in paid labour by Australians of ‘prime working age’ (25 through 54 years of age). There had, however, been a significant redistribution of paid labour from men to women, and time spent in paid labour had also become more dispersed. Although the combined time spent in paid labour by older couples decreased on average, among those of prime working age combined time in paid labour increased. Australian workers also spent increasing amounts of time in paid labour outside 9 a.m. to 5 p.m. on weekdays. Similarly increasing diversity in working hours underlies the relatively constant US averages (Rones et al. 1997; Clarkberg and Moen 2001; Jacobs and Gerson 2004). Alessina et al. 2005. OECD 2005a, table F, p. 255; note however that the OECD warns that differing sources might compromise cross-country comparisons based on this table.

74

Time pressure

Another big change in the labour market towards the end of the twentieth century was in the composition of the workforce. A large part of that story lies in the increasing participation of women in the paid labour force. That occurred sooner, and more fully, in some countries than others. The employment rate for prime-aged (25 to 54) women has long hovered around 70 per cent in the US and Australia and around 80 per cent in Sweden and Finland. Traditionally, it has been much lower in the countries of continental Europe. Over the course of the 1990s, however, it rose dramatically.23 By 2000, the employment rate among prime-aged women in France and Germany was roughly on a par with that in the US or Australia.24 Among men, labour-market experiences have become increasingly bifurcated, with some (the younger unskilled; the prematurely retired) having not enough work while others have too much. Over the last decade of the twentieth century, the employment rate dropped for both young men (aged 15 to 24) and older men (aged 55 to 64) across the EU-15.25 The employment rate of the younger and older groups of men is now around half that of prime-aged male workers across the European countries discussed here, and only slightly higher in the US and Australia.26 (These differences constitute part of the reason why, in the empirical chapters that follow, we will be focusing our own analyses on prime-aged workers exclusively.) All of those changes in the paid labour market have dramatic consequences for the amounts of time different people in different places spend in paid labour.

4.2.2 The demands of unpaid household labour In terms of unpaid household labour, there seem to be three ‘big stories’ to be told. One – a ‘non-story’, actually – is that, despite dramatic improvements in household technology, the amount of time spent on household tasks has not actually shown any correspondingly dramatic

23 24 25

26

By fully 15 per cent, across the EU-15 members of the OECD. OECD 2005a, pp. 247–9. Including in all four European countries here under study in the case of the former, and all but Finland in the case of the latter. OECD 2005a, pp. 244–6. For further discussion see, e.g.: Epstein and Kallenberg 2004; Jacobs and Gerson 2004.

Time pressure: a new problem?

75

decline.27 Indeed, the amount of time a full-time US homemaker spends in unpaid household labour has hovered around 50 hours a week since at least the 1920s.28 It remains so to this day. The composition of ‘time in unpaid household labour’ has shifted, away from ‘routine domestic work’ (cooking and cleaning) and towards other components (child care particularly, but also shopping and odd jobs).29 But the total time in unpaid household labour of all kinds has held constant (or if anything slightly increased) over the twentieth century. A second ‘big story’ bearing on these themes is, of course, the increasing time pressure on women as they move into the paid-labour market. As they do, women reduce their time in unpaid labour at home, but not remotely hour-for-hour for every hour they spend in paid labour. Their male partners increase their own time in unpaid household labour, but again not by nearly as much as working wives reduce theirs. The upshot is that rather less unpaid household labour gets done overall in the dual-earner household – but the woman’s total combined time in paid and unpaid household labour is substantially greater than is the typical non-employed woman’s in unpaid household labour alone.30 In the words of the original Multinational Comparative Time-Budget Research Project report: There is nothing surprising in the fact that women who try to maintain a dual role of paid work and housekeeping tend to fall between employed men and housewives in the amount of time spent in both domains [of paid and unpaid

27

28

29

30

The initial Multinational Comparative Time-Budget Research Project team reports ‘variations in time devoted to household obligations . . . correlate with household technology only very weakly’ (Robinson et al. 1972, p. 125). Schor 1992, pp. 86–8; based on Vanek 1974. Note however that Lebergott (1984, table 38.1, quoted in Esping-Andersen 1999, p. 53) puts the number in 1900 much higher, at over 80 hours per week. For example, the Middletown study reports the proportion of working-class women spending more than 9 hours a week in washing and ironing dropped from 61 per cent before the introduction of the washing machine to 24 per cent afterwards (Lynd and Lynd 1937, quoted in Gershuny 2000, p. 67). Note well that the sharp reductions reported by Gershuny and Robinson (1988; Gershuny 2000, ch. 5; Gershuny and Sullivan 2003; Sullivan and Gershuny 2001) in women’s time in unpaid household labour pertain exclusively to ‘routine domestic work’ (called ‘core domestic work’ in the later versions). Indeed, Gershuny (2000, p. 126) reports that ‘the broader category of unpaid work . . . shows small overall increases from the 1970s onwards once we control for structural change’. Gershuny 2000, chs. 5, 7.

76

Time pressure

household labour time]. What is important, however, is the extreme time pressure that the dual role creates. For when the times spent on the two types of work are summed together, the working woman is much busier than either her male colleagues or her housewife counterpart. After her day’s obligations are done, she finds herself with an hour or two less time than anyone else, and this pattern . . . appears ‘universally’, at all of our survey sites.31

This is a finding that has been constantly replicated in time-use surveys ever since.32 A third ‘big story’ bearing on these themes concerns changing patterns of family life.33 Over the last decade of the twentieth century, there was a fall in both rates of marriage (at least in the US, Australia and Germany) and childbearing (particularly dramatically in Sweden; not at all in France). Divorce rates rose (dramatically so in France, Germany and Australia). By 2000, nearly half of all marriages ended in divorce across all six countries.34 All of these factors have important temporal consequences. Having children dramatically increases time pressure; avoiding having them avoids that. The breakdown of marriages involving children can radically exacerbate time pressure if a lone parent is forced to serve as both breadwinner and homemaker combined.35 The childless, particularly the single childless, suffer far fewer strains.

4.2.3 The combined effects on people’s time Perhaps the best way of calibrating the magnitude of the effects of these changes on people’s time is to quote at length from the ‘Executive

31 32

33 34

35

Robinson et al. 1972, p. 119. See, e.g.: Baxter and Gibson 1990; Bittman 1992; Gershuny 2000. According to the most recent US report, employed mothers’ combined time in paid and unpaid household labour exceeds fathers’ by 5 hours a week and non-employed mothers’ by 19 hours a week (Bianchi et al. 2006). Bittman and Pixley 1997; Lewis 2001. OECD 2005b, table GE5.3. In the US, the rise in the divorce rate was similarly precipitous in an earlier period, more than doubling over the two decades from 1960 to 1980 (Sawhill 1999, p. 100; Castles and Flood 1993, p. 304). Taking only a slightly longer historical perspective, however, the rate of contemporary single-parenthood is nowise unique. Calculations suggest that a child in the US ‘was as likely to grow up with only one parent in 1870 as in 1970’. All that has changed is the cause, with divorce replacing death (Esping-Andersen 1999, p. 51, based on Masnick and Bane 1980).

Time pressure: a new problem?

77

Summary’ of the 1999 US Council of Economic Advisers’ report Analyzing the ‘Time Crunch’: The American family has experienced dramatic changes over the last three decades – changes in the amount of time parents work for pay; changes in income and who earns it; changes in family size; and changes in how child care and household tasks are accomplished . . . *

*

*

*

Increase in Hours Worked. The hours American parents work in paid jobs have increased enormously since 1969 due to a dramatic shift of mothers’ time from the household to the labor market. In 1969, 38 per cent of married mothers worked for pay; in 1996, 68 per cent did so . . . Burdens on Women. Virtually all of the increase in total hours families spend on paid work has come from increases in women’s hours. While annual hours of paid work by all wives increased greatly – by . . . 93 per cent – husbands’ hours of paid work decreased slightly from 1969 to 1996. The ‘time crunch’ falls heavily on employed women who spend over one third less time on child care and household tasks than women without paid jobs, but still have 25 to 30 per cent less free time . . . Rise of the Single Parent. At the same time, the share of families with a single parent has expanded greatly since 1969. Single parents have half as much total time as two parents have and typically have less than half as much potential income. The rising number of single parents has increased the proportion of families that are ‘cash-strapped’ and ‘time-poor’. Need for Policies to Help Families. Increased time in market work among parents raises a key set of challenges to policy-makers seeking to help promote strong families, including the need for flexibility in paid work hours, the need for available and affordable child care, the need for effective ways to support the earnings of families with low-wage earning parents, and the need to encourage two-parent families to form and stay together.36

4.3 Better off but busier At least compared to people in the early years of the industrial revolution, people today are generally much better off in terms of time as well as money. In general, as GDP rises, so too does leisure time. Time-use studies show that to be true not only cross-nationally but also longitudinally over the decades for which we have reliably comparable time-use data. A pooled regression of all countries and surveys represented in the MTUS files suggests that ‘for every 1000 US dollars [at 36

US Council of Economic Advisers 1999, p. i.

78

Time pressure

1980 value] per capita of extra national income, our data show some 12 minutes less work (paid and unpaid) . . . and 16 minutes of extra leisure’.37 While that is true in comparisons across countries, it is not true of comparisons within countries. As countries get richer, they become more leisured on average.38 But as individuals get richer, they become less so.39 That is counterintuitive. Statistically, it is only possible because the ‘average’ masks shifts in different directions for different groups whose time use is being averaged together. Those whose incomes are static or falling become less intensively employed, while those whose incomes are rising become more so.40 That explains the statistical paradox, but not the substantive one. Why do people work themselves harder as they get richer, rather than taking advantage of some of their new-found wealth to buy themselves a more leisurely life? There are various possible answers. One is essentially sociological in character. It is a status symbol to be ‘busy’. That shows that you are socially valued, that your time is in demand.41 A second is essentially microeconomic in character. When your hourly wage rate increases, that has two effects that pull you in opposite directions. On the one hand, there is the ‘income effect’: you are richer now, and can more easily afford to buy things (including leisure). You simply do not need to work as much. On the other hand, there is the ‘substitution effect’: since your time is more valuable, the opportunity cost (i.e., the wage forgone) of any given hour’s leisure is higher. The first consideration would incline you to consume more leisure as

37 38

39

40 41

Gershuny 2000, p. 131. Although if economic growth is accompanied by increasing income inequality, average work hours increase as poorer individuals work longer hours to ‘keep up with the Joneses’ in consumption terms (Bowles and Park 2005). Hamermesh and Lee 2007. This is also reflected in Gershuny’s (2000, pp. 219–21) finding, pooling all twenty countries represented in the MTUS files, that both men and women spend more time in paid labour the higher their status, measured by educational attainment. Gershuny 2000, pp. 69–104. As Gershuny (2005a, p. 289; see similarly 2000, p. 9) speculates, ‘the growth in busy feelings may in part reflect an increasingly positive view of ‘‘business’’ that results from its association with the increasingly busy lifestyle of the most privileged groups in developed societies’.

Time pressure: a new problem?

79

your wage rate rises; the second would incline you to spend more time in paid labour as your wage rate rises. There is nothing in economic theory to tell us which of those two factors will win out in that tug-ofwar for your time.42 Sociologists say they ‘find it implausible’ that ‘high earners . . . choose intensively enjoyable activities and work longer to pay for them’.43 But economists demur.44 Allied to that is a third argument that is broadly socio-economic in character. That argument rests on the observation that consumption is itself a time-consuming activity. It takes time to spend money; and the more money you have with which to consume, the more pressure there is on your time in doing so.45 That would not in itself explain why getting richer makes you spend more hours in paid labour. (To do that, this third argument has to be combined with the second.) But this third line of analysis definitely would explain why the ‘leisure class’ feels increasingly ‘harried’: its leisure is increasingly ‘intensified’. A fourth line of analysis, again socio-economic in character, would analyse the phenomenon in terms of the ‘work-spend squirrel cage’.46 The more you work, the more you have to spend; and the more you spend, the more habituated you become to high-spending patterns, causing you to need to work even longer to get the same ‘buzz’ as before. ‘Like drug addicts who develop a tolerance, consumers need additional hits to maintain a given level of satisfaction.’47 As you get richer, you have a felt need to get still richer, leading to longer working hours. A fifth line of analysis would offer an economistic twist to the sociological talk of ‘status’ with which we began. In economics, ‘status goods’ are ones that have no fixed absolute value but are only valuable to you relative to how much of them others have.48 Weaponry is like that, if you are engaged in an armed struggle. Lawyers are like that, if 42 44 45 47 48

Wilensky 1961, p. 32. 43 Gershuny 2000, p. 9. Becker 1965; Linder 1970, ch. 3; Hamermesh and Lee 2007. Becker 1965; Linder 1970, ch. 3. 46 Schor 1992, chs. 5–6. Schor (1992, p. 124), building on Scitovsky (1976). In the words of Keynes (1930, p. 365), needs ‘fall into two classes – those which are absolute in the sense that we feel them whatever the situation of our fellow human beings may be, and those which are relative only in that their satisfaction lifts us above, makes us feel superior to, our fellows’. Dusenberry (1949), Shubik (1971), Hirsch (1976), Goodin (1990a) and Frank (1985), in their different ways, provide rational grounding for what was in Keynes an essentially sociological insight.

80

Time pressure

you are engaged in a courtroom battle. Money itself is like that, if you are engaged in a bidding war.49 And if you are competing for jobs or promotions with people who put in long hours, you must put in at least as long hours as they do to stand a chance of winning.50 Competitive dynamics can drive self-exploitation in this way, and might be another thing that leads people’s leisure time to decrease even as their incomes increase. Whichever of these (or whatever combination of them) constitute the real reasons for the phenomenon, the tendency for time pressure on people to increase as their income increases is another important source of the increasing sense of time pressure among better-off people in expanding economies.

49

50

If as in Bengal in 1943 there is a fixed stock of rice and you find yourself bidding for it against people whose incomes are suddenly rising, you will end up starving unless your income also rises (Sen 1981; 1983). Schor 1992, ch. 5.

5

Time pressure: a new measure

There is one striking but seldom-remarked feature of all the discussions of ‘time pressure’ described in the previous chapter. They all talk in terms of what people do with their time – how much of it they spend in paid labour, for example – with nary a mention of why. But surely ‘pressure’ implies compulsion. To say people are under ‘time pressure’ suggests that they are somehow forced to commit as much of their time as they do to the various activities they do, necessarily leaving so little remaining for other things they might do instead. Sometimes that unspoken assumption is enormously plausible. Subsistence farmers simply have to work as long as it takes to get their crops in, on pain of starvation. Sweatshop workers with no alternative means of subsistence likewise have to work as long as the factory boss demands, or starve. ‘Salarymen’ in Japan might be in the same position: maybe they would indeed be fired, and their families would starve, if they did not put in the long hours Japanese employers expect. Other people, however, work long hours purely because they enjoy their jobs, or because they have nothing else to do. Or they spend long hours preparing gourmet meals just for fun. Thus, simply looking at how much time people spend at work (for pay or at home) without asking why – which is what standard discussions of ‘time pressure’ invariably do – is a very imperfect way of assessing how much ‘time pressure’ people are really under or how ‘time poor’ they really are. Such studies may well find that people are ‘overworked’, in the sense of having a work–life balance that is seriously out of kilter. But strictly speaking, they may be under no pressure to do so. Our methodology is designed precisely to capture what those standard discussions of ‘time pressure’ miss – the extent to which people’s allocation of their time is dictated by strict necessities. There are certain activities of daily life that are unavoidable, and certain amounts of time that simply must be devoted to them. The more of your time it is strictly 81

82

Time pressure

necessary for you to commit to those activities, the less ‘discretionary time’ you have – and the more genuinely ‘pressed for time’ you truly are, in consequence. ‘Discretionary time’, in contrast, is precisely that: discretionary. It is ‘potential spare time’, time that you are perfectly free to spend outside paid or unpaid labour or personal care if you choose to do so. Of course, you might choose instead to spend some of that discretionary time doing more paid labour than strictly necessary, either because you enjoy your job or because you want things that the extra income buys you. If you do that, then it may be perfectly true (given the number of hours you are actually ‘rostered on’) that you have ‘little time left’ for anything else. But insofar as you could have freely chosen to roster yourself on far fewer hours than you did, any sense of ‘time pressure’ arising from that fact would be illusory. (Or fraudulent: but we will here systematically opt for the other, more charitable construction.) In this chapter, we will explore the extent of this gap between potential and actual spare time – a gap that we refer to as the potential for ‘time-pressure illusion’. We do so through the MTUS/LIS data and the measures we have constructed on the basis of them. As we have seen in chapter 3, people’s discretionary time (which we here interpret as ‘potential spare time’) varies considerably depending on their circumstances. So too, this chapter will show, does the extent of the potential for ‘time-pressure illusion’. The gap between how much ‘spare time’ people actually have and how much they could have had is much greater for some sorts of people than others. Ironically, the potential for ‘time-pressure illusion’ turns out to be the greatest for those who objectively have the most ‘discretionary time’, time that they could potentially have freed up for other activities had they chosen to do so. The chapter begins with some preliminary remarks on how ‘time pressure’ is conceptualized in ordinary time-use studies, and what we find wrong with that conceptualization (section 5.1). We proceed to examine the extent of genuine time pressure (which we conceptualize as a lack of ‘discretionary time’) and potential illusions people might have about it (based on a shortage of ‘actual spare time’), across the six countries under study. We discuss the magnitude and sources of time pressure and the potential for illusions about it (section 5.2). We show how that is distributed among different subgroups of the population, defined in terms of gender and parental status (section 5.3.1) and in terms of household types (section 5.3.2). We go on to examine the ‘best

Time pressure: a new measure

83

and worst’ cases of time pressure and of potential for illusions concerning it – dual-earner couples with no children, compared to lone parents (section 5.4). We close by observing just how misleading it would be to assess the relative time pressures on those groups, and also on parents within traditional male-breadwinner households, in terms of ‘spare time’ rather than ‘discretionary time’ (section 5.5).

5.1 Conceptualizing time pressure In standard time-use studies, the conceptualization of ‘time pressure’ starts by specifying a list of activities of daily life which are themselves ‘obligatory’.1 Those categories are the ones discussed in chapter 2: ‘paid labour’, ‘unpaid household labour’ and ‘personal care’. You have to eat and sleep and shower; you have to earn a living; you have to keep your household ticking over. Time-use studies then conventionally calculate your ‘spare’ (‘free’) time by looking at how much time you have left over, after those obligatory activities of daily life have been completed. Operationally, ‘spare time’ is calculated as the amount of time you have remaining, net of how much time you actually spend in all those ‘obligatory’ activities (as discussed in section 2.2.5 above). ‘Time pressure’, in conventional time-use studies, is simply the converse of ‘spare time’. The less ‘spare time’ (‘free time’) you have remaining, the more time-pressured you are conventionally said to be. However, that basic methodology is fundamentally flawed. True, the activities in view are genuinely ‘obligatory’: it is undeniably necessary to engage in them; they represent things that definitely must get done. But a person might choose to spend more time than strictly necessary engaging in those activities, or to achieve more in each of those realms than is strictly necessary. People generally like to (and generally manage to) get more sleep and earn more money than the minimum that is strictly necessary, for example. How much time people actually spend at their desks or over their stoves might be an important factor in determining their subjective sense of being ‘pressed for time’.2 It is 1 2

Robinson 1977, ch. 3. See similarly, e.g.: A˚s 1978; 1982; ABS 1998a, b. Analysis of data reported in section 2.3.1 above suggests that people’s sense of ‘feeling rushed’ is a function of both the lack of ‘spare time’ and the lack of ‘discretionary time’.

84

Time pressure

not, however, a good way of finding out whether they are genuinely ‘time poor’ in any deeper or more objective sense.3 Consider this analogy (to which we return in chapter 8 below). In ordinary poverty research, you are reckoned money poor if you do not have enough money to buy what you strictly need; and you are ‘nearpoor’ if you have little money left over after what you need to spend your money on. But money poverty is reckoned on the basis of spending needs, not spending choices. A spendthrift millionaire who chose to spend all her money on fancy food, clothing and mansions would rightly be regarded as having had a rich life, not an impoverished one.4 Time poverty, like money poverty, ought similarly to be defined not in terms of actual expenditures but instead in terms of necessary expenditures. You are time poor to the extent that you have little time left over after what you need (not after what you choose) to spend your time on. In order to assess genuine ‘time poverty’, we must focus not on how much time people actually spend on the necessary activities of daily life but rather on how much time they strictly need to spend on them. That is precisely what our indicator of ‘discretionary time’ is designed to do. It measures how much or little time you have left over, after doing the minimum necessary in three realms (paid labour, unpaid household labour and personal care). The difference between the amount of ‘spare time’ people actually have (after all the time they actually spend in all those activities) and their ‘discretionary time’ indicates the extent to which people are ‘busier’ than they strictly need to be, in the sense that they devote more time than strictly necessary to the admittedly obligatory activities of daily life. We dub that people’s potential for ‘time-pressure illusion’. Two further comments are in order before proceeding to our findings. One concerns the extent to which people really have a ‘free choice’ over how much extra time to devote to those tasks. We will discuss that matter more in chapter 6 below. For now, let us confine ourselves to a couple of more methodological remarks that might quell at least certain sorts of concerns.

 At the micro level, we may worry that any particular individual might be in some very particular circumstances: people with serious illnesses or peculiar physiologies may need to sleep longer than 3

Vickery 1977.

4

Ringen 1988.

Time pressure: a new measure

85

others; people with disabled children may need to spend more time in unpaid caring labour than others; some people with unusually inflexible jobs and unusually few employment options might have to work lots of hours in order to remain in paid work at all. All those things are true. But those are just the sorts of personal idiosyncrasies that talking in terms of ‘averages’ and ‘means’ washes out.  At the macro level, we might worry that there are ‘social expectations’ constraining people’s choice over how much time they devote to any particular task. But those are just the sorts of social norms that are supposed to be captured by relativizing our standards of ‘necessary’ to what others in one’s country typically do, in the ways described in chapter 2. Another concern might surround the appropriateness of a term like ‘illusion’ in this context. In one sense, the time pressure that people are actually under is not at all an illusion. They truly are working, cooking and sleeping all those hours; they really do have only that much spare time actually left over. What is potentially illusory is merely any sense of ‘pressure’ – any suggestion that they are ‘forced’ to do all those extra hours, above and beyond what is necessary (by the standards we have developed and justified in chapter 2). In another sense, too, the use of the psychologized term ‘illusion’ may be somewhat out of place in the present context. People may or may not feel ‘pressure’; and they may or may not be under any ‘illusion’ as to its true source. Apart from the data already reported (in section 2.3.1), we have no information on people’s subjective mental states. What we will be talking about here are more objective facts about people’s time use and commitments, which might potentially give rise to the illusion that they are under more time pressure than actually they are. Given how little time people actually have left over (outside of paid labour, unpaid household labour and personal care), they might well be forgiven for feeling pressed for time. But insofar as they could have had much more, had they only chosen to, any sense of being ‘pressed for time’ would to that extent be illusory. Any implication that they were forced to do what they had in fact freely chosen to do would be simply mistaken (or worse). When in describing our results we refer to people’s ‘time-pressure illusion’, that should always be read as shorthand for the more cumbersome phrase ‘potential for time-pressure illusion’. The

86

Time pressure

reference is not to people’s subjective states, as a literal reading might suggest. Instead, reference is to objective facts – the differences between the amounts of time that people actually spend on certain activities and how much they strictly need to spend on them – that might create the potential for that illusion to arise.

5.2 Magnitude and sources of time pressure The remainder of this chapter is devoted to interrogating the MTUS/ LIS data along those lines. In the figures presented here (and in the tables in Appendix 2 upon which those figures are based), we report the mean amount of ‘discretionary time’ (potential spare time), the mean amount of actual ‘spare time’ and the mean gap between the two (‘timepressure illusion’) for various groups of people. We report those statistics separately, for each of our six countries. By and large, the patterns are very similar across all six countries, however. So much of our discussion will simply be based on the first column of each figure, which averages across all six countries taken together. Figure 5.1 offers the broadest overview, averaging across the entire population within each country (and in the first column, across all countries). The top of the black bar shows how much ‘discretionary time’ people have on average in each of the countries. That represents how much time they potentially have free of obligatory tasks of daily life. The grey portion of each bar shows the far lesser amount of actual ‘spare time’ that is left to people on average, net of the time that they actually devote to those tasks. The difference between the two – the portion of the bar that appears in black in Figure 5.1 – represents the mean ‘time-pressure illusion’ among the population as a whole. As we see from Figure 5.1, there is a huge difference between the amount of ‘spare time’ that people actually have and the amount that they could have if they confined themselves to doing only what is strictly necessary. Potential spare time (i.e., ‘discretionary time’) is more than twice ‘actual spare time’ in every country under study, and nearly thrice in one of them (Sweden). Conversely, thinking (as time-use researchers typically do) of ‘time pressure’ as how little ‘spare time’ people have, the ‘real’ pressure on people’s time (calibrated in terms of ‘discretionary time’) is less than half what it might seem to be, looking (as time-use researchers typically do) merely at ‘actual spare time’. Looking not at how people choose to

Time pressure: a new measure

87

100 90 80.78

80.34

80

78.18

85.22

81.08

83.80

76.08

hours per week

70 60 50 40 30

39.93 33.16

35.73

33.48 29.62

29.99

30.19

France

Sweden

20 10 0 all countries

US

Australia discretionary time

Germany

Finland

spare time

Figure 5.1 Magnitude of time-pressure illusion nationwide

spend their time but rather at how people have to spend their time to calibrate the extent of the genuine ‘pressure’ they are under, time pressure is far less acute. The gap between those two ways of measuring ‘time pressure’ – and hence the grounds for illusions about how much time pressure you are ‘really’ under – is enormous. Across the whole population, it averages almost 48 hours a week across our six countries (from under 44 in Finland to over 55 in Sweden). If people fall into the trap of judging time pressure by the wrong (‘actual spare time’) measure, there is little wonder that they think themselves severely pressed for time. If they thought in terms of the right measure (‘discretionary time’), they would realize that the pressure on their time was not nearly so severe and that their time was much more under their own control – almost 7 hours a day more so, than if they looked at their ‘actual spare time’ alone. In Figure 5.2 we decompose the time-pressure illusion by showing exactly where those extra hours are spent. Averaging across the population as a whole and across all countries (which are all very similar in these respects), people generally spend about twice as much time as strictly necessary in both paid labour and unpaid household labour, and quarter again more time than strictly necessary in personal care. In terms of hours a week, that amounts to some 20 hours a week more than strictly necessary in paid labour, and almost 14 hours

paid labour unpaid household labour personal care

paid labour unpaid household labour personal care

paid labour unpaid household labour personal care

paid labour unpaid household labour personal care

paid labour unpaid household labour personal care

100 90 80 70 60 50 40 30 20 10 0

paid labour unpaid household labour personal care

Time pressure

paid labour unpaid household labour personal care

hours per week

88

all countries

US

Australia

Germany

France

Sweden

Finland

actual

necessary

Figure 5.2 Components of time-pressure illusion nationwide

a week more than strictly necessary in both unpaid household labour and personal care. It might be thought the similarities in the difference between ‘actual’ and ‘necessary’, both across categories of time use and across countries, are merely artifacts of our way of specifying ‘necessary time’ in each of these activities. The thought might be that, if you define ‘necessary’ as half the median amount ‘actually’ done, then naturally the one is about half the other in every case. But remember that calculations of both ‘necessary time in paid labour’ and ‘necessary time in unpaid household labour’ involve moderately dramatic nonlinear adjustments for household size.5 So, for example, ‘necessary time in paid labour’ could turn out to be much more or much less than half the actual – as indeed it does in Sweden, where ‘necessary’ is nearer a third the ‘actual’. The fact that in our other countries we find that people on average spend around twice as long as strictly ‘necessary’ in paid labour is, thus, a genuine finding rather than purely an artifact of our procedures.

5

Through the equivalence scale of the square root of household size, as explained in sections 2.2.2 and 2.2.3.

Time pressure: a new measure

89

5.3 Distribution of time pressure among subgroups of the population The patterns reported in Figures 5.1 and 5.2 provide nationwide baselines against which the patterns of subgroups of the population can be compared. First we look at how time pressure and illusions about it vary according to gender and parental status (section 5.3.1). Next we look at how those vary according to household types (section 5.3.2).

5.3.1 By gender and parental status

men without kids women without kids men with kids women with kids

men without kids women without kids men with kids women with kids

men without kids women without kids men with kids women with kids

men without kids women without kids men with kids women with kids

men without kids women without kids men with kids women with kids

all countries

men without kids women without kids men with kids women with kids

100 90 80 70 60 50 40 30 20 10 0 men without kids women without kids men with kids women with kids

hours per week

As we see from Figure 5.3, among people without children there is little difference between men and women in the amount of actual ‘spare time’, potential spare time (‘discretionary time’) or in the gap between them (the ‘time-pressure illusion’). That is true within each country, on average; and it is broadly true averaging across countries as well. Parents have rather less spare time, both actual and potential, in virtually all our countries. The gap between them (the ‘time-pressure illusion’) is however largely the same for parents and non-parents. Parenthood generally imposes more time pressure on mothers than fathers, however. On average across all six countries, having children

US

Australia

Germany

France

Sweden

Finland

discretionary time

spare time

Figure 5.3 Magnitude of time-pressure illusion, by gender and parental status

90

Time pressure

costs men 5 hours of actual ‘spare time’, compared to a little over 4 for women. But in terms of what we regard as the much more appropriate measure of potential spare time (‘discretionary time’), having children costs women almost 7 hours compared to under 4 for men. Germany is an exception (there mothers have about half-an-hour more discretionary time than fathers); but in all the other countries under study, the opposite is true. On average across all six countries, mothers have about 3-and-a-half fewer hours of discretionary time than fathers. That difference is not insignificant. In the terms we used in chapter 3 to help give a sense of how important such differences are, we might think of it as women having to work an extra half-day a week more than men. Still, the difference between men and women, and even between mothers and fathers, in the time pressure they are under is rather less than some rhetoric might have led us to expect.

5.3.2 By household types Much larger differences are found comparing the time pressure on people in different types of households. The different ‘household types’ we will be discussing are:

 households with two adults, both of whom are in paid labour, with     

no children present (‘dual-earners without kids’); households with two adults, only one of whom is in paid labour, with no children present (‘single-earner couples without kids’); households with one adult, who is in paid labour, with no children present (‘singles’); households with two adults, both of whom are in paid labour, with one child or more present (‘dual-earners with kids’); households with two adults, only one of whom is in paid labour, with one child or more present (‘single-earner couples with kids’); households with one adult, who is in paid labour, with one child or more present (‘lone parents’).

That does not exhaust all possible household types, to be sure. Most importantly, it ignores households in which there is no adult in paid labour. That is a particularly glaring omission in the case of lone parents, a large proportion of whom (in some countries, anyway) are not employed. But given the constraints of our methodology explained in chapter 2, that omission is simply unavoidable.

Time pressure: a new measure

91

Figure 5.4 displays the actual ‘spare time’, the potential spare time (‘discretionary time’) and the gap between them (the ‘time-pressure illusion’) for all six of those household types, for all six countries and averaging across all six countries. The country profiles differ a little – most importantly in their treatment of lone parents, more of which later. But before turning to that, let us focus on the averages across all six countries, in the suite of bar charts at the far left of Figure 5.4. In discussing the time-pressure costs of parenthood, and its differential impact on mothers and fathers, we were talking in terms of 7 hours a week difference at most. Looking across different household types, we see a difference of over 4 times that: a difference of 30 hours a week in ‘discretionary time’ (comparing lone parents to dual-earners without kids). Even comparing likes with likes, we find much more substantial differences than previously: people without children gain over 13 hours’ a week more potential spare time by partnering in dual-earner couples, as compared to remaining single; people with children lose over 20 hours a week becoming lone parents, compared to being dual-earner couples. Those differences across household types are substantially greater than differences across either gender or parental status. A second thing to notice in Figure 5.4 is that the actual ‘spare time’ available to people in different household types is generally pretty similar. People in childless single-earner couples invariably have a little more than the others; people in Finland have more across the board. But the grey bars in Figure 5.4 are all about the same height. The black bars are not: ‘discretionary time’ (i.e., potential spare time) is very substantially more variable than actual ‘spare time’. Assuming (as we believe) that potential spare time is the right way to measure genuine ‘time pressure’, that implies that the genuine time pressure on people in different types of households is very much more different than it might appear when measuring that (as time-use researchers conventionally do) in terms of actual ‘spare time’ alone. The third thing of significance to notice in Figure 5.4 is that, while the time pressure on different household types varies, the case that clearly stands out is that of lone parents. Averaging across all six countries, they have fully 30 hours a week less ‘discretionary time’ than the most temporally privileged household type (childless dual-earners). And in some countries they are substantially worse off than that, even: as we have already noted in chapter 3, lone parents in the US have fully 42 hours a week less ‘discretionary time’ than childless dual-earners;

hours per week

dual-earners without kids single-earner couples without kids singles dual-earners with kids single-earner couples with kids lone parents dual-earners without kids single-earner couples without kids singles dual-earners with kids single-earner couples with kids lone parents dual-earners without kids single-earner couples without kids singles dual-earners with kids single-earner couples with kids lone parents

Figure 5.4 Magnitude of time-pressure illusion by household type

all countries US Australia Germany

dual-earners without kids single-earner couples without kids singles dual-earners with kids single-earner couples with kids lone parents

France

spare time

dual-earners without kids single-earner couples without kids singles dual-earners with kids single-earner couples with kids lone parents

Sweden

discretionary time

dual-earners without kids single-earner couples without kids singles dual-earners with kids single-earner couples with kids lone parents

Finland

100 90 80 70 60 50 40 30 20 10 0 dual-earners without kids single-earner couples without kids singles dual-earners with kids single-earner couples with kids lone parents

Time pressure: a new measure

93

that is like forcing lone parents to work more than 5 days a week more, compared to childless dual-earners. It is a truly massive difference.

5.4 The best and the worst

men in dual-earner couples without kids women in dual-earner couples without kids lone mothers

men in dual-earner couples without kids women in dual-earner couples without kids lone mothers

men in dual-earner couples without kids women in dual-earner couples without kids lone mothers

US

men in dual-earner couples without kids women in dual-earner couples without kids lone mothers

all countries

men in dual-earner couples without kids women in dual-earner couples without kids lone mothers

men in dual-earner couples without kids women in dual-earner couples without kids lone mothers

100 90 80 70 60 50 40 30 20 10 0

men in dual-earner couples without kids women in dual-earner couples without kids lone mothers

hours per week

Let us probe more deeply that last comparison: between the households with the most time pressure, objectively construed, and those with the least. By ‘objectively construed’ we mean those households with the least versus the most potential spare time (i.e., ‘discretionary time’). By that standard, childless dual-earner couples are under the least time pressure, while lone parents are under the most, of all the household types shown in Figure 5.4. The statistics reported in Figure 5.4 for each household type average across gender. The first thing we will do, in looking more closely at those two particular household types, is disaggregate by gender. As we see in Figure 5.5, when we do it does not make much difference for childless dual-earners; men and women in such households are

Australia

Germany

France

Sweden

Finland

discretionary time

spare time

Figure 5.5 The difference between households with the most and least time pressure

94

Time pressure

essentially identical, both on average across all countries and within each country, in the amounts of actual ‘spare time’ and of ‘discretionary time’ that they have. Disaggregating ‘lone parents’ by gender, however, reveals that ‘lone fathers’ (few though there are in each country) are importantly different from lone mothers; their profile is more similar to ‘singles’ than it is to ‘lone mothers’. Hence, for purposes of this ‘best versus worst’ comparison, we will ignore ‘lone fathers’ and confine our attention to ‘lone mothers’ alone. Looking at the height of the grey lines in Figure 5.5, we see that in terms of their actual ‘spare time’ – the way conventional time-use researchers measure ‘time pressure’ – ‘childless dual-earners’ and ‘lone mothers’ appear strikingly similar. That is true both averaging across all six countries and, indeed, within each of the six countries. Looking only at actual ‘spare time’, it would appear that ‘childless dual-earners’ are under just as much time pressure as are ‘lone mothers’. However, looking at it in terms of potential spare time – ‘discretionary time’ – we see that the two groups differ dramatically.6 ‘Discretionary time’ is indicated by the top of the black bars. The lower the bars, the less ‘discretionary time’ people have and the more objectively time-pressured they are. Focusing on that, it is clear that lone mothers are objectively under much greater time pressure than are people in dual-earner couples without children. As we have already observed, lone mothers have over 30 hours a week less potential spare time, averaging across all countries. In the most dramatic case – that of the US – they have over 45 hours a week less. Country clearly matters here (so in France, the difference is more like 25 hours a week). But the difference is still enormous, even in that country where that difference is smallest. The other thing to notice from Figure 5.5 is that the ‘time-pressure illusion’ – the gap between potential and actual ‘spare time’ (the

6

Our choice of equivalence scale in sec. 2.2 contributes minimally to this result. Taking the square root of household size assigns a lone mother’s first child a weight of 0.41, her second child a weight of 0.32 and so on in decreasing fashion for each subsequent child. The principal alternative to the square-root-ofhousehold-size equivalence scale, the so-called ‘OECD scale’, would assign each child under 14 a weight of 0.5 (Atkinson 1998, p. 15). The OECD scale would tend to lead to the lone mother’s poverty line to be higher, and her necessary time in paid labour higher and hence her discretionary time even lower, especially for households with many children.

Time pressure: a new measure

95

portion of the bar that appears in black in Figure 5.5) – is very, very much smaller for lone mothers than for childless dual-earners. Indeed, on average it is about half as large. Those who are genuinely under the most time pressure (those who have the least potential spare time, i.e., the least ‘discretionary time’) are also under the least illusion concerning that fact. Researchers employing ordinary time-use methodologies would infer – quite wrongly – that childless dual-earners are under the same time pressure as lone mothers, when they observe the fact that the actual ‘spare time’ available to the one group is about the same as that available to the other. Taking our measure of ‘discretionary time’ as a better measure of ‘potential spare time’, that is simply crazy. An interesting question remains. How exactly do childless dualearners spend their time, so as to make them seem as time poor as lone mothers to ordinary time-use researchers? In Figure 5.6 we decompose the time use of childless dual-earners, in an attempt to answer that question. Figure 5.6 is a slightly complicated diagram. But basically what it tells us is that childless dual-earners spend vastly more time (by a factor of around four) in paid labour than they strictly need to do. They spend over twice as much in unpaid household labour as they need to do as well; and women in such households spend a bit more of their discretionary time in unpaid household labour and a bit less of it in paid labour than do men in such households, on average. But basically, it is the extra time that both men and women in childless dual-earner couples devote to paid labour, over and above what is strictly necessary, that is driving these results.

5.5 Another example: burden-sharing in male-breadwinner families We have just seen how, looking at ‘spare time’ alone, we might be led to conclude that childless people in dual-earner couples are just about as time-pressured as are lone mothers. Looking instead at ‘discretionary time’, we see just how badly mistaken any such conclusion would be. The same might be said about burden-sharing among parents in traditional male-breadwinner households. Looking purely at ‘spare time’, it would seem that both the breadwinner father and his stayat-home wife are shouldering very similar temporal burdens. That is broadly true in all six countries under study. But, as seen in

hours per week

Figure 5.6 Decomposition of the time-pressure illusion of childless dual-earners

all countries women men US women men Australia women men Germany women men France women men Sweden women men Finland women men

necessary actual

100

90

80

70

paid labour unpaid household labour personal care paid labour unpaid household labour personal care

60

paid labour unpaid household labour personal care paid labour unpaid household labour personal care

50

paid labour unpaid household labour personal care paid labour unpaid household labour personal care

40

paid labour unpaid household labour personal care paid labour unpaid household labour personal care

30

paid labour unpaid household labour personal care paid labour unpaid household labour personal care

20

paid labour unpaid household labour personal care paid labour unpaid household labour personal care

10

0 paid labour unpaid household labour personal care paid labour unpaid household labour personal care

all countries

Australia

discretionary time

Sweden

84.47 40.16

78.27 34.93

employed fathers

stay-at-home mothers

82.18 31.09

73.74 27.69

employed fathers

79.94 32.72

France

stay-at-home mothers

employed fathers

32.08

66.30

81.21 35.79

Germany

stay-at-home mothers

employed fathers

stay-at-home mothers

35.31

72.37

76.21 31.62

67.76 27.50

employed fathers

36.83

US

97

stay-at-home mothers

86.12

73.84 30.04

employed fathers

stay-at-home mothers

81.69 34.70

72.05 31.26

employed fathers

100 90 80 70 60 50 40 30 20 10 0

stay-at-home mothers

hours per week

Time pressure: a new measure

Finland

spare time

Figure 5.7 Time pressure on parents in traditional male-breadwinner households

Figure 5.7, it is particularly true in Germany and France, the two corporatist countries where this ‘traditional’ household type is particularly salient. In those countries, stay-at-home mothers and their husbands put in almost literally the same number of hours in their respective jobs. Looking at ‘spare time’ is once again grossly misleading, however. True, stay-at-home mothers actually spend as much time working for the household as do their breadwinner husbands. But looking not at how much time they actually spend but at how much time they strictly need to spend on their tasks, stay-at-home mothers have vastly more discretionary time than their breadwinner husbands – over 13 hours more in France and around 9 in Germany (and indeed across all six countries on average). Of course, there is always the ‘was it really voluntary?’ question to be asked, perhaps particularly pointedly as regards the sort of patriarchal household under discussion here. Perhaps the sort of husband who would allow his wife to stay home with the children might also insist upon far higher standards of housekeeping than we here deem ‘strictly necessary’. Calibrating the stay-at-home mother’s ‘necessary time in unpaid household labour’ instead in terms of her patriarchal husband’s expectations-cum-demands, she may well have far less genuinely ‘discretionary time’ than we suppose.

98

Time pressure

Alternatively, of course, self-selection may be at work. The sort of wife who would prefer to stay at home with the children may also be the sort of wife who herself has higher standards of housekeeping than average, and does more unpaid household labour than strictly necessary not because her husband insists but rather because she herself wants to do so. Or perhaps it is simply a matter of Parkinson’s law at work, as people with lots of ‘discretionary time’ expand their workloads to fill their days. The disparity between the relative amounts of ‘spare’ and ‘discretionary time’ available to parents in traditional malebreadwinner households thus admits of multiple interpretations. Our main point here is merely that there is an important difference between breadwinning fathers and stay-at-home mothers in these traditional households. Even if it can be explained away, it would be simply wrong to elide the difference, in the way that would be done by focusing on their relative amounts of ‘spare time’ alone. Looking at their relative amounts of ‘discretionary time’ instead, we at least see that there is an important difference here that is in need of further investigation.

6

Is it really an illusion?

Economists have coined the phrase ‘the money illusion’. That refers to the way in which people often mistakenly cue on the amount of money that they are offered, rather than on what that amount of money will buy for them.1 The money itself is no illusion. That is indeed what they are being offered. The illusion lies in their conflating two things (there, money and consumption opportunities) which are not the same and which can sometimes come dramatically apart. And insofar as they do, focusing on the money rather than what the money will buy you amounts to focusing on the wrong thing. It amounts to fetishizing money, when (at least for most of us) money is only of value by virtue of what it can buy for us. We offer our notion of the ‘time-pressure illusion’ in a similar spirit. By that, we refer to the difference between the ‘time pressure’ that people actually experience (in terms of their actual ‘spare time’) and that which they strictly need to experience (in terms of their ‘discretionary time’). As with the money illusion, so too with the timepressure illusion: the time pressure that people actually experience is not in the least illusory. People really are spending all those hours in paid labour, unpaid household labour and personal care; they really only do have that many hours to spare, in their actual lives. Here again, the illusion lies in conflating two things (here, actual ‘spare time’ and potential ‘discretionary time’) which are not the same and which can come dramatically apart. Insofar as they do, focusing on actual ‘spare time’ rather than on ‘discretionary time’ amounts to focusing on the wrong thing. It amounts to fetishizing the particular choices people have made, when what really matters (from most people’s own point of view, as well as from the philosopher’s) is their ability to choose. Some commentators resist any argument of that sort. Juliet Schor, for example, complains that it ‘embodies a blame-the-victim approach. 1

Keynes 1930, p. 369; Fisher 1928; Shafir et al. 1997.

99

100

Time pressure

It assumes that the problem lies in personal shortcomings, and counsels that we ‘‘try to do too much’’ or do not organize our lives efficiently.’2 Others might resist our argument because of what they see as its affinities with Catherine Hakim’s much-maligned ‘preference theory’ that puts the problems of working mothers down to their own choices.3 ‘Women’s choice between family work and market work’, she insists, is ‘a genuine choice in affluent modern societies.’4 If mothers choose to go out to work, that is their choice and ought to be respected. If they are under increased time pressure because of that choice, well, that is their choice. Governments, on Hakim’s view, ought to do no more to alleviate the plight of working mothers than they do to support the choices of mothers who have chosen to stay at home with their children.5 To some critics, our way of talking about the ‘time-pressure illusion’ seems uncomfortably similar. It might seem to be saying, ‘If people choose to work longer hours than strictly necessary, their choice ought of course 2 3

4 5

Schor 1993, p. xv. We here present Hakim’s (2000) argument in the vaguely caricatured way it has entered the general debate, since it is that caricatured version which bears most affinity with the sorts of arguments that might be levelled against our own analysis. Hakim’s own emphasis is not so much anti-working mothers as prostay-at-home mothers, who she believes should receive parity of treatment in public policy. When asking what government can do to help stay-at-home mothers, Hakim (2000, pp. 227–36) points to strictly money measures: changes to tax and benefit policies. Putting more money into their pockets, however, does nothing to alleviate the specifically temporal pressures on stay-at-home mothers (unless they spend it on home help or child-minders, which on Hakim’s own account they would typically prefer not to do). Working mothers could of course benefit in temporal terms both from subsidized child care and from generous taxand-benefit policies; putting more money into their pockets reduces their ‘necessary time in paid labour’, and hence increases their ‘discretionary time’, in a way that putting more money into the pockets of their stay-at-home counterparts who bear no responsibility for raising money through paid labour cannot. Government-subsidized pre-school schemes do help to increase the discretionary time of stay-at-home mothers in Germany and France, on the evidence of sec. 11.3. Thus, if we are concerned with time pressure specifically, and if we wanted to help stay-at-home mothers on a par with working mothers, those would be the sorts of policies we should recommend, rather than the fiscal ones on which Hakim concentrates. Hakim 2000, p. 1. In the closing words of her keynote address to a conference of Australian policymakers, ‘Most important, we need to redress the current bias towards policies supporting working women exclusively, at the expense of policies supporting fulltime homemakers and full-time parents’ (Hakim 2003a, p. 61; see further Hakim 2000, ch. 8).

Is it really an illusion?

101

to be respected; but being purely a matter of choice, the long hours they work ought to attract no particular public sympathy or subsidy.’ These are important challenges, which we devote this chapter to addressing. Doing so affords us the opportunity to draw attention to various other important aspects of our findings and how they might best be interpreted.

6.1 Blaming the victims Schor’s objection to ‘blaming the victim’ assumes, first and foremost, that the people concerned are indeed victims. In the sense relevant here, ‘victims’ are people (a) to whom something bad has happened and (b) for whom that was (largely) outside their own control.6 We can well imagine victims of that sort. Think for example of a mother of young children who, having been widowed (and lacking any other source of income), is forced to spend long hours in paid employment in order to support her family.7 It would indeed be churlish to deny her any public sympathy or subsidy on the grounds that she went out to work ‘of her own choosing’. She has no choice, or anyway no reasonable choice, to do otherwise. Other working mothers might be in broadly the same situation. Single mothers who have no other source of income obviously have to go out to work to feed their family. So too do wives of men who are unemployed or disabled or on incomes so low as to be insufficient to sustain the family. All those other sorts of working mothers, too, have ‘no choice’ but to enter the labour market, and it would again be churlish to pretend they do so of ‘their own choosing’. But what of the sorts of people who, on our analyses, suffer the greatest ‘time-pressure illusion’ – childless dual-earner couples? Are they victims in either of the relevant respects? Do they truly have

6

7

One might of course be said to be ‘a victim of one’s own folly’ or ‘carelessness’ or ’stupidity’. But those are things for which we think people should be (at least partly) blamed. Objections to ‘blaming the victim’ are primarily objections to blaming victims of the other sort. Under the influence of ‘welfare-to-work’ policies that have swept the world, even public benefits might be conditional upon her engaging in paid employment or actively preparing for it, at least if her children are school aged (US Congress 1996; UK Parliament 1999).

102

Time pressure

‘no choice’ in the matter? Are they social unfortunates, or relatively privileged?

6.1.1 Are they victims of external forces? Victims are people whose fates are imposed upon them by external forces rather than their own choices. ‘Overworked Americans’, Schor claims, are victims in that sense. They work long hours because their jobs require them to do so. ‘In contrast with freedom in the spending of the money we earn, the modern industrial regime denies us a similar freedom in choosing the work routine by which we earn those dollars,’ Schor says (quoting Paul Samuelson, no less).8 Apocryphally, being a corporate lawyer is not something you can do for only 10 hours a week: the job is 60 hours, or nothing.9 And if mothers want to enter the labour force, they have to conform to an ‘ideal worker norm’ of fulltime employment that implicitly presupposes a non-working partner taking care of the home and family.10 Let us be clear what we are claiming here, in contrast. We do not claim that people can work any number of hours that they choose in their present jobs, necessarily; if they want to reduce their hours in paid labour, they may well have to change jobs. Nor do we claim that it would be possible for everyone to reduce their hours in paid labour allat-once. Were they to do so, there would be consequences for the macroeconomy and government budget that we simply cannot hope to model; so we are just asking, in more limited fashion, what would happen if they changed their working hours one-household-at-a-time. For purposes of that more limited analytic exercise, we need only claim that: there are at least some jobs (and, to foreshadow our concluding chapter, there could therefore be more jobs) at every relevant wageand-hours combination; and hence that any given person could therefore find a job paying pretty much the same wage rate s/he presently enjoys for substantially fewer hours a week. That is what we claim to have shown, in section 2.2.2 above, at least apropos of high-waged

8 9

10

Schor 1992, p. 128. Landers et al. 1996. Although notice, even there there might be selection effects at work, with workaholics who enjoy those long hours self-selecting for the practice of corporate law. Williams 2000.

Is it really an illusion?

103

workers whom our calculations envisage working substantially different hours than at present. Schor’s contrasting claim is that people are generally forced to work the long hours that they do: they have no alternative. Logically, it is perfectly possible for her claim and ours both to be correct. All we strictly claim is that any given person, one-at-a-time, could change his or her number of hours in paid labour. That is perfectly consistent with a claim (akin to, if not identical with, Schor’s) that people in general could not do so all-at-once. Indeed, the evidence we have offered in section 2.2.2 above might be seen as supporting that latter claim as well as our own. There we have seen that, while there are some high-waged jobs presently being performed for only a very few hours a week, there are not many jobs presently being performed in that way. Any given high-waged person could occupy one of them – that is our claim. But there may not be enough of them for all high-waged people to find one. The temporal autonomy that we thus show is available to any given person in isolation may not be available to everyone in general.11 In our concluding chapter we will be making public-policy recommendations designed to close that potential gap. Just how big the gap is remains an open question, however. The evidence of section 2.2.2 is inconclusive on the crucial question of whether people in general are forced, or whether they merely choose, to work longer hours than would be required merely to keep themselves out of poverty. Schor wants to claim that it is more a matter of ‘force’ than of ‘choice’. But the evidence she offers in support of that claim is mixed at best. It consists almost wholly of surveys asking people about their ‘ideal’ or ‘preferred’ hours in paid labour, and comparing the responses to their actual hours in paid labour. Those surveys, however, admit of radically different interpretations, making them problematic in ways we shall discuss shortly. Before doing that, let us pause to consider what exactly could be inferred from such surveys, even if they were not problematic in those other respects. Suppose we find people systematically saying they preferred to work fewer hours than they actually do. Why should we 11

An analogous distinction underlies Cohen’s (1983) analysis of ‘the structure of proletarian unfreedom’.

104

Time pressure

necessarily privilege what people say over what they do, and conclude that they are working longer hours than they really prefer? Why not privilege what people do over what they say, concluding that they are working the hours that they genuinely prefer and that they are merely misrepresenting their true preferences in their verbal reports? There are at least two good reasons for thinking that people might understate their preferred working hours. One is that it is not socially respectable to say that you prefer to abandon your family for long hours – even if you do.12 Another is that being pressed for time is a badge of social status: it shows that you are in demand; it shows that you hold an important job.13 Either motive, or both, could lead people to understate their true preferred hours of work, making the difference between actual and (ostensibly) ‘preferred’ hours of work an unreliable measure. Furthermore, notice that people’s statements of the number of hours they prefer are unstable and vary depending on the way the question is put.14 Asked in the abstract whether they would prefer to work fewer hours, fully a third of Americans in 1998 responded that they would. Asked whether they would prefer to work fewer hours but earn less money, fewer than 10 per cent said they would. Asked if they would ‘be willing to give up a day’s pay each week for an extra day of free time’, a quarter said they would.15 Asked how much of a 10 per cent raise they would be willing to forgo for additional free time, nearly half of respondents said they would be willing to forgo it all.16

12

13 14

15

As Hochschild (1997; cf. Maume and Bellas 2001) speculates might actually be quite common. Gershuny 2005a. As Bowles and Park (2005, p. F401) justifying their own model’s ‘assumption that individuals choose their hours of work’ say: ‘Not surprisingly, a significant fraction of employees in the advanced economies would prefer hours different from what they have (Bell and Freeman 2001). However in the studies reported, a majority preferred current pay with current hours (rather than more hours and more pay, or less hours and less pay) and Bell and Freeman report evidence that most European Community workers would prefer increases in pay (at the current hours) to decreases in hours (at the current total earnings), suggesting that they are close to the hours they would have chosen even if the institutional setting allows no direct relationship between individual hours choices and outcomes. Bo¨heim and Taylor (2004) report similar results using the British Household Panel Survey.’ Jacobs and Gerson 2004, pp. 74–5. 16 Schor 1992, pp. 130–1.

Is it really an illusion?

105

Any way you look at that mixed set of findings, the upshot seems to be this: While a minority (a sizeable one on some question formulations; a small one on others) might prefer to work shorter hours, the majority of respondents are content with their present number of hours in paid labour (indeed, a few would prefer more). What to make of that finding, normatively, may well be another matter.17 But it is the more empirical aspects of the matter that concern us here. The issue here is whether people are generally forced to work longer hours, contrary to their preferences. On the survey evidence on preferred-versus-actual work hours, most are not. One further piece of survey evidence seems particularly telling in this context. In a 1997 survey, when people said they preferred to work fewer hours they were then asked why they did not do so. Much the most common response was that they needed the money. Less than a fifth said it was because their employer would not let them.18 Panel data from both the US and Germany confirm that, if people genuinely prefer to work fewer hours, they eventually will. These studies compare respondents’ ‘preferred work hours’ in their first interview with their actual work hours when the same respondents are reinterviewed 6 to 10 years later. Most of them are indeed working fewer hours. That was true of some 60 per cent of Americans in the National Study of Families and Households between 1987/8 and 1993/4.19 The German labour market is notoriously ‘stickier’ for ordinary employees. But among professionals (who predominate among those who suffer the largest ‘time-pressure illusion’, on our analysis), the German Socio-Economic Panel Survey found 45.5 per cent of those who said in 1985 that they preferred to work fewer hours were actually working fewer hours in 1994.20

17

18

19

They may be suffering false consciousness or adaptive preferences (Schor 1992, p. 129). Jacobs and Gerson 2004, p. 75. The further follow-up question in which we would have been most interested was not asked: ‘could they find some other employer who would let them work the hours they preferred (or is their preference for working fewer hours sufficiently weak as just not to be worth going to the trouble of changing jobs)?’ But the proportion who could not find any employer who would let them work the number of hours they prefer is obviously a fraction of the fifth of the workforce whose present employer will not let them do so. Clarkberg and Moen 2001, p. 1129. 20 Merz 2002, p. 331.

106

Time pressure

None of this evidence is conclusive of course. But all of it taken together is highly suggestive. The conclusion towards which it all points seems to be that – certainly for the sorts of people who experience the largest ‘time-pressure illusion’ on our analysis, childless dualearner couples – the number of hours they spend in paid labour is not necessarily externally imposed and beyond their own power to alter.21 They are not necessarily ‘victims’, in the sense of being victims of circumstances wholly or even substantially outside their control.

6.1.2 Are they badly off? A second aspect of ‘victimhood’ is that victims are badly off. Anyone who has something bad happen to them (provided it was outside their control) can count as a victim. But the victims who are of most concern to us are those who were badly off to start with, and who have been made still worse off. Can the people whom we find to be experiencing most ‘time-pressure illusion’ be properly counted as ‘victims’ in that sense? In chapter 5, the group of people we find having the greatest ‘time-pressure illusion’ are dual-earner couples. The ‘time-pressure illusion’ is particularly strong among dual-earner couples without children; but even among households with children, dual-earner couples suffer the ‘time-pressure illusion’ to a greater extent than any other parents. Their ‘time-pressure illusion’ is high for two reasons. First, dualearners almost invariably spend more actual time in paid labour than their counterparts in either of the other two household types. Second – and of particular relevance to the present discussion – dual-earners’ ‘necessary time in paid labour’ is low. Indeed, theirs is the lowest average ‘necessary time in paid labour’ among all the household types

21

The belief that long working hours are more a matter of ‘choice’ than ‘force’ is widespread in Australia. Drawing on a 2003 survey of the Australian workforce, van Wanrooy and Wilson (2006) report that some 50% of Australian workers agree with the view that most people who work long hours choose to do so (only 31% disagree, while the remainder neither agree nor disagree or else can’t choose). Tellingly, the belief that long working hours are a choice is even more prevalent among those who actually work long hours. Among Australian workers who usually work 60 hours a week or more, some 57% agree that most people who work long hours choose to do so (only 32% disagree, while the remainder neither agree nor disagree or else can’t choose).

107

125.68

127.55

97.57

98.41

76.75

74.80 Finland

120.47

Sweden

82.38 94.81

94.89 94.34

France

56.81

80

78.75

per cent

100

91.79

94.94

120

86.87 92.74

140

110.77

126.45

160

122.86

151.39

Is it really an illusion?

60 40 20 0

all countries

US

Australia

bottom third

Germany middle third

top third

Figure 6.1 Mean wage rates of prime-aged wage-earners with high, medium and low time-pressure illusion (as percentage of national mean)

for any group that is in paid labour at all.22 The reason ‘necessary time in paid labour’ is low, in turn, is primarily that the hourly wage rates of people in dual-earner couples are relatively high.23 In general, people experiencing the greatest ‘time-pressure illusion’ tend to have higher wage rates than others in their community. This is shown in Figure 6.1. That figure splits prime-aged wage-earners in each country into thirds, according to the extent of ‘time-pressure illusion’ that they suffer. It then displays the mean wage rate of that group, as a percentage of the mean wage rate for all prime-aged wage-earners in the country. Averaging across the six countries, the third experiencing the most ‘time-pressure illusion’ enjoys a wage rate on average oneand-a-half times that of the third experiencing the least ‘time-pressure illusion’. The magnitudes differ from country to country, but in every country that same basic pattern is observed: the most time-pressured

22 23

i.e., excepting the non-earner in one-earner couples. ‘Primarily’, because ‘necessary time’ in both paid and unpaid labour can also be affected by household size, which (for households with children) might vary from one household type to another.

108

Time pressure

are in general economically privileged by higher wage rates, not lower ones.24 That these people have a relatively high wage rate means two things, in terms of the rhetoric of ‘victimhood’. First, these people are in a relatively privileged position, economically. They are not the sorts of unfortunates we would feel most sorry for, should something further bad befall them. Second, given their wage rates, it is not all that clear that when working long hours in paid labour anything particularly bad has befallen them. They have less time available for other things, to be sure. On the other hand, they are compensated more handsomely than the average wage-earner for the time they spend on the job. Thus, on neither score does it seem to us that the people we have found to be under the greatest illusions about their time pressure can be convincingly characterized as ‘victims’ whom it would be wrong to blame for their plight. Any given person might of course be an exception.25 But in general, it is unclear that the sorts of double-income couples who are under the greatest ‘time-pressure illusion’ are being compelled by external forces to work long hours; and it is not even clear that working long hours is necessarily a terrible thing to happen to them, given the relatively high wage rates they command for every one of those extra hours that they work. We would conclude instead that, in general, they choose to work those hours, and given their wage rates perfectly sensibly so.26 24

25

26

It is over two-and-a-half in the US, only one-and-a-sixth in Australia. The pattern of those with the greatest time-pressure illusion having higher average wage rates becomes even more striking the more narrowly we define ‘greatest time-pressure illusion’: averaging across all six countries, people in the top third of that distribution have average wage rates 1.26 times the national average; people in the top quartile 1.31; people in the top quintile 1.35. Our way of calculating the time-pressure illusion makes it easier for high-wage-rate people to score high on that measure (because of their higher wage rates, they do not have to work as many ‘necessary hours in paid labour’ to secure a poverty-level income); but our procedures do not make that result inevitable (they will score high on the timepressure illusion indicator only if they go on actually to work many more hours than by that calculation is deemed ‘necessary’). It is also possible that dual-earner couples live disproportionately in big cities where expenses are high, and poverty lines calculated on the basis of nationwide averages misrepresent their true costs of living. But that raises larger issues with poverty-rate calculations in general, beyond the scope of this book to explore. As we point out in sec. 4.3, some economists have predicted that those with the highest wage rates would work the longest hours: after all, the opportunity cost of an hour’s leisure is highest to them (Becker 1965; Linder 1970, ch. 3).

Is it really an illusion?

109

6.2 Choice and the quality of options Among philosophers, ‘luck egalitarians’ draw a sharp distinction between conditions that people are in because of their own choices, and conditions that they are in because of circumstances beyond their control (‘luck’). Broadly speaking, luck egalitarians think that people ought to bear responsibility for the consequences of their own choices. In respect of outcomes that are the consequence purely of luck that is wholly outside the agent’s own control, however, luck egalitarians suppose that justice does indeed require us to intervene to eliminate (as best we can) the effects of sheer luck.27 The controversies over time pressure can be recast in those terms. People like Schor who would have us sympathize with the plight of people working long hours argue (among other things) that those workers have no choice in the matter. People like Hakim who would have us deny any particular sympathy to working mothers argue (among other things) that the choice of combining a career with children was indeed the woman’s own. We think that a genuine choice is (usually) involved in both cases. Luck egalitarians would say that makes them both ‘responsible’ for any time pressure they suffer in consequence. Unlike cruder luck egalitarians, however, we are not prepared always just to leave losses (however large) lying with the person whose choice was somehow responsible for them. At a minimum, there ought be some notion of proportionality here. When a careless driver’s minor mistake causes a major accident, the driver is in some sense responsible for the accident, but sheeting the full costs of the accident home to him would be vastly disproportionate to his blameworthiness. Furthermore, for people to be held fully responsible for the consequences of their choices, the choices must be made from among a fair and reasonable set of options. Saying to someone at gunpoint, ‘your money or your life’, does not shift responsibility for the robbery onto the victim’s shoulders, just because she ‘chose’ to relinquish her purse rather than her life. Finally, some choices are ones that society wants people to make and for which we think they should be socially rewarded. Soldiers engaging in heroic acts above and beyond the call of duty are not told that they 27

Dworkin 1981; Barry 1989; Arneson 1997.

110

Time pressure

‘have only themselves to blame’ when they are injured in the course of heroic dare-doing. Instead they are cared for by ‘a grateful nation’, honoured and rewarded. Bearing those thoughts in mind, the time pressure experienced by working mothers really does look very different from the time pressure that childless people might gratuitously inflict upon themselves. Motherhood is a socially valuable choice for a woman to make; it is crucial to the reproduction of society itself (as well as, more selfishly, to our own support in old age).28 At seminal moments in the development of social policy, analogies have indeed been drawn between the service to the nation of mothers and the national service of soldiers.29 Theirs clearly is a choice that should be socially rewarded. When facing the choice of whether or not to conceive, however, women are often confronted with an unfair and unreasonable set of options, in temporal terms. Either they forgo motherhood altogether; or else they suffer much more severe time pressure than their childless counterparts, and necessarily so (by something on the order of 7 hours a week less discretionary time, on our reckoning).30 Of course, many of the demands that young children put on parents’ time may not be something that we can do much to alleviate through social policy.31 But insofar as there are policies that might help, we ought to give them serious consideration. Foremost among them would be policies to give people (particularly but not exclusively mothers) the capacity to engage in paid labour part-time, if that is their preference.32 Can the same be said about dual-earner childless couples who are time-pressured only because they gratuitously choose to do far more work than really necessary? They, too, do socially useful labour of course. But they are well rewarded for it. As we have shown in section 6.1.2, the groups suffering the most gratuitous time pressure

28 30

31

32

Folbre 2008; World Bank 1994. 29 Lake 1992; Skocpol 1992. So too do fathers (albeit only half as much); and to the extent that they do our social concern ought to be with parents quite generally. Standard proposals of the sort offered by Phillips (1983) and Schor (1992, ch. 6) could help reduce actual time in paid labour, but not ‘necessary time in paid labour’; and virtually none of them addresses ‘necessary time in unpaid household labour’. Mothers particularly suffer from being held to the ‘ideal worker norm’ of fulltime work, which in effect presupposes a non-employed wife at home taking care of the children (Williams 2000).

Is it really an illusion?

111

have relatively high wage rates. What they choose among is, by and large, a perfectly reasonable set of options. It is in no way part of our argument to say that these dual-earner childless couples have somehow made the ‘wrong’ choice, in choosing to work far longer hours than strictly necessary. We merely say that they genuinely had a choice in that matter, from a perfectly respectable set of options. They therefore do not have the same call on public sympathy and support as others, such as parents (especially lone parents), who make their ‘choices’ from a much less attractive set of options and whose time pressure is, in any case, much more a matter of unavoidable necessity than of free choice.

6.3 Temporal neutrality State policies ought to be neutral among the many ways people might reasonably choose to organize their lives. That is Hakim’s central theme, but it is hardly hers alone.33 Philosophically phrased as ‘neutrality among alternative conceptions of The Good’, that is the central tenet of the dominant liberal creed itself.34 Most liberals are not prepared to adhere absolutely scrupulously to that policy, of course. Society as a whole has a strong interest in having someone produce the next generation. Parenthood is not a lifestyle choice about which societies (or most liberals, come to that) are usually prepared to be indifferent. It is something that is rightly encouraged and supported through public policy, and arguably it should be more so.35 Let us focus, however, on those lifestyle choices among which it seems that public policy genuinely ought to be neutral. In what does neutrality consist? On the luck-egalitarian argument, what we ought to ‘neutralize’ is people’s conditions of choice, not the effects of their choices. Luckegalitarianism says that, ideally, everyone should be given the same option set – the same choices, on the same terms and conditions. We should use social policy to ensure that, as far as we are able. People should then be allowed to make their own choices from among those options, and they should bear responsibility for the consequences of

33 35

Hakim 2000, p. 253. 34 Dworkin 1978. Esping-Andersen 1999, ch. 4; 2002, ch. 2; Folbre 2008.

112

Time pressure

their choices. Rectifying any inequalities that emerge at that level should not be the business of social policy. Expressed in temporal terms, that means that we should (insofar as we are able, which might be strictly limited) strive to equalize people’s ‘necessary time’ in the various necessary tasks of daily life, regardless of their social circumstances (household type, parenthood, etc.). How people then actually choose to exercise their discretion and spend that time should be left up to them; and they should bear responsibility for the temporal consequences of those choices. In short, luckegalitarian thinking suggests that social policy should be interventionist in egalitarian directions with respect to opportunities (discretionary time); but social policy should be impervious to differences in the actual amount of ‘spare time’ people have left over after the various choices they have made as to how to use their time. Such a policy would, it seems to us, be ‘neutral’ in the ways that matter. True, it would require us to give more help to some people than others, depending on their social circumstances. Specifically, it would generally require us to do more to temporally assist (insofar as we are able through social policy) parents in general and lone parents in particular. But such interventions would truly be ‘neutralizing’, rather than neutrality-violating. As part and parcel of that policy, of course, we should also strive to ensure insofar as possible that people have the option of working in paid labour any number of hours they choose. That may not be completely possible: some jobs might have to be done for a certain number of hours, or certain people might have to do the same number of hours as certain others (if they work on the same automated production line, for example). And most people will of course choose to work longer hours than the minimum necessary merely to achieve a poverty-level income. But if their temporal autonomy is to be maximized, the choice in this matter ought to be their own. We shall return to this theme in our concluding chapter.

PART III

Welfare regimes matter

7

How welfare regimes differ

It has long been a function of the state to promote the welfare of its people. How ‘welfare’ is conceived, and what the state is supposed to do to promote it, has, however, varied widely from place to place and time to time. After some general remarks on the conceptualization of ‘welfare’ (section 7.1), we offer a very short history of state efforts to promote people’s welfare (section 7.2). We then proceed to introduce the three standard ‘welfare regimes’ (section 7.3) and some standard indicators commonly used to situate countries in relation to them (section 7.4). We close with a reminder of ‘other dimensions of welfare’, things that clearly ought to be seen as part and parcel of the welfare state’s core activities but which are often left out of such discussions – pre-eminently, for our purposes, child-care policy (section 7.5).

7.1 Defining ‘welfare’ ‘Welfare’ has both objective and subjective dimensions. Subjectively, ‘welfare’ consists in a state of mind: happiness, or satisfaction. ‘Objective welfare’ – sometimes called ‘standard of living’1 – may or may not be related to that. Sometimes ‘objective welfare’ is taken to be simply a list of things that it is objectively good for people to have, whatever their subjective states now or later. That is a ‘hard paternalist’ line. There is, however, a ‘soft paternalist’ line that depicts ‘objective welfare’ as one’s capacity (or opportunity) to achieve those subjective states. When welfare-state theorists talk about ‘welfare’, it is one or other of those ‘objective’ senses that is being evoked. What the state distributes and redistributes through its so-called ‘welfare sector’ is not subjective states but rather material goods. Its rationale for doing so may well be 1

Johansson 1973; Sen 1987.

115

116

Welfare regimes matter

(and on the ‘soft paternalist line’, would be) that those material resources are believed to contribute causally to subjective states. But the welfare state, through its taxes, transfers and subsidies, affects people’s subjective welfare only indirectly. Welfare-state studies – like welfare policy-makers – often employ what is sometimes called an ‘objective list’ conception of welfare.2 There are certain things that are presumably ‘good’ for people (in either a ‘hard paternalist’ sense or a ‘soft’ one); and certain ‘threshold’ amounts of those things that people ‘need’ in order to thrive. Welfare policy-makers strive to ensure that everyone has at least those threshold amounts of those needed commodities, and commentators assess their performance accordingly. How we are to determine what people need, and how much of it they need, are hard questions. What is strictly required for biological functioning sets a lower limit. But welfare-state analysts and operatives both typically supplement that bare-bones standard of what is absolutely needed with more context-sensitive analyses of what is ‘socially needed’ in one’s own society.3 (Indexing the poverty line to ‘half the median equivalent income nationwide’, as in section 2.2.2, is one such relativization.) Welfare-state analysts sometimes specify what is needed in terms of a list of specific commodities. (That is the ‘basket of goods’ approach to defining poverty, also discussed in section 2.2.2.) More often, however, summary assessments of the overall performance of a country’s welfare sector are based on a monetary standard – what proportion of the population lives below the poverty line, defined in terms of income.4 In short, ‘welfare’ as it is assessed by welfare-state researchers and promoted by welfare policy-makers refers to people’s objective ‘opportunities’ and ‘resources’.5 A person’s welfare, thus construed, is a matter of the objective stock of needed resources that that person has, rather than in terms of what subjective satisfactions the person derives from the way in which those resources are deployed. Perhaps the best

2 3 4 5

Griffin 1986. Braybrooke 1987; Doyal and Gough 1991; Goodin 1988, ch. 2; 1990a; Sen 1985. Atkinson 1998. Perhaps (on the ‘hard paternalist line’) because they are good in themselves, perhaps because (on the ‘soft paternalist line’) they constitute opportunities for achieving welfare in the subjective sense (Arneson 1989).

How welfare regimes differ

117

summary statement comes from the magisterial Swedish Level of Living Survey. There, ‘welfare’ in its most general terms is defined as ‘the individual’s command over resources . . . through which the individual can control and consciously direct his living conditions’.6 Control over resources sufficient to live one’s life as one pleases – autonomy, in short – is thus, at root, the central concern of all studies of people’s welfare and of all policies aimed at promoting it.

7.2 Welfare policy: a potted history Modern social-welfare policy has its roots in ‘poor relief’. A religious duty in medieval Europe, poor relief was a function that states increasingly took upon themselves as they consolidated their authority. That was often done with ‘public order’ purposes in mind. The famous English Poor Law of 1601, like many others across Europe, was designed in no small part to put an end to vagabonds and itinerants, offering paupers public assistance on condition that they ‘settle’ in some particular parish.7 Power-and-order dynamics have governed welfare policy ever since. Bismarck’s reforming policies at the end of the nineteenth century were explicitly designed to ‘buy social peace’ and prevent the emerging working class from falling into the arms of socialist parties.8 The strings (‘conditionalities’) now attached to contemporary workfare policies serve to ‘regulate the poor’ even more fiercely than when Piven and Cloward coined that phrase.9 And everywhere, the concessions that capital makes to labour, through the ‘social wage’ (social-welfare benefits) as well as through pay and conditions, are everywhere a function of the relative power and resources available to each.10 Having noted and duly acknowledged the power dynamic, however, let us proceed to tell the history of welfare policy in the more principled and problem-solving terms that is the preferred mode of discourse in these realms.

6 7

8 9

10

Johansson, quoted in Erikson (1993, pp. 72–3). See further Johansson (1973). De Swaan 1988, ch. 2. The text of the Elizabethan Poor Law of 1601 is reproduced in Goodin and Mitchell (2000, vol. II, pp. 3–6). Rimlinger 1971, ch. 4. Piven and Cloward 1971. On workfare see, e.g., Goodin (2002; 2004); on precursors see King (1995). Korpi 1983.

118

Welfare regimes matter

That version of the modern welfare-state story starts, in effect, with the ‘Great Transformation’.11 With rapid industrialization, the problems of ‘unemployment’ came to loom large. Whereas the paupers of old were typically the ‘impotent’ – those too old or young or disabled to support themselves through their own labour12 – the industrial-era poor were increasingly ‘able-bodied’, who were simply out of work. Poor relief originally aimed primarily at the former gave rise to perverse incentives as applied to the latter. Writing in 1835, Tocqueville reported his ‘indescribable astonishment that one-sixth of the inhabitants of this flourishing kingdom [of England] live at the expense of public charity’, concluding that ‘the inevitable result of public charity was to perpetuate idleness among the majority of the poor and to provide for their leisure at the expense of those who work’.13 The policy response, common across Europe but finding fiercest expression in Britain, was the policy of ‘lesser eligibility’: deliberately making living on public assistance a less attractive option than the least attractive job available in the community.14 This was done in two ways. One was to ensure that public assistance paid less than any job. (‘To promote industry and economy it is necessary that the relief given to the poor be limited and precarious’, as one contemporary put it.15) The other plank of the ‘lesser eligibility’ strategy was to increase reliance on a regime of ‘indoor relief’, targeted particularly at the ‘undeserving’ able-bodied poor who were forced into residential ‘workhouses’ if they wanted public assistance.16

11

12

13 14

15

16

What follows is essentially the story of Polanyi (1944) and Wilensky and Lebeaux (1958). In the words of the English Poor Law of 1601: ‘Lame, Impotent, Old, Blind, and such other among them, being Poor, and not able to work’ (Goodin and Mitchell 2000, vol. II, p. 3). Tocqueville 1835/1983, pp. 105, 114. This is the underlying principle of the ‘New Poor Law’ enacted by the Poor Law (Amendment) Act of 1834. For contemporaneous discussions, see Blomfield et al. (1834) and Tocqueville (1835); for more recent assessments, see Polanyi (1944) and Blaug (1963). Rev. Joseph Townsend, quoted in Goodin (1988, p. 229). See similarly Malthus (1826), the Poor Law Commissioners (Blomfield et al. 1834) and Tocqueville (1835). The institutions were not themselves new; provision for workhouses is made in the first clause of the Poor Law of 1601. The 1834 amendments merely promoted greater use of them.

How welfare regimes differ

119

This is essentially the policy inheritance of Anglo-Saxon countries, Britain and its former colonies. There have of course been a great many overlays, since. Basically, however, people in those countries are supposed to be responsible for supporting themselves through paid labour, if they can.17 If they genuinely cannot, the public will support them. But public support is not on particularly generous terms, and the support is strictly means-tested. That is the essence of the ‘liberal welfare regime’, which we will abstract from this history in section 7.3 below. A later response to the risks posed by industrialization came through schemes of ‘insurance’, first ‘mutual’ and then ‘public’. The basic premise of insurance is risk-pooling: everyone involved pays premiums into a pool that is then used to pay compensation to anyone who suffers the contingency being insured against (unemployment; fire; sickness; accidental injury, incapacity or death; or whatever). Such risk-pooling was originally arranged privately, through voluntary ‘mutual benefit societies’ (‘Friendly Societies’ in Anglo-Australian terms).18 States first began sponsoring, regulating and underwriting those arrangements in earnest in the 1880s. The most famous example was Bismarck’s German reforms, which soon spread across the industrialized world. Almost simultaneously in France, ‘the concept of e´tat providence arose under the second Empire, fostered by Le Play and Napoleon III in the form of expanded mutual insurance plans (mutualite´)’.19 Bismarck’s own intentions were clearly conservative. He wanted the Kaiser to be seen as the protector of the working classes in order to block the rise of socialist parties. Structurally, his reforms operated through essentially feudal guilds (‘corporate associations’, on the principle of Soldaten der Arbeit).20 In addition to reinforcing existing macro-power structures in that way, Bismarckian social insurance – paid as it is on the basis of employment history and employment-based contributory insurance premiums21 – reinforced existing micro-power structures within the family, entrenching the position of the male breadwinner.22 17 18 19 20 21 22

Goodin 1988, ch. 12; 1998; Fraser and Gordon 1994. De Swaan 1988, ch. 5. Ashford 1991, p. 32; Rosanvallon 2000, ch. 1. See, e.g.: Rimlinger 1971, chs. 4–5; Esping-Andersen and Korpi 1984. In Titmuss’ (1974, p. 31) terms, ‘industrial achievement’. As feminist critics complain; this is discussed more fully in section 10.3 below.

120

Welfare regimes matter

Beyond those particularities of the early German case, there is something in the nature of social insurance more generally that makes it deeply conservative. When insurance premiums are earnings-related and the insurance payouts are likewise, social insurance serves to preserve the existing status order. Its explicit aim is to ‘stabilize expectations’ by replacing lost earnings when people are out of work, at some proportion (higher or lower, depending on the details of the country’s scheme) of the earnings they enjoyed when they were in work.23 Workers with higher incomes pay higher social-insurance contributions, but they get higher social-insurance payouts. What the state takes with one hand it gives back with the other (perhaps in some later period, the welfare state acting as ‘piggybank’24). In one way, that might look like pointless ‘fiscal churning’.25 But in another way, these earnings-related aspects of social insurance constitute a profoundly important contribution to preserving the social pecking order as well as individuals’ peace of mind.26 This is the fundamental policy inheritance of the countries of continental Europe. Of course, other currents have also shaped social policy there as elsewhere. But whereas the Anglo-Saxon countries are fundamentally liberal-individualist, the countries of continental Europe basically see people as being embedded in networks of social relations, centring primarily on the family and the workplace. Those are seen as loci of mutuality, linking people in relations of mutual support. Risk-pooling through social insurance is the natural extension of those relationships, in the social-policy sphere. It helps preserve not only the natural networks of mutuality but also the ordered hierarchies 23 26

Goodin 1990b. 24 Barr 2001. 25 Palda 1997. More generally, actuarially fair insurance with a 100 per cent replacement rate would serve simply to transform risks into certainties (at appropriately probabilistically discounted rates) and average lifetime income into annual income flows (Goodin 1982, ch. 8; 1988, pp. 155–71). There would be redistribution of a sort: from your earning years to your non-earning years; from those who got lucky and did not suffer a workplace accident to those who were unlucky enough to suffer one (Zeckhauser 1974). There are good reasons for people to welcome redistribution of that sort: reducing life’s uncertainties in just such ways is why people buy insurance, after all. But actuarially fair insurance with a 100 per cent replacement rate would not produce any redistribution of the sort that egalitarians talk about, when talking about redistribution. Those who are badly off on average just become badly off continuously, and those who are well off on average just become well off continuously, thanks to insurance of this sort.

How welfare regimes differ

121

within them. That is the essence of the ‘conservative-corporatist welfare regime’ abstracted from this history in section 7.3. A final policy response to problems of industrialization is the egalitarian response, based on solidarity not just within family or occupational groups but across society as a whole.27 This was the response of the late-industrializers in Scandinavia. Instead of linking social benefits to a previous record of workplace-based insurance contributions, entitlement there is universal and funded by general taxation. Instead of making social benefits a function of your previous earnings, benefits there are flat-rate, the same for everyone. ‘Why’, right-wingers like Enoch Powell might ask, ‘should any social service be provided without test of need?’28 ‘Because we are all part of the same community’, would be the solidaristic Scandinavian response. This solidaristic response sees society as a single-status moral community. Social benefits are not a public largesse to be dispensed at the discretion of state officials after some test of character or means. They are not some contractual entitlement, earned through membership in and contributions to some occupationally based insurance scheme. Instead, on this world-view, they are rights of citizenship which should be bestowed equally and unconditionally on all citizens.29 This is the policy inheritance of Scandinavian countries. Its lofty ambitions are well expressed in the text of the Swedish Social Services Act (Socialtja¨nstlag) of 1982: Public social services are to be established on the basis of democracy and solidarity, with a view to promoting economic and social security, equality of

27

28 29

Baldwin 1990, chs. 1–2. Perhaps the most eloquent statement of this position is Le Chapelier’s 1791 speech in the Constituent Assembly of Revolutionary France, arguing against mutual benefit societies: ‘The bodies in question have the avowed object of procuring relief for workers in the same occupation who fall sick or become unemployed. But be no mistake about this. It is for the nation and for public officials on its behalf to supply work to those who need it for their livelihood and to succor the sick . . . It should not be permissible for citizens in certain occupations to meet together in defense of their pretended common interests. There must be no more guilds in the state, but only the individual interest of each citizen and the general interest. No one shall be allowed to arouse in any citizen any kind of intermediate interest and to separate him from the public weal through the medium of corporate interests’ (quoted in Goodin 1988, p. 74 n. 1). Quoted in Titmuss 1955/1987, p. 40. Marshall 1949; King and Waldron 1988.

122

Welfare regimes matter

living conditions, and active participation in community life . . . Social services are to be based on respect for the self-determination and privacy of the individual.30

This is the essence of the ‘social-democratic welfare regime’ abstracted from this history in section 7.3.

7.3 Three welfare regimes The history of any given country’s welfare policies is inevitably far more complicated than can be conveyed in so short an overview as we have just offered.31 So too are their institutional embodiments far more complex than can be conveyed in a handful of generalizations about ideal-typic welfare regimes. Country specialists probe those differences in depth, and express often well-warranted scepticism about ideal-type regime generalizations. Comparative welfare-state researchers, however, stand back from those fascinating matters of fine detail, to try to get a better grip on the ‘big picture’. They talk in terms of the ‘three worlds of welfare capitalism’.32 As we acknowledged earlier (section 1.4), there are important variants within each, and no country fits any of them perfectly. Still, as we have said, the ‘three worlds’ are settled, fixed points for orienting comparative welfare-state studies. And they fit the main countries of the developed OECD world well enough to be genuinely interesting. The liberal regime, exemplified by the US and (arguably less well) Australia, is a residualist welfare regime.33 The liberal regime relies mainly upon the capitalist market economy for promoting people’s welfare. The state is relegated to a ‘nightwatchman’ role. ‘Poor relief’ and its modern equivalent, ‘social assistance’, is regarded as public 30 31

32

33

Quoted in Korpi 1990, p. 4. For one interesting example, many of the earliest universalist, flat-rate, generaltax-financed schemes of the sort we today associate with social democrats in Scandinavia were supported (and in Finland, actually introduced) by agrarian parties. For subsistence farmers with small flows of income in the formal economy, earnings-related insurance schemes predicated on a history of workplace contributions held no charm. See Baldwin 1990, pp. 83–94; Kangas 1992, pp. 31–5; Therborn 1995, p. 90. Esping-Andersen 1990. Cf. Titmuss 1974; Castles 1998; Goodin et al. 1999; Goodin and Mitchell 2000, vol. II. In the terms of Titmuss (1974, p. 30). The case for Australian exceptionalism is discussed at the end of sec. 7.4 below.

How welfare regimes differ

123

charity, not as a rights-based entitlement. Certain people such as the old, the young and the disabled are excused from paid labour as a matter of public policy, and they are paid ‘categorical benefits’ sometimes at generous rates. But as regards the ‘undeserving poor’ – people who liberals think ought to be expected to earn their own keep – liberals are much more miserly. Welfare benefits are targeted tightly on the poor, and they are paid at a rate only barely adequate to alleviate the worst distress, precisely to avoid creating disincentives to paid labour. The corporatist regime, exemplified by Germany and Austria and to lesser extents France and other countries of continental Europe, is a ‘conservative-corporatist’ welfare regime.34 It is ‘corporatist’ in seeing society as a cooperative venture in which various social groups each have their distinctive roles to play: labour and capital, men and women, and so on. Gender roles are particularly central to corporatist thinking, with men being breadwinners and women mothers and homemakers (as will be discussed in more detail in chapter 10). The corporatist regime is ‘conservative’ in employing public policy to reinforce those roles, thus underwriting social cohesion and social stability. Corporatists make welfare benefits earnings-related precisely with a view to making them status-preserving. Thus, corporatist regimes give back to people in benefits roughly what they take from them in taxes, sometimes in any given period (‘churning’) but especially over the life cycle (the corporatist welfare state serving as a ‘piggybank’).35 The social-democratic regime, exemplified by Sweden and Scandinavia more generally, is highly egalitarian. Politics is organized around class lines and the economy around socialist principles.36 Macroeconomic policy promotes high levels of employment and a compressed wage distribution; taxes are steep and welfare benefits generous, typically universal and often flat-rate. As will be further discussed in chapter 10, the social-democratic regime is also femalefriendly, bringing women into the paid-labour force (often via public employment) fully on a par with men. These standard differences between welfare regimes are effectively summarized by Gøsta Esping-Andersen, the pre-eminent theorist of welfare regimes, in a table we reproduce here as Table 7.1. Various technical terms that appear there will be explained as our story unfolds. 34

van Kersbergen 1995.

35

Palda 1997; Barr 2001.

36

Korpi 1983.

124

Welfare regimes matter

Table 7.1. Welfare regime characteristics

Role of: family market state Welfare state dominant mode of solidarity degree of decommodification (rights-based entitlements) Exemplars

Liberal

Social democratic Corporatist

marginal central marginal

marginal marginal central

individualist universalist minimal

maximal

US

Sweden

central marginal subsidiary kinship/corporate group high for breadwinner

Germany

Source: Esping-Andersen 1999, p. 85 (truncated).

It bears emphasizing that some features are common across all three of these welfare regimes. Each, for its different reasons, shares a concern for the welfare of those who are especially disadvantaged, and strives to provide some kind of social safety net as a result. This is seen by liberals as a matter of ‘poor relief’, by corporatists as a way of manifesting and promoting ‘social cohesion’, and by social democrats as an expression and instrument of ‘social equality’. Whichever the rationale, redistributing towards the bottom (to a greater or lesser extent) is a common feature across all three regimes. So too is a concern with freedom and autonomy, although once again the meanings of those terms vary. The freedom liberals promote is the ‘negative liberty’ of free markets: freedom from purposive intervention by particular others in one’s affairs. What liberals see as ‘freedom to choose’ socialists deride as ‘freedom to lose’.37 What social democrats promote is not ‘freedom from’ but rather ‘freedom to’, by providing people with the resources that would allow them to actually implement their preferred choices. Corporatists see freedom in more Hegelian terms, in which people realize their true nature as fundamentally social beings living in organic groups (first and foremost, the family). 37

Cf. Friedman and Friedman 1980 and Roemer 1988.

How welfare regimes differ

125

7.4 Standard ways of classifying countries In trying to ‘fit countries to regimes’, there are various standard indicators that are typically employed. The first and most basic is simply the ‘welfare effort’ that the country makes. Liberal welfare regimes are supposed to be low-spenders; socialdemocratic welfare regimes high-spenders; and corporatist regimes medium-spenders. Illustrative data are found in the first column of Table 7.2 displaying each of our six countries’ total (non-health) social expenditures as a percentage of GDP. Clearly, on the criterion of ‘welfare effort’ – what proportion of GDP they devote to public social expenditures – the six countries under study do indeed fall into the three sharply distinct regime clusters with which they are typically associated. The two liberal regimes (the US and Australia) devote around 10% of GDP to social expenditures. The two corporatist regimes (Germany and France) devote around 20% to it. The two social-democratic regimes (Sweden and Finland) devote around 30% to it. Those are very big differences indeed. Venerable as it is, that simple ‘welfare effort’ measure has been subject to many criticisms, many of them eminently justifiable. There may be other more effective ways of promoting welfare than simply Table 7.2. Classifying countries into welfare regimes

Liberal welfare regimes US Australia Corporatist welfare regimes Germany France Social-democratic welfare regimes Sweden Finland

Welfare effort

Average universalism

9.79 10.65

54 33

18.63 21.45

72 70

31.84 28.49

96 88

Source: Welfare effort (non-health social expenditure as % of GDP, c. 1993), OECD 1996, table 1.23, p. 20. Average universalism (average of the proportion of population aged 16–64 eligible for sickness benefits, unemployment benefits and pension benefits c. 1980), Esping-Andersen 1990, table 3.1, pp. 70–1.

126

Welfare regimes matter

spending money. Just counting how much money we spend wrongly equates inputs with outputs.38 Different programme structures can cause some countries to rank artificially high or low. Much of the more recent work of comparative welfare-state researchers has focused on that latter issue, on unpicking details of programme structure and basing comparisons on that.39 Perhaps the most influential such work is that of Gøsta Esping-Andersen. As his preferred measure, he constructs a ‘de-commodification index’ that is a complex function of several variables at once, making it hard to unpack. A better intuitive grasp of what is going on can be adduced by just looking at one of the major components of that index in isolation.40 Here we focus on the indicator of the ‘universality’ of entitlement to social benefits. Social-democratic welfare regimes should (in theory) be keen on high degrees of universality; to liberal regimes, universality should be anathema; and the corporatist regimes should lie somewhere in between. As we see from the second column of Table 7.2, our six countries once again line up just as conventional wisdom about welfare regimes leads us to expect. The statistic there reported is what proportion of the working-aged population, on average, is eligible for sickness benefits, unemployment benefits and pension benefits. While there is some variation within regimes (and the figure for Australia might be artificially low through an artifact of coding), the general picture is that the ‘universality’ of those benefit programmes is about half as high in the two liberal welfare regimes as it is in the two social-democratic ones, with the two corporatist regimes once again falling in between. All these cross-national classificatory schemes inevitably elide important facts about particular countries. Ideal types, by their nature, necessarily abstract from existing social reality. No country perfectly fits any of the ideal types, and some countries fit less well than others.

38 39 40

See Goodin et al. 1999, pp. 80–2 and references therein. See, e.g.: Korpi 1989; Esping-Andersen 1990; Palme 1990. Much the same pattern emerges for our six countries looking at the ‘decommodification index’: the two liberal countries cluster at the bottom (US 13.0; Australia 13.8), the two corporatist ones cluster in the middle (Germany 27.7; France 27.5); and the two social-democratic ones cluster at the top (Finland 29.2; Sweden 39.1). The order is ‘right’, but with the de-commodification index Finland is closer to Germany and France than it is to Sweden. See EspingAndersen 1990, p. 52.

How welfare regimes differ

127

So, for example, the sort of ‘status-preserving social insurance’ that welfare-regime models characteristically associate with corporatist regimes is found in the unemployment, sickness and accidentalinjury policies of every country under study. In terms of pension policies, it is as characteristic of our liberal countries as it is of our corporatist; and there are even components of it (now, if not originally) even in the social democracies under study. Or again, the ‘universal entitlements’ characteristically associated with social-democratic regimes are actually in the sorts of family allowances or child benefits that are found virtually everywhere in the industrialized world (the US apart). The devil is often in the detail of how exactly programmes are set up and of the context in which they operate. For an example of the former, focus on contributory old-age insurance. We typically think of that as ‘conservative’ – reproducing the same income inequalities among pensioners as prevailed among them in their earning years. But contributory state pensions can actually be redistributive if (as typically they are) they are income-capped (so that people continue paying contributions based on income above that, but earn no higher pension in consequence).41 Or, for another example, there could be a world of difference among means-tested programmes: a high means test that essentially just excludes the rich from benefits (as in Australia) is vastly different from a low means test that excludes everyone except the very poor. For an example of how context matters, consider how in an earlier period Australia’s labour-market institutions interacted with its welfare-state institutions. Australia’s welfare regime was distinctively ‘liberal’ – poorly funded, means-tested and all the rest. But its labourmarket institutions were, historically, more social-democratic, with a strong system of wage arbitration. If unemployment is low and wages are adequate, a generously funded welfare state is less urgently needed.42 Those labour-market institutions eroded over the period under study, and have now been essentially eliminated; so the particulars are, from the present perspective, of purely historical interest.43 41 42

43

US SSA 2006. Hancock 1979; Castles 1985; Castles and Mitchell 1992. Esping-Andersen (1999, pp. 89–90) accepts the point, noting however that changes in Australia and New Zealand in this respect render it now irrelevant. Castles 1996; 2001.

128

Welfare regimes matter

Still, that remains a salutary reminder of how cross-national generalizations can often go awry if they are insufficiently sensitive to national contexts.

7.5 Other dimensions of welfare: child care, for example Comparative welfare-state researchers are notoriously narrow in their foci. They focus much more commonly on income-maintenance programmes than on service-delivery programmes. Originally, they focused upon old-age pensions almost exclusively. Critics rightly complained that the sharp contrasts among countries that give rise to their talk of ‘welfare regimes’ only ever really applied to old-age pension policy, where some Scandinavian countries (and even then, only in some periods) truly did provide flat-rate universal benefits, while other countries truly did organize their pensions on a contributory-insurance basis, while others offered pensions or supplements that truly were meanstested. Critics of the narrow focus on income-maintenance policies go on to point out that, by any reasonable standards, much other state activity is oriented at promoting people’s welfare – even relatively narrowly conceived – and ought to be taken into account in any proper discussion of the welfare regime. Minimally, it is standardly said, we ought to take into account health, education and housing policy.44 We may well want to look at each of those domains of policy separately, for they may well be subject to different social forces and display differing patterns of development within the same country. Still, they are all surely part of the ‘broader welfare regime’. So too are ‘active labour-market policies’ of education and training, as their Scandinavian advocates have long insisted and contemporary practitioners of ‘workfare’ now agree.45 In those recent discussions of how to ‘activate’ people who are presently outside the labour force, the major focus has been on policies designed to ‘make work pay’.46 There are many aspects to that, of course. But one of them, aimed particularly at people with young children, is the provision of affordable child care to enable them to go out to work. 44 45 46

As do, e.g., Heidenheimer et al. 1990. Calmfors and Skedinger 1995; Moene and Wallerstein 1995. OECD 1997b.

How welfare regimes differ

129

We shall here focus on that issue, as a token of all those other welfare-state activities that a full analysis ought to take into account. We do so not because child-care subsidies are financially larger than education or health-care subsidies, for example, and certainly not because they are more important to the quality of people’s lives. Instead we focus here on child care merely because the cost and state subsidization of it is most directly related to the other issues we are here discussing. The cost of caring for children while you are at work reduces your net take-home pay in just the same way as taxes do. In assessing the costs (which we will, once again, be converting into timecosts) of working, child-care costs are an important component. There is no neat regime mapping of state policies on the provision or subsidization of child care for young children.47 Once the children hit school age, of course, all countries provide free state schooling for them – which doubles up as child-care provision during school hours, as well. In some countries there is a free, universal system of public preschooling. In France the e´cole maternelle and in Germany kindergarten is provided in this way half-days, and both are widely utilized. In other countries, the state offers support on a means-tested basis for preschool-aged children to participate in ‘early learning and development’ programmes. The US Head Start programme is a prime example of that model. Sometimes in addition, sometimes instead, states also provide families with financial support to assist them in paying for child care. All of the states under study (the US apart) provide child benefits or family allowances, of course; and those funds can be used to pay childcare costs, but are not restricted to that. Beyond that, the state can assist parents with children in paid care with rebates or tax credits (as in the US and Australia, in the period under study48); alternatively, the state can pay the rebate directly to the provider of the child care (as was also done in Australia, in the period under study). Finally, the state can assist parents with the costs of child care by ‘paying parents to stay home’, and provide the child care themselves. In the period under discussion, this was done in both Finland and Sweden 47

48

Morgan (2006) persuasively argues that differences between countries in this regard are traceable instead to historic patterns of church–state relations and conflicts over religion, feeding into ideologies about gender roles and the family. Whiteford et al. 2001, p. 30; Dapre` 2006, p. 553.

130

Welfare regimes matter

(in a very short-lived policy that just happened to coincide with the time-use survey we are using: this policy was introduced by a nonsocial-democratic regime and promptly immediately repealed by Social Democrats upon their return to government). Certain approaches to subsidization of child-care costs might more naturally be associated with some welfare regimes (and, to anticipate chapter 10, some gender regimes) than others. Liberal welfare regimes try to make people support themselves through paid labour, so a tax allowance reducing the tax levied on earnings would be a natural way for them to proceed (as is, of course, means-testing support for subsidized places in pre-schools). Corporatists, attached to traditional values, most naturally want to support stay-at-home mothers – and to do so in such a way that integrates the next generation into the national traditions. Social democrats, being pro-natalist and femalefriendly in equal measure, want to support mothers in their many forms but also to facilitate female labour-force participation by subsidizing the child-care costs of working women.49

49

Gornick and Meyers 2003; Gornick et al. 1997.

8

A temporal perspective on welfare regimes

‘Welfare’ as it is conceived in welfare-state studies is an objective measure of people’s ‘opportunities’, cashed out in turn in terms of ‘control over resources’. Minimally, the aspiration of this chapter is to get ‘time’ included among those ‘resources’ in such discussions. Indeed, the larger aspiration is to get ‘time’ included high on that list, fully on a par with income itself. Ordinary welfare-state studies focus heavily on people’s income streams, and the adequacy of those for meeting their needs. Time is just as important, however, for all the reasons that we gave at the outset: time and money are conjoined in the production function for welfare, just as labour and capital are conjoined in the production function for commodities; it takes money to buy goods, but it takes time to consume them. Having discretionary control over time is thus as important as having discretionary control over money, when it comes to producing subjective welfare. Not only are there those a priori reasons for considering time and money co-equals in promoting wellbeing. We also have corroborating empirical evidence from ‘happiness’ studies, reported in section 2.3.2. This chapter begins with a few conceptual remarks on how ‘welfare’ is measured in welfare-state studies, drawing attention to the strong analogies between ordinary measures of ‘income’ and our ‘discretionary time’ indicators (section 8.1). We then proceed to report the impact of governments’ ‘welfare-state sector’ on people’s discretionary time, first across each country’s population as a whole (section 8.2) and then focusing upon particular subgroups. There, we look at the welfare-state impact on the discretionary time of parents, both in general (section 8.3) and disaggregated according to partner status, whether they are lone parents or in dual-earner or single-earner couples (section 8.4). We conclude with a discussion of the impact of each welfare regime on the population subgroup of most particular concern to that regime: lone parent, as the poorest of the poor, for poor-relief-oriented 131

132

Welfare regimes matter

liberal regimes (section 8.5.1); parents in general, for pro-natalist socialdemocratic regimes (section 8.5.2); and stay-at-home parents for traditionally minded corporatist regimes (section 8.5.3). Discussion of gender differences is reserved for part IV of the book, where they get three chapters all their own.

8.1 Welfare measures: money and time When welfare-state researchers try to assess people’s welfare, they standardly look at their income. They do so in full knowledge that it is not a perfect proxy for their welfare. How much welfare people derive from their income obviously depends on how they spend it – and, indeed, on whether they spend it at all. Income is, at most, a measure of ‘potential welfare’.1 A millionaire who is so miserly that she ends up starving is nonetheless counted as ‘rich’, however poor her diet. She is counted as ‘rich’ by virtue of her ‘command over resources’, regardless of her actual consumption of resources. And rightly so, for purposes of social policy (if not for purposes of physiological or psychological assessments). Being ‘richer’ in time terms increases your ‘potential welfare’ in ways strictly analogous to the ways in which being richer in money terms does. Of course, as with income so too with time, how much welfare people derive depends on how they spend it. But whether one is counted as ‘rich’ or ‘poor’, in terms of time just as in terms of money, ought (as we have said in chapter 5) to be seen to depend on one’s ‘command over resources’, not on one’s ‘consumption of resources’. The focus of this book is on people’s command over the particular resource of ‘time’.2 The conventional time-use category of ‘spare time’ is a function of how much time people actually spend in paid labour, unpaid household labour and personal care. That makes ‘spare time’ more akin to a measure of one’s ‘consumption’ of the resource of time. ‘Discretionary time’, in contrast, is a function of how much time people strictly need to spend in those activities. Measuring as it does the extent to which their allocation of time is not dictated by strict necessity, ‘discretionary time’ is thus an indicator of people’s ‘command’ over the resource of time. 1 2

Ringen 1988. With the emphasis very much on ‘command’ (Schwartz 1975).

A temporal perspective on welfare regimes

133

Public policies that increase people’s ‘discretionary time’ increase their ‘temporal autonomy’.3 They provide people with opportunities to use time in other ways of their choosing, instead of ways in which they otherwise would have had to in order to meet the necessities of daily life. There is nothing that anyone, the state included, can do to ‘mint more time’. There is, however, much that the state can do to ease or exacerbate time pressures that people would otherwise suffer. Of course, what the state transfers to people is not time itself, directly. Rather, it transfers money, goods and services, which have the effect of altering the amount of time that people would otherwise have to spend in pursuit of those things. The more the state taxes people, the more hours people have to spend ‘working for the government’. Conversely, the more the state subsidizes people’s necessities, the more time it frees up for people to use in ways other than earning money to pay for them.4 No politician or social planner intentionally targets ‘discretionary time’ (measured exactly as we do, anyway) as an explicit policy objective. What social-welfare policy-makers do nonetheless has clear implications for people’s temporal autonomy, in ways that we will here be measuring. Assessing the impacts of those public policies in temporal terms is perhaps the most revealing way of judging the actual effects of such policies on the daily lives of real people.

8.2 Differing state impacts on temporal welfare: the big picture Our focus in this chapter is on the difference that government makes to people’s discretionary time and how that differs from one country to another and for various different subgroups in the population. Empirically, our indicator of ‘state impact’ will be the difference between people’s ‘post-government discretionary time’ and their ‘pregovernment discretionary time’. It bears emphasizing from the outset, however, that we are here only attempting to capture certain very particular activities of governments in these ‘pre-’ and ‘post-’ comparisons. Specifically, we will only be 3 4

Goodin 2001; van der Veen and Groot 2006. See similarly Van Parijs 1995. Tax critics quoted in sec. 1.2.3 emphasize the former proposition while ignoring the second.

134

Welfare regimes matter

looking at effects that governments have through their taxes, transfer payments and child-care subsidies. Those are the only things we can explore through the LIS data set. ‘Post-government discretionary time’ is akin to ‘net disposable income’ in ordinary welfare-state poverty research. It is the amount of discretionary time that people have left to them, after paying taxes and receiving state transfer payments and child-care subsidies.5 More precisely (because, remember, discretionary time is calculated on the assumption people work only long enough to earn a poverty-level income), it is how much time they would have available after paying the taxes and receiving the transfers and child-care subsidies that they would were their income at the poverty line. Of course, governments do many other things that powerfully influence how much discretionary time people have. Macroeconomic policy affects wage levels, and typically more for some groups of people than others. Labour-market policy affects employment rates, again more or less differentially across the population. Industrial relations policies lead to more or less wage compression. Anti-discrimination policies affect the disparity in wage rates between races, genders and ethnic groups. Transportation and planning policies affect how long people have to spend getting to work and how long it takes to do their shopping. On our procedures, however, all those influences show up in ‘pre-government’ time, necessary and discretionary. That label should therefore be taken with a pinch of salt. Remember, every time you see it, that what it really means is just ‘before government taxes, transfers and child-care subsidies’ – and that alone. We emphasize that fact because ‘pre-government discretionary time’ of that sort actually differs, rather substantially, across countries. We see that clearly by comparing the left-hand columns in Figure 8.1 across all our countries. As we said apropos of a similar diagram in chapter 3, the scale of Figure 8.1 makes the differences look smaller than they really are, in absolute terms. Just look at the actual numbers. ‘Pregovernment discretionary time’ in Finland is fully 8 hours a week higher than it is in France. That is as if the French had to work a day a week more than Finns: a big difference. 5

It is not just ‘net disposable income by another name’, though: remember, the conversion between money and discretionary time is a complex, non-linear function as described in chapter 2.

85.10 83.80

84.39 85.22

84.75 84.51

135

77.03 76.08

US

80.67 78.58

liberal countries

78.56 78.18

80

82.04 80.34

90

80.30 79.26

100

84.31 81.08

A temporal perspective on welfare regimes

hours per week

70 60 50 40 30 20 10 0 Australia corporatist Germany France countries pre-government

social- Sweden Finland democratic countries

post-government

Figure 8.1 Pre- and post-government discretionary time, whole population

As we have said, a lot of that cross-national variation may be due to other aspects of government policy that the LIS data set does not allow us to explore. Insofar as it is, the variation across countries in ‘pregovernment discretionary time’ is really just variation in ‘unmeasured post-government discretionary time’. And it is worth noting that ‘pre-government discretionary time’ is indeed highest in the two socialdemocratic countries (Finland and Sweden) famous for ‘progressive’ policies in the macroeconomic, labour-market and anti-discrimination spheres. Another part of the cross-national variation in ‘pre-government discretionary time’ is due to wider societal influences. Being a more egalitarian society, for example, Finns and Swedes spend (and by our methodology for specifying ‘necessary time’, therefore also need to spend) less time preening before the mirror each morning. The French spend (and by our methodology, therefore need to spend) more time over meals. These differences, while not due to government policies as such, have much to do with social ethos of a more diffuse sort – the sort of thing sometimes captured in talk of ‘welfare society’ or ‘a culture of equality’, in contradistinction to ‘the welfare state’.6 Thus, those sources of cross-national variation in ‘pre-government discretionary 6

Robson 1976.

136

Welfare regimes matter

time’ might also be treated as part of countries’ differing ‘welfare regime’ in that much broader sense. Comparing the right-hand column to the left for each country in Figure 8.1 allows us to reckon the impact of the welfare regime – its taxes, transfers and child-care subsidies – on discretionary time of the average person in the country. As we see from Figure 8.1, in all but one of our countries that impact is negative (but generally only by an hour or two a week). That is to say, averaging across the whole population, the negative effects of taxes on discretionary time generally exceed, by a small margin, the positive effects of transfer payments and child-care subsidies. The singular exception to that generalization is Sweden. There, alone among the countries under study, the state policies in view actually work to increase the average person’s discretionary time overall. The effect is not huge: it is less than an hour a week, for the average person. But as we will see in subsequent discussions, the effect is very much greater on different population subgroups – and that will be true across all the countries under study. There is one last thing we should also note from Figure 8.1. Looking at the right-hand columns again, we see that the two social-democratic countries of Scandinavia – Sweden and Finland – have the highest average ‘post-government discretionary time’ among all the countries under study. Averaging across regime types, social democrats have on average over 5 hours a week more ‘post-government discretionary time’ than corporatists or liberals. In Figure 8.2, we decompose that state impact on discretionary time, apportioning responsibility for those effects between taxes and transfers, on the one hand, and child-care subsidies, on the other. The first bar for each country represents the amount and direction of change in average ‘pre-government discretionary time’ that comes from adding in the effects of taxes-and-transfers. The second bar represents the amount and direction of change that comes from adding, to that, the effects of child-care subsidies. At this level of aggregation, anyway, welfare regimes look broadly similar in terms of the impact of their child-care subsidies on discretionary time. (Australia is strikingly low, but all the other countries are around the same level.) Where welfare regimes differ is in the net impact of their taxes-and-transfers on people’s discretionary time. Averaging across the population as a whole, taxes-and-transfers have a negative impact in all countries except Sweden (where their

A temporal perspective on welfare regimes

137

8 6

0.58

0.14 0.69

0.63

0.83

0.60

0.71

0.17

0.81

0.49

2

–6 –8

liberal countries

US

Australia

–1.89

–0.88

–1.78 –3.84

–2.81

–4

–2.51

–2

–0.55

0 –1.53

hours per week

4

corporatist Germany France countries taxes-and-transfers

social- Sweden Finland democratic countries

child-care support

Figure 8.2 State impact (decomposed) on discretionary time, whole population

impact is weakly positive). But the magnitude of those negative effects of taxes-and-transfers varies notably across countries. Furthermore, those differences do not track in conventional welfare-regime types. Of the three countries whose taxes-and-transfers have the most negative impacts, one is corporatist, one is liberal and one is social democratic. The net effect of taxes-and-transfers is to reduce discretionary time on average by almost 4 hours a week in Germany, 2.5 in the US and just under 2 in Finland.

8.3 Differing state impacts on parents Nationwide averages can obscure as much as they reveal, however. No one – not even universalistic social democrats – expect the welfare regime to affect everyone in society literally equally, exactly the same. Some people need more help than others, and hopefully they get more. Some of the most important groups in that respect are outside the scope of our study, purely on methodological grounds. Among them are households in which all adults are unemployed or retired.7 For 7

We do, however, include in our study households where an unemployed or retired person has a partner in paid labour: they appear as ‘single-earner couples’.

138

Welfare regimes matter

purposes of our methodology, we need a wage rate to calculate ‘discretionary time’ in the way that we do; and no one in those households has a wage or hence a wage rate. The very thing that makes them of special concern for the welfare state forces us to ignore them in our study. That is an unfortunate fact about methodology, but an unavoidable one.8 Other major areas of welfare-state concern we address only somewhat obliquely. Take ‘child poverty’, for example. The proportion of children living in poverty is often taken as a key indicator of the performance of a country’s welfare sector, just as infant mortality rates are of the performance of its health sector.9 Child poverty rates vary dramatically across countries, and across household types within countries: they are invariably much higher in households with a lone mother, and much higher still where that lone mother is not in paid labour.10 The latter cases we fail to pick up at all, owing to the limitation of our methodology just discussed. The former cases we pick up only obliquely, registering the suffering of the children through the low ‘discretionary time’ of the parent. In other cases, however, our methodology does offer us insight into the comparative well-being of different social groups, calibrated in temporal terms. It allows us to explore in some depth the relative temporal well-being of men and women, in various different social circumstances. We devote a whole other chapter (chapter 11) to that

8

9

10

For now, anyway. Perhaps it might be possible to impute a wage rate to the unemployed based on their human-capital characteristics. Presumably it would typically be a pretty low wage rate, so their discretionary time would be low on the grounds that they would have to work long hours in paid labour (if they could only find some) to get a poverty-level income. Discretionary time calculations of this sort might thus serve to emphasize the importance of the education-and-training components of activation policies in the labour market. OECD 2005b, tables EQ3.1–3.2; World Bank 2005; WHO 2005. See similarly, e.g., Esping-Andersen 2002, ch. 2. OECD (2005b, table EQ3.1) reports the following statistics on the ‘share of children 17 years and under living in households with equivalised disposable income less than 50% of median income’ for the countries of concern to us here in 2000: US 21.7%; Australia 11.6%; Germany 12.8%; France 7.3%; Sweden 3.6%; Finland 3.4%. Child poverty rates are two or three times that for households headed by lone parents, and around twice that again where the lone parent is unemployed. Child poverty rates for households with a lone parent who is unemployed ranged in 2000 from a low of 25% in Finland to a high of 93.8% in the US, among the countries here under study (OECD 2005b, table EQ.2).

A temporal perspective on welfare regimes

139

8

2.80

6

0.89

1.85

0.77

2

–3.62

–1.79

–2.70

–3.23

–4.62

–6

–1.97

–0.60

–0.38 –0.37

–3.92

–4

–2.95

–2

–0.85

0 –1.66 –0.61

hours per week

4

–8 liberal countries

US

Australia

corporatist Germany France countries without kids

social- Sweden Finland democratic countries

with kids

Figure 8.3 State impact on discretionary time, by parental status

issue. It also allows us to explore the temporal well-being of people with and without children, in various different household types. Underwriting the conditions of social reproduction – producing and rearing the next generation – has long been seen as a central state concern. The poor law was always particularly sympathetic to the plight of widows and orphans, and education was one of the earliest social functions assumed by the state. Virtually all countries of the developed world have now enacted family allowances, child benefits or tax credits to assist families financially with the costs of raising children. Child care is sometimes publicly subsidized (again, directly or indirectly through the tax system). In all those and myriad other more specific ways, the benefits of the welfare state flow particularly towards parents. What then is the differential impact of welfare-state taxes, transfers and child-care subsidies on the ‘discretionary time’ available to parents, as compared to childless people in the population? Figure 8.3 offers some evidence. In it, ‘parents’ refers to people with at least one child under age 18 living in their household; people with grown children count as non-parents for these purposes. The bars for each country represent ‘state impact on discretionary time’, that is, average ‘postgovernment discretionary time’ minus average ‘pre-government discretionary time’.

140

Welfare regimes matter

As we see from Figure 8.3, all welfare regimes have, on average, a negative impact on the discretionary time of people without children. That is to say, all welfare regimes typically take more discretionary time from them through taxes than they give to them via transfer payments. Average state impact on discretionary time is less negative for people with children than for those without across all three welfare regimes and, indeed, in all six countries under study (albeit only by a whisker in Australia). There are important differences in net state impact across the three types of welfare regime, however. In the two liberal welfare regimes, the average net impact of all state taxes, transfers and child-care subsidies on people with children is still negative. In the two social-democratic welfare regimes, in contrast, the net impact the state has on the average discretionary time of people with children is actually positive. The social-democratic welfare regime increases average discretionary time of parents by around 2 hours a week (an hour more in Sweden, an hour less in Finland). The two corporatist welfare regimes split in their treatment of parents, the French state giving them more discretionary time while the German state reduces it. Indeed, the German welfare regime has the strongest negative impact on people both with and without children of any of the six countries under study. In Figure 8.4, we once again decompose state impact on discretionary time, disaggregating the separate effects of taxes-and-transfers and of child-care subsidies on the discretionary time of people with children. The first bar for each country once again represents the magnitude and direction of change in ‘pre-government discretionary time’ that comes from taking into account the effects of taxes-and-transfers. The second bar represents the magnitude and direction of change that comes from taking into account the effects of child-care subsidies. The main message of Figure 8.4 is once again one of diversity within regime types as regards which policy instruments are most responsible for producing those overall effects. In Figure 8.3 we saw both liberal regimes having a mildly negative impact on parents’ discretionary time. But from Figure 8.4 we see that they produce that similar ‘bottom line’ in different ways. In the US, the state’s impact is strongly negative on the tax-and-transfer side but strongly positive on the child-care subsidy side; whereas Australia is weakly negative on taxes-and-transfers but even more weakly positive on child-care subsidies.

A temporal perspective on welfare regimes

141

8 6

1.14

1.60 1.21

0.68 1.17

1.45

1.15

1.30

0.26

1.36

0.81

2

–0.24

–0.69 –3.12

–1.90

–4

–0.63

–2

–2.21

0 –1.42

hours per week

4

–6 –8 liberal countries

US

Australia

corporatist Germany France countries

taxes-and-transfers

social- Sweden Finland democratic countries

child-care support

Figure 8.4 State impact (decomposed) on discretionary time of parents

Among social-democratic regimes, too, there is a diversity of patterns disaggregating state impact by policy instrument. In both Sweden and Finland, the impact of child-care subsidies is solidly positive. But in Sweden the impact of taxes-and-transfers on average parental discretionary time is even more strongly positive, whereas the impact of Finnish taxes-and-transfers is weakly negative. In corporatist regimes, child-care subsidies make a strong and uniformly positive contribution to discretionary time. The negative impact of taxes and transfers is more negative in Germany than France, however, and that makes all the difference overall.

8.4 Differing state impacts by household types The category of ‘households with children’ is of course hugely heterogeneous, in ways that matter greatly to people’s welfare within them and hence to states trying to promote people’s welfare. To explore those differences, we here disaggregate by household type – households with a lone parent, households where both parents are employed, and households where one parent is employed and the other is not. From our previous analyses (in section 5.3.2) we know a lot about how much ‘post-government discretionary time’ people living in various household types enjoy. We know, in particular, that lone

142

Welfare regimes matter

8 6 hours per week

4 2 0 –2 –4

liberal countries

US

Australia

dual-earners

corporatist Germany France countries single-earner couples

without kids with kids

without kids with kids

without kids with kids

without kids with kids

without kids with kids

without kids with kids

without kids with kids

without kids with kids

–8

without kids with kids

–6

social- Sweden Finland democratic countries one adult

Figure 8.5 State impact on discretionary time, by household type

parents have strikingly little ‘post-government discretionary time’. Our concern here is with how this ‘post-government’ result was produced – that is to say, with how much extra discretionary time was given to people (or taken from people) by the effects of the state’s taxes, transfers and child-care subsidies. Figure 8.5 reports average state impact on discretionary time for each of those household types, with and without children. ‘State impact’, as before, is simply ‘post- minus pre-government discretionary time’. Where state impact is negative, that means that on average people in that household type lose more discretionary time through the influence of taxes than they gain through the influence of state transfers and child-care subsidies. Our primary concern here is with the differences in state impact on discretionary time among different types of household with children. But we include in Figure 8.5 bars for the corresponding household type where there are no children, for comparison. Looking at the first suite of three ‘childless’ bars in each country’s entry in Figure 8.5, we see that the state impact on the discretionary time of childless people is always negative. We knew from Figure 8.3 that that was true, averaging across all household types. But from Figure 8.5, we can now confirm that that is true of each type of childless household, in each of the countries

A temporal perspective on welfare regimes

143

under study. The pattern of penalizing the childless is not uniform, however. It is stronger in some countries (Germany and Finland, for example); and it is stronger for some household types (childless singles, for example, who have almost as much ‘pre-government discretionary time’ as childless dual-earners, but who see more of it taxed away). Let us now shift our focus to households with children. There we see a very mixed pattern overall. There is considerable variation from country to country in the state’s impact on the discretionary time of parents; and while some of that variation fits neatly into regime clusters, some of it does not immediately seem to fit. The issue of ‘regime fit’ will be taken up in section 8.5 below. For now, we shall confine ourselves to some general remarks, looking across the whole set of countries. One striking feature of Figure 8.5 is that it shows that states almost invariably have a strong positive impact on the discretionary time of lone parents. The state’s taxes, transfers and child-care subsidies generally give lone parents an extra 4 to 6 hours’ a week more discretionary time than they would have had without them. That is a hefty gain. It is like being able to leave work early (perhaps very early) every Friday afternoon. The one exception to that general pattern is Germany, where the state penalizes lone parents by as much as any of the other states benefit them. A second striking feature of Figure 8.5 is that it shows that states almost invariably have a negative impact on the discretionary time of dual-earner parents. The impact is more strongly negative in Germany and the US (where it is 2 to 3 hours) than in Australia, France or Finland (where it is less than 1 hour). But the general pattern is found across all five of those countries. In all of them, the taxes dual-earner parents pay cost them more discretionary time than they gain from state transfers or child-care subsidies. The lone exception to that general pattern is Sweden, where dual-earner parents on average gain more than 2 hours a week in discretionary time from those state policies.

8.5 State impacts on subgroups of regime-specific concern Our discussion in the previous chapter suggests that different welfare regimes take particular interest in the welfare of different groups within society. To assess welfare regimes’ differential impacts, it is therefore necessary to conduct a more detailed assessment of their impact on the

144

Welfare regimes matter

discretionary time and temporal autonomy of the particular groups of special concern to each of them. We begin our discussion in each case with a reminder of what particular groups we would expect to be of special concern to each welfare regime, given the welfare-regime logics set out in chapter 7 above. We then examine the impact of each sort of welfare regime on the discretionary time of those particular groups, to see if those expectations are indeed borne out.

8.5.1 Liberal regimes and lone parents Liberal welfare-gender regimes, recall, concentrate their largesse first and foremost on the ‘deserving poor’ (the old, the young and the disabled) and, second, on the poor as such. Regimes with more pro-natalist orientations of the sort we shall discuss later (section 10.2) might consider parenthood a deserving status in its own right. Even liberal welfare regimes were historically solicitous of the needs of lone parents, particularly when they were envisaged as being largely widows with children. Programmes bearing witness to that fact include Aid to Families with Dependent Children and the Australian Sole Parent Pension (the former defunct but the latter still extant in the period here under study).11 As unmarried mothers became predominant within the class, lone mothers ceased being regarded as ‘deserving poor’, in the US at least. Nonetheless, liberal welfare regimes should still be expected to concentrate benefits on single parents and their children as the ‘poorest of the poor’. For that reason if no other, we ought to expect that liberal regimes ought to be expected to assist lone parents, and largely them alone. Some (but not all) of the other welfare regimes ought to be expected to assist lone parents as well, of course. It is the exclusivity of its focus on lone parents as the ‘poorest of the poor’, and them alone, that would be expected to be the hallmark of the liberal welfare regime. Figure 8.5 has offered evidence on that point. There we see:

 The liberal welfare regimes of the US and Australia do indeed have strong positive impacts on the discretionary time of lone parents. Both liberal regimes give lone parents around 6 extra hours a week. 11

Weaver 2000; Barrett 2001; Dapre` 2006, p. 117.

A temporal perspective on welfare regimes

145

(Sweden does about the same, but none of the other countries do nearly as much.)  Lone parents is the only household type on which those liberal regimes of the US and Australia have any positive impact. On average, the impact of state taxes, transfers and child-care subsidies is negative on the discretionary time of people in all other household types – even other types of households with children. The assistance that liberal welfare regimes give to lone parents must be seen in context, of course. Lone parents have less ‘pre-government discretionary time’ in Australia and especially the US than in any of the other countries. In the US, at least, that is largely because the wage rate of lone mothers is remarkably low, compared to the average wage rate for all workers.12 Even after the liberal regime’s strong performance in assisting them, lone parents in the US anyway still end up with very substantially less post-government discretionary time than their counterparts in other countries (over 8 hours a week less than in the US’s nearest rival in this respect, Germany). Still, if we are judging welfare regimes by the priorities that they set, the liberal regimes of the US and Australia run true to form. In terms of discretionary time, they help lone parents, as the ‘poorest of the poor’ – and them alone.

8.5.2 Social-democratic regimes and parents in general Social-democratic welfare regimes are focused on social equality, in its many dimensions. Thinking in temporal terms, this would lead them to focus particularly on ways to relieve the time pressure where it is greatest – which is (as we have seen in chapter 5) on parents in general, and on lone parents in particular. The egalitarianism of social-democratic welfare regimes has important gender dimensions as well. This will be discussed in much greater depth in part IV below. Suffice it for now to say that the quest for gender equality leads social-democratic welfare regimes to promote the equal participation of men and women in the paid-labour market, even 12

The mean wage rate for lone mothers, as a percentage of the mean wage rate for all workers, is 65.88% in the US whereas in all the other countries under study it is over 80% and sometimes over 90%. Percentages in the other countries are: Australia 88.45; Germany 93.97; France 81.19; Sweden 86.73; Finland 83.59.

146

Welfare regimes matter

(or especially) if they have children. This consideration leads us to expect social-democratic welfare regimes to be particularly supportive of parents in two-earner couples. That expectation is borne out by the evidence. Figure 8.3 has shown that, averaging across the two social-democratic countries, that welfare regime has the strongest positive impact on the discretionary time of parents as a whole. Figure 8.5 has shown that the Swedish state, alone among those under study, has a positive impact on parents in all three types of household. Both Finland and particularly Sweden have a strong positive impact on discretionary time of lone parents in particular, and to a lesser extent on discretionary time of parents in oneearner couples. Sweden, alone among all countries under study, has a positive impact (of a couple of hours a week) on discretionary time of dual-earner parents; and although Finland’s impact on them is weakly negative, its impact is less negative on them than in any of the other countries under study. Again, this performance should be set in context. The two countries with social-democratic regimes provide parents in all household types with more ‘pre-government discretionary time’ than do any of the other countries under study. In Sweden, dual-earner couples get yet more discretionary time, post-government, thanks to the workings of the welfare regime there; in Finland, they get a little less. But even in Finland they are still left with more post-government discretionary time than in any of our other countries except Sweden (as we saw in Figure 5.4). Judging welfare regimes by the priorities that they set, then, it seems that the social-democratic regimes of Sweden and Finland run broadly true to form. In terms of discretionary time, they help parents in general, lone parents in particular and dual-earner parents (at least in Sweden).

8.5.3 Corporatist regimes and stay-at-home parents Corporatist welfare regimes, as conservative regimes, are most concerned to promote traditional family structures, leading them to take a particular interest in stay-at-home mothers. We will examine the specifically gendered aspects of that preference in part IV below. Here we will focus instead simply on the household type: corporatists should be expected to favour parents living as single-earner couples.

147 6.85

A temporal perspective on welfare regimes

–6 –8

liberal US countries

0.00

0.00

4.13

5.49 0.00

0.58

0.00

–3.43

–4

–1.91

–2

0.00

0

–1.50

0.00

2

–1.70

hours per week

4

2.83

3.48

6

4.58 4.14

8

Australia

corporatist Germany France countries

employed partner

social- Sweden Finland democratic countries

stay-at-home partner

Figure 8.6 State impact on discretionary time of single-earner couples with children, by earner status

On their face, the results of Figure 8.5 are not particularly clear-cut on this score. True, corporatist France has the strongest positive impact of all six countries on single-earner coupled parents. But in the other corporatist country, Germany, the state’s impact on discretionary time in such households is weakly negative; and in the two socialdemocratic countries, state impact on people in such households is positive, although not as strongly so as in France. In short, looking at Figure 8.5, it would not appear that favouritism towards single-earner couples with kids is the distinguishing feature of the corporatist welfare regimes. Things look very different, however, when we disaggregate those effects on single-earner coupled households, looking at which person within the household received the benefits. Those results are shown in Figure 8.6. In Figure 8.6, we see that Germany and France are strikingly different from the other countries under study in increasing the average discretionary time of the stay-at-home parent among one-earner couples with children.13 Furthermore, they do so by a noticeable extent – by 3 or 4 hours per week. At the level of the household overall, that 13

The stay-at-home parent is almost invariably the woman, but we will discuss the more specifically gendered aspects in part IV below; here we couch our discussion in more generic terms.

148

Welfare regimes matter

effect is masked in Germany by the equally large negative impact of taxes on the earner-parent’s discretionary time. And in the two Scandinavian countries, similar effects are produced at the household level, but again by policies that impact (positively, there) on the discretionary time of the earner-parent.14 Upon closer inspection, then, the two corporatist countries run true to form. They and they alone increase the discretionary time of the stayat-home parent in their preferred family form of one-earner couples with children.

14

To some extent, this finding might be seen as an artifact of our methodology. Where the state pays families money, in our terms that reduces the amount of time that the earner-partner has to spend in paid labour to achieve a povertylevel income for the family; it shows up as a benefit to him, even where the policymaker is thinking of the payment as a ‘mother’s wage’ for the stay-at-home mother. The only way that any of the specific tax, transfer or child-care-subsidy policies can increase the stay-at-home parent’s discretionary time is through the provision of free or subsidized child care that is utilized by stay-at-home parents. That is precisely what happens in France (with the e´cole maternelle) and Germany (with kindergartens), which is what produces the findings described in the text. But it is not just a matter of our methodology, but rather a genuine question in the real world: if the money gets paid to the employed partner (as a tax credit, for example) it is an open question how intra-household bargaining plays out, whether that money gets passed on to the non-employed partner (to hire a babysitter, for example, and thus increase the discretionary time of the stay-at-home parent).

9

Welfare regimes and temporal autonomy

So what have we learned about welfare regimes, by looking at them through a temporal lens? Judging from the effects of state taxes, transfer payments and child-care subsidies – which are only parts of the ‘welfare regime’, but important parts – the following broad-brush conclusions emerge. In general, people have more temporal autonomy in social-democratic regimes. They have about 5 hours a week more discretionary time there than in corporatist and liberal regimes, which are similar to one another in this respect. That is largely due to factors outside the scope of our study. Those may or may not be social, political or economic factors that would fall under the rubric of the ‘welfare regime’ or ‘welfare society’, conceived more broadly than our data permit us to explore. Budgets must broadly balance, in temporal terms as well as financial ones. Governments can provide people with benefits and subsidies that increase their discretionary time only by collecting taxes that reduce the discretionary time. Given the complex operationalizations described in chapter 2, however, money does not translate straightforwardly into discretionary time on a one-for-one basis. So depending on whom exactly they take the money from and whom exactly they give it to, tax-transfer policies can have either negative or positive overall effects on temporal autonomy. In most of the countries under study, the net effect of government taxes, transfers and child-care subsidies is to reduce temporal autonomy slightly, averaging across the population as a whole. The amount of the reduction in discretionary time varies greatly across regime types, however. It is only a quarter-hour a week averaging across social-democratic regimes, compared to an hour or an hour-and-three-quarters in liberal and corporatist ones respectively. Alone among the six countries under study, the Swedish state actually increases mean discretionary time across the population as a whole (by about three-quarters of an hour a week). 149

150

Welfare regimes matter

All regimes reduce discretionary time of people without children substantially more than that. For people with children, regime performance differs. Social-democratic welfare regimes actually increase the discretionary time of parents (by a couple of hours a week). Liberal regimes reduce the discretionary time even of parents, but by only about a third as much as they do that of people without children. Corporatist regimes vary, with Germany reducing the discretionary time of parents (but only half as much as they do that of non-parents), while France increases it (by around 45 minutes a week). Each welfare regime makes a substantial contribution to the discretionary time of the particular groups in the population of particular ideological concern to it.

 For liberal regimes, that is lone parents, as the poorest of the poor. Lone parents enjoy 6 hours more discretionary time a week than they would have done without the liberal state’s taxes, transfers and child-care subsidies.  The social-democratic regimes, being pro-natalist, focus on helping parents across the board. Parents enjoy a couple of hours more discretionary time a week thanks to the social-democratic state’s taxes, transfers and child-care subsidies. Lone parents enjoy fully 5 hours a week more.  Conservative corporatist regimes promote traditional family values by helping one parent to stay at home with the children while the other goes out to work. Stay-at-home parents enjoy around 3 to 4 hours more discretionary time a week than they would have done without the corporatist state’s taxes, transfers and child-care subsidies. Thus, welfare regimes help the people we would expect them to help. Taking a temporal perspective enables us to say ‘how much’ in a way that is genuinely interpersonally and cross-nationally meaningful. The upshot of our analysis is that each sort of welfare regime gives its ‘preferred customers’ 2 to 6 hours a week more discretionary time than they would have had without its activities. Think of that as akin to not having to go into work on Monday mornings. In some larger sense, that would not make a huge difference to one’s life, perhaps. Still, any union leader who won that concession from employers would be regarded as a real hero. So too, we suggest, should welfare regimes.

PART IV

Gender regimes matter

10

How gender regimes differ

Among feminists themselves it has long been axiomatic that what social-welfare arrangements best promote the interests of women is a question that admits of no single, simple answer.1 The interests of women are many and varied. Given that heterogeneity, there is no one outcome that is ‘good for all women’.2 All theorists of the welfare state surely know that, if they only stopped to think about it. In attempting to ‘gender’ their analyses of welfare states, however, they often proceed in ways that elide it. Among the Scandinavian voices that often dominate these discussions, it is typically taken for granted that what it is for a regime to be ‘woman-friendly’ is to promote female labour-force participation.3 Ironically, almost the opposite was taken for granted in US debates surrounding the 1996 welfare reforms: it was seen as distinctly ‘woman-unfriendly’ to push poor single mothers into paid labour by withdrawing public assistance.4 Women’s interests vary in part because their social circumstances vary.5 Women’s interests vary, too, in part because their preferences vary. Catherine Hakim, as we have already seen, makes much of the heterogeneity in women’s preferences over issues of ‘work–life

1 2 3

4 5

Orloff 1997, p. 190. Orloff 1997, p. 191. See also: Walby 1992; O’Connor et al. 1999, pp. 34–6. Hernes 1984. In a report commissioned by the Belgian Presidency of the Council of the European Union, for example, Esping-Andersen (2002, p. 94) writes, ‘There exists a broad consensus on what constitutes women-friendly policy. It includes: (1) Affordable day care . . . (2) Paid maternity but, arguably much more importantly, also paid parental leave. A more equal division of family tasks may occur at the margin by encouraging fathers to make more use of paternal leave entitlements. (3) Provisions for work absence when children are ill.’ Moynihan 1995; Orloff 2002a, p. 99. Both their personal circumstances and their larger social circumstances: institutionalized models of the life course make different paths available, and validate them as appropriate in different countries.

153

154

Gender regimes matter

balance’. On her (much contested) count, perhaps a fifth of mothers are firmly home-centred and prefer to make a career of marriage and childraising, with only a fifth being resolutely work-centred (the large remainder being prepared to go either way depending on opportunities and incentives).6 Hakim’s critics claim that it is not so much personal preferences as systemic pressures that force women to make the decisions that they do: many who stay home would prefer to go out to work, and many in paid work would prefer more of it; many mothers in paid labour would prefer fewer hours, or none at all.7 Perhaps all we can really conclude is that it is ‘good for women’ to have the right to choose, on fair terms, for themselves. Much praised though women are for their ‘ethic of care’, the fact remains that ‘for the choice to engage in caring work to be genuine, women must be able to choose not to care’.8 More generally, we can say that it’s good for women to have a right to ‘independence’ and ‘autonomy’ across many dimensions.9 Ending the subordination of women was a key goal of feminism from the start.10 Central to that is the right in turn to form an ‘autonomous household’ – to live ‘independently’ if she so chooses.11 Having that as an option in turn empowers her to form a joint household in which she can live inter-dependently, on terms genuinely of her choosing. In this chapter we will be offering a very partial perspective on these wider issues. There is far, far more to a country’s overall gender regime than its policies of tax, transfer and public subsidy. Still, those are the only aspects of the overall gender regime that our data will allow us to assess at all directly. So, after a brief allusion to wider issues (in section 10.1), we will confine the further remarks of this chapter to those narrower aspects of the ‘gender regime’ on which the subsequent chapter’s data analysis will ultimately bear. We begin our discussion of the social-policy aspects of gender regimes with a pair of historical vignettes (in section 10.2) illustrating 6 8

9 10

11

7 Hakim 2000, p. 158. McRae 2003; OECD 2001, p. 136. Land and Rose 1985, quoted in Lewis and A˚stro¨m 1992, p. 80. Cf. Gilligan 1982; Folbre 2001. Pateman 1988a; Fraser 1994; Young 1995. On Linda Gordon’s (1979, p. 107 n.) account, feminism is literally defined as ‘an analysis of women’s subordination for the purpose of figuring out how to change it’ (quoted in Eisenstein, 1984, p. xii). Orloff 1993; O’Connor 1993.

How gender regimes differ

155

the differing ways in which France and Sweden responded to pronatalist concerns early in the twentieth century. We then proceed (in section 10.3) to abstract from those vignettes the two classic models of gendered welfare regimes (male breadwinner and dual-earner), surveying (in section 10.4) some of the standard statistical indicators conventionally used to classify countries into one or the other. We end (in section 10.5) by complicating that simple classificatory scheme with a discussion of lone mothers; countries’ attitudes towards them cut across the traditional dichotomy, yielding the richer classificatory scheme proposed in section 10.5.2.

10.1 Capturing gender: the challenge In her classic work Theorizing Patriarchy, Sylvia Walby identifies ‘six main . . . structures which together constitute a system of patriarchy’:

 a patriarchal mode of production in which women’s labour is expro    

priated by their husbands; patriarchal relation within waged labour; the patriarchal state; male violence; patriarchal relations in sexuality; and patriarchal culture.12

A full account of overall gender regimes would need to incorporate all six. Clearly, there is much that the state can do either to reproduce patriarchy or to reduce it in all those dimensions.13 In all our countries under study, employers were legally permitted to fire women upon marriage until 1939 (when Sweden first legislated against it). In Australia, women were not allowed to work in clerical or administrative posts in the Commonwealth Public Service until 1949, and they were automatically discharged upon marriage until 1966. As late as 1977, a German husband was legally empowered to prevent his wife from taking paid employment that ‘he felt . . . to be detrimental to

12 13

Walby 1989, p. 220; expanded in Walby 1990. The following discussion draws heavily on Therborn 1995, pp. 104–11. See more generally MacKinnon 1989.

156

Gender regimes matter

family life’.14 Under common law, it was traditionally legally impossible for a husband to be charged with raping his wife, and that remained so even in Sweden until 1962.15 Feminists are thus certainly right to criticize welfare-state studies for their ‘narrow and unrepresentative vision’ of the wide range of ways in which the state enforces gender codes of societies.16 They did tend to focus on ‘a limited set of chosen policy institutions’, traditionally ‘almost exclusively on cash transfers, and often even more narrowly on unemployment compensation and pensions’. That narrowness undoubtedly did indeed lead to a traditional ‘scholarly blindness to gender in social policy’.17 When analysts of welfare regimes attempt to extend their models to include some ‘gender’ dimension, they do so primarily by trying to take greater account of Walby’s first two dimensions – the gendering of employment relations, and of unpaid household responsibilities affecting them. That was the focus of most early feminist commentaries on welfare regimes, just as it was of mainstream welfare-regime theorists trying to come to grips with gender.18 Later work by both has increasingly emphasized the distribution of unpaid caring labour as a central dimension of gender-welfare regimes.19 But feminist critics continue to complain, quite rightly, that that still ‘offers no point of entry for discussing the significance of gender in the politics that configure’ those patterns of paid and unpaid caring labour.20 We explore the ‘micro-politics of gender’, at the household level, in part V. There, we discuss various different ways of running one’s household, and the impact of alternative choices of ‘household regime’ on the discretionary time available to women and men in the household. We

14

15

16 18

19 20

‘Or, alternatively, to force her to earn money if his income was deficient’ (Ostner 1993, pp. 99–100). No less venerable a source than Sir Matthew Hale’s History of the Pleas of the Crown holds that ‘the husband cannot be guilty of a rape committed by himself upon his lawful wife, for by their mutual matrimonial consent and contract the wife hath given up herself in this kind unto her husband, which she cannot retract’ (Hale 1778, vol. I, p. 628; quoted in Pateman 1988b, p. 123). Shaver 2002, pp. 204, 205. 17 Shaver 2000, p. 218. Among the former are: Hernes 1984; Lewis 1992 (as Lewis 1997 concedes); O’Connor 1993; Land 1994; Sainsbury 1996; Orloff 2002b. Among the latter see especially: Esping-Andersen 1999, pp. v, 47–72; 2002, ch. 3; Korpi 2000. Lewis 1997; Esping-Andersen 1999, ch. 4; 2002, ch. 3. Shaver 2000, p. 217.

How gender regimes differ

157

also discuss there the way in which countries’ gender regimes ease or exacerbate the effects of those household choices. As for the ‘macro-politics of gender’, we can get some sense of that simply by looking at the ‘gender gap’ – the proportion of female representatives, minus the proportion of male – in the national legislatures.21 We will present those statistics later, in conjunction with various other indicators of gender-regime type for each of our countries. Suffice it for now to say that, unsurprisingly, countries with more women in politics also have (either as cause or consequence) a wide range of other features conventionally regarded as ‘woman-friendly’. In the bulk of this chapter, however, we focus narrowly on the socialpolicy aspects of gender regimes, and more narrowly still on policies that impact upon women as employees and as caregivers. We persist in that narrower focus in part because that has become the tradition within the social-welfare literature with which we are here engaging. We do so in larger part because (as we have said) those are the only sorts of policy effects that we will be able to explore empirically, given the data we are analysing in this book.

10.2 Two modes of maternalism Having acknowledged those many larger aspects of gender regimes that are necessarily beyond our purview, let us now turn to distinctively social-welfare aspects of gender regimes. The gendering of welfare regimes has historically been, first and foremost, maternalist in nature.22 The ‘women’s welfare state’ has long consisted primarily in assistance, in cash and in kind, for mothers and their children. Initially that took the form of poor-law relief to widows and orphans, as discussed in section 10.5.1 below. But over the course of the twentieth century, it came increasingly to take the form of ‘Mothers’ Aid’, initially means- and character-tested but increasingly universal in form. By the century’s end, virtually all the countries of the OECD had a moderately generous universal ‘child benefit’ or ‘family allowance’ – with the US and now also Australia being conspicuous exceptions.23 21 23

Korpi 2000, pp. 136–7. 22 Lewis 1980; Koven and Michel 1990; Lake 1992. A (fairly lenient) means test was imposed on Australia’s family allowance in 1988 (Whiteford et al. 2001, pp. 27–8; Gauthier 1996, p. 166). The US offers much more rigorously means-tested assistance to families (on an ongoing basis until 1996, on a temporary basis thereafter). Among the other countries under

158

Gender regimes matter

Maternity is much celebrated in its own right, of course.24 ‘Motherhood issues’ are ones that politicians dare not oppose.25 But motherhood became a particularly urgent item on the political agenda with the demographic panic that swept much of Europe in the early twentieth century. Not all countries were equally panicked. The US and Australia, being countries of immigration, were unperturbed and they remained so until the very end of the century.26 In continental Europe, however, there was great concern with declining populations and what government might do about them.27 Here we want to recall how the differing political uptake of that early demographic panic shaped – and, as we shall see in section 10.5, continues to shape – the gender regimes of the European countries here under study.28

10.2.1 Home-oriented pro-natalism: the French way (initially) France early on was spooked by the prospect of ‘depopulation’, shocked by defeat in its 1870 war with Germany. ‘If France intended to reconquer Alsace and Lorraine, she needed more children.’29 A popular movement and various lobby groups emerged, but they made only modest political headway until the further devastation of World War I.30 ‘During the years 1919–1924 . . . a change in the political majority had brought into government the first of those bloc national

24 25

26 28

29

30

study, the generosity of family allowances varies from 4.9 per cent of average male wages in manufacturing in Germany to 7.2 per cent in Sweden; the family allowances paid to high-income families in Germany is even lower than that (Gauthier 1996, p. 166, data c. 1990). Ruddick 1980; 1989; Gilligan 1982. Note, for example, how ‘Mothers’ Aid’ policies ‘spread like wildfire’, not only across the American states (Skocpol 1992, pp. 424 ff.; Gordon 1994, ch. 3) but also the Western world (Gauthier 1996, p. 46). Hakim 2000, p. 223. 27 Glass 1940; Kamerman and Kahn 1978, pp. 10–11. This differing uptake was moulded in no small part by differences in the power of confessional parties (Morgan 2006). Questlaux and Fournier 1978, p. 124; see similarly Spengler 1979, pp. 121–2. Seventy years later, in ‘explaining to the country the reasons that led him to ask for an armistice’, Marshal Pe´tain summed up the problem as: ‘Too few children, too few arms, too few allies’ (Lenoir 1991, p. 150). ‘Around 1925, leading obstetricians and paediatricians launched a campaign to persuade public opinion to accept ‘‘that a woman who wishes to give a child to France should have a baby even as a result of an illegal and temporary union’’ ’ (Lefaucheur 1996, p. 427).

How gender regimes differ

159

Ministers known as familiaux . . . The Superior Council for Natality [was] set up and invited to review the over-all aspects of the problem.’31 The legislative breakthrough finally came in the 1930s, with the introduction of an allowance paid to ‘women at home raising children’ (la me`re au foyer). This allowance was eventually extended to all mothers, and it greatly increased in value throughout the 1930s and especially the 1940s. By the end of the war, this allowance had increased to the point that, on average, it ‘enabled a[n average] French family of four to double its income’.32 These payments were ‘conceptualized . . . as a compensation for earnings foregone’ by the mother.33 Historically, in France ‘the family’ had long been ‘a right-wing cause’.34 Apart from a brief flurry of Revolutionary fervour, the dominant family ideology was decidedly patriarchal. The Code Napole´on of 1804 re-established the husband as sovereign over all matters concerning the family, and imposed a duty of ‘obedience’ on the wife. Until 1965, the husband retained a formal legal right to prohibit his wife from engaging in paid labour.35 Against this ideological backdrop, it should come as no surprise to find pro-natalism, French-style, taking the form of payments to ‘women at home raising children’. As ever, ideology fails to track perfectly reality-on-the-ground. In practice far more women have always engaged in paid labour in France than in many other countries. Over time, public policy increasingly reflected those realities. A benefit for families with working mothers was introduced in the early 1970s, and by the end of the decade that benefit had itself been rolled together with the stay-at-home mother’s allowance into a single comple´ment familial.36 Such developments lead some commentators to complain that French family policy is ‘not entirely coherent’ and others such as Jane Lewis to call the French regime a ‘modified male-breadwinner’ model.37 But it is

31 32 33 34 35 36 37

Questlaux and Fournier 1978, p. 117. Lewis 1992, p. 208; Questlaux and Fournier 1978, pp. 119–20. Lewis 1992, p. 209. It was spoken of as a ‘maternal salary’ (Lenoir 1991, p. 162). Questlaux and Fournier 1978, p. 125. Therborn 1995, pp. 104, 109; Lewis 1992, p. 209; Lenoir 1991, p. 167. Lewis 1992, pp. 209–10; Lenoir 1991, pp. 165–85. Questlaux and Fournier 1978, p. 129; Lewis 1992, pp. 165–6. As Orloff (2002b, p. 19) says, ‘France stands out among continental European countries – also grouped as the conservative/general family support cluster – for featuring extensive publicly supported childcare services, leaves and so on, and mothers’

160

Gender regimes matter

actually pretty clear what French policy is doing: supporting motherhood in all its forms. All those later changes were in place come the period under study here. That is the policy environment to which our data on France pertain; and the support that France now provides for mothers in paid labour, particularly important in its approach to lone mothers, is also of consequence for dual-earner parents as well. For purposes of our initial ‘conceptual-mapping’ exercise here, it is the initial French reaction to the demographic panic that is of greater interest. Pro-natalism, French-style, focused – at least initially – on supporting ‘women at home raising children’. That represents one strategy for responding to problems posed by declining fertility.38

10.2.2 Employment-oriented pro-natalism: the Swedish way (ultimately) Sweden represents the contrasting case, focusing (ultimately) on strategies of promoting female independence and employment as a solution to declining fertility.39 The story starts with a wildly influential book published in 1934 by Alva and Gunnar Myrdal, which drew public attention to the fact that Swedish birth rates had dropped below replacement rate.40 The political uptake across all parties was dramatic. ‘Here was Public Problem No. 1,’ all agreed.41 ‘No people with . . . the will to live can . . . fail to undertake strong measures to combat the situation,’ the Minister of Social Affairs said in appointing a Population Commission to investigate the issue and formulate policy responses to it. To some extent, it was obvious what would be required. ‘First of all’, the Minister continued,

38

39

40 41

employment rates are higher than in other continental European countries, where service provision is conspicuous in its absence.’ As it long did in Germany, especially but hardly exclusively in the Nazi period. Reforms in German family policy to make it more like the contemporary French model are currently under discussion. Finland similarly, in large part influenced by these Swedish debates (Lindgren 1978); note however the Finnish Home Care Allowance introduced in the 1980s, discussed in section 10.2.3 below. Myrdal and Myrdal 1934. For an English gloss, see Myrdal 1940. Myrdal 1944, p. 165.

How gender regimes differ

161

measures will have to be instituted to encourage marriage, particularly in the younger age groups, and the bearing of children . . . But no matter how important this may be, the need of a large amount of socio-political intervention in order to create economic security and to improve the material welfare of our people must be frankly faced. It will be necessary to weigh different alternatives for attaining a state in which children will not be a pressing economic burden to the parents to the same extent as present.42

Most especially: Consideration must . . . be given to widening the possibilities for the care and education of children during the part of the day when the mother is gainfully employed outside the home . . . The raising of a family is now further being discouraged when, as so often happens, an employer refused to continue the employment of a woman upon marriage or when she becomes pregnant.43

That, anyway, was the Social Democratic Party Minister’s take on the situation, in establishing the Population Commission. According to Alva Myrdal’s own subsequent commentary, At this juncture anything could have happened. The remarkable thing is that, at the crucial moment, the nativity argument was wrested from the hands of the antifeminists and turned instead into a new and effective weapon harnessed to the woman’s aspirations. The earlier debate on the married woman’s right to work transformed into a fight for the employed woman’s right to marry and have children.44

The Population Commission issued various interim reports, and its final report in 1938. ‘So decidedly was [its] social program . . . accepted and so free from opposition were the major reforms enacted that this session has been called the ‘‘Riksdag of Mothers and Children’’ .’45 The civil service was reformed to become ‘a model employer and . . . a leader in . . . harmonizing women’s work with motherhood’. New laws were introduced to prohibit any employer from dismissing a woman ‘when marrying or giving birth to children, legitimate or not’. Laws on contraception and abortion were liberalized, on the grounds that ‘it is necessary that the starting point be the principle of voluntary parenthood’. And ‘subsidies for nursery education’ were introduced.46

42 44 45

Quoted in Myrdal 1944, p. 161. 43 Myrdal 1944, p. 162. Myrdal 1944, p. 456; quoted in Liljestro¨m 1978, p. 32. Myrdal 1944, p. 166. 46 Myrdal 1944, pp. 164–6, 170.

162

Gender regimes matter

Thus was born the thought that the way to increase childbearing is, first, to make a country fit for children to be born into and, second, to reassure women that they do not have to forsake careers in paid employment when becoming mothers.47 Now, these first waves of Swedish reforms presupposed that that reassurance should take the form of guarantees that mothers could resume their careers in paid labour, once their maternal duties were completed. While the children were young, it was assumed that mothers would ‘take a break’ from paid employment and devote themselves full-time to child-rearing. That was the ‘dual-role’ model of Alva Myrdal and Viola Klein: ‘women ideally entered the labor market from school and remained until the birth of their first child, returning again when the children left school’.48 Thus, even in Social Democratic Sweden, the initial response to the demographic panic was a ‘malebreadwinner’ one, at least during child-rearing years. That model of sequenced ‘dual roles’ came in for serious criticism, however. In 1968, it was officially repudiated and replaced by this vision: The aim of a long-term programme for women must be that every individual, irrespective of sex, shall have the same practical opportunities, not only for education and employment but also in principle the same responsibility for his or her own maintenance as well as a shared responsibility for the upbringing of children and the upkeep of the home.49

Various policies flowed from that vision. Separate taxation was introduced in 1971, making it advantageous for both partners to earn incomes (particularly so, given the high marginal tax rates for which Sweden is notorious); provision of public child care expanded over that decade; maternity (and paternity) benefits were indexed to income, replacing 90 per cent of previous earnings. Swedish mothers responded to these incentives by entering the paid-labour force in droves. Over the course of the third quarter of 47 48

49

See more generally Tilton 1990. ˚ stro¨m 1992, p. 66, glossing see further Myrdal and Klein 1956. See Lewis and A further Lewis 1990. Sweden 1968, p. 5; quoted in Lewis and A˚stro¨m 1992, p. 71; Liljestro¨m 1978, pp. 33–4. Initially prepared by a joint committee of the Social Democratic Party and the LO (Lands-organisattionen) as a report to the UN’s Economic and Social Council, this subsequently became official government policy.

How gender regimes differ

163

the century, Sweden went from having one of the lowest female employment rates in Europe to having one of the highest. By 1986, fully 85 per cent of women with children under 7 were in paid labour and only 7 per cent of women of prime working age (25 to 54) were classified as ‘housewives’.50 The eventual Swedish response to the demographic panic thus represents a second distinctive way of being ‘pro-natalist’: instead of confining breeder-mothers to the home, ensure that they can continue as earners outside the home.

10.2.3 Pro-natalism redux These debates have recently been revisited in the wake of a renewed ‘demographic panic’.51 Across Western Europe, birth rates are now well below replacement rate. Southern European countries are particularly hard-hit; but northern European countries are not far behind. The same is true in Australia, and would be true in the US if high fertility among Latinos did not mask low fertility among whites. Below-replacement birth rates pose a particular problem for old-age pension systems set up (as virtually all originally had been) on a ‘payas-you-go’ basis. There, payments from people currently in work are used to fund the benefits for those who have retired. The higher the ‘dependency ratio’ of beneficiaries to contributors, the more you have to make contributors pay (or the less you can afford to pay beneficiaries, or the more debt you need to allow the system to run). The ‘demographic bulge’ of baby-boomers was always going to be a problem, but it is greatly exacerbated when combined with low birth rates as well.52 Gøsta Esping-Andersen, in good Scandinavian fashion, calls for a ‘new gender contract’ to redress the situation. He points out that, today, ‘the correlation between fertility and women’s paid employment is . . . exactly the opposite of what one might expect. The higher the rate of female employment, the greater the level of fertility.’53 Explaining the correlation, he observes that ‘at one end, this cross-national correlation is driven by the Nordic countries’ unique ability to maximize female [labour force] activity and birth rates. At the other end, the 50 52

Lewis 1992, p. 169. World Bank 1994.

51 53

Gauthier 1996, ch. 8. Esping-Andersen 1999, p. 67.

164

Gender regimes matter

correlation is being driven by . . . women in large parts of Europe postpon[ing] having children as long as they cannot find affordable day care or remain stuck in seemingly endless job queues.’54 Whereas policies to facilitate female employment are EspingAndersen’s preferred solution to the current demographic crisis, policies to assist stay-at-home mothers are Catherine Hakim’s. As she points out, home-centred women are the most reliable breeders. ‘Work-centred women are unlikely to respond to pronatalist incentives’, whereas ‘home-centred women will be the most enthusiastic in their response’.55 Governments attempting to increase the birth rate through female-employment-oriented strategies are offering incentives to groups least likely to respond to them. Hakim’s preferred strategy is a ‘homecare allowance’, reminiscent of the initial French allowance for ‘women at home raising children’. The paradigm for Hakim is the Finnish scheme, which she glosses as follows: In the late 1980s, Finland introduced a new policy that gave parents the right to choose between publicly provided childcare services and a cash benefit for childcare at home . . . The new Home Care Allowance . . . is paid to all families who do not use the public daycare services, effectively providing a subsidy to fund one parent at home full-time or to help pay for private childcare services . . . Funding is generous. While the exact value of the HCA varies according to family and local circumstances . . . the allowance for one child can amount to [up] . . . to 40% of the average monthly earnings of female employees in Finland. The scheme is popular and successful. From the start, two-thirds of all mothers with a child under three used the scheme.56

10.3 Abstracting models of gender regimes Such historical vignettes offer important insights into the myriad ways in which states might approach issues of gender, maternity, work and welfare. The point of rehearsing all those old debates has not been to resolve them one way or another, however, but rather to let them help inform us concerning the range of possible approaches to these issues.

54 56

Esping-Andersen 2002, p. 71. Hakim 2000, pp. 232–3.

55

Hakim 2000, p. 226.

How gender regimes differ

165

Stepping back from that rich detail, let us now attempt to abstract from it some more stylized ‘regime types’. As with welfare-regime types, so too with gender-regime types: they are Weberian ‘ideal types’, abstractions from the real world. ‘No country has ever matched the model[s] completely, but some have come much closer than others.’57 But as with welfare-regime types, so too with gender-regime types: they point to important dimensions along which actual countries in the real world do genuinely differ, often substantially. Recent discussions of the gendering of welfare regimes focus predominantly on a contrast between ‘male-breadwinner’ and ‘dual-earner’ regimes.58 The difference lies purely in their attitudes towards married women (mothers, most particularly) engaging in paid employment. The ‘male-breadwinner regime’ discourages that. The ‘dual-earner regime’ encourages it.

10.3.1 The male-breadwinner gender regime The ‘male-breadwinner’ model is based on a family ideology that celebrates marriage and a sharp division of labour within it as a natural, God-given fact. In the words of Pope Leo XIII’s 1891 Workers’ Encyclical, ‘it is a most sacred law of nature that the father of a family see that his offspring are provided with the necessities of life’; and women are ‘intended by nature for the work of the home’. Eighty years later, while also urging an end to discrimination against women, Pope Paul VI continued to rail against ‘the misinterpreted equality which denies the differences God himself has created and . . . the woman’s special and especially important role in the heart of the family and society’.59 57 58

59

Lewis 1992, p. 162. Terminology varies. Lewis (1992) just talks about countries being ‘high’ or ‘low’ on the male-breadwinner dimension. Sainsbury (1996, p. 42) distinguishes the ‘male-breadwinner model’ from the ‘individual model’. Gauthier (1996, pp. 203–5) distinguishes the ‘traditional’ (i.e., male-breadwinner) model from the ‘egalitarian’ (i.e., dual-earner) one. And Korpi (2000; Orloff 2002b, p. 15) distinguishes three models: one of ‘general family support’ (which turns out to be the male-breadwinner model); another of ‘dual-earner support’; and a third that is ‘market-oriented’ (liberal regimes, neutral as between single and dual-earners). While the terminology varies, the substantive distinction is the same. Both are quoted in Korpi (2000, p. 149).

166

Gender regimes matter

Today, the male-breadwinner ideal remains strong principally in countries of continental Europe where ‘confessional’ parties are strong.60 But it started out as the dominant pattern across all the countries under study at the beginning of the twentieth century.61 Across all our countries, economic and social policy was historically organized around the presumption that there would (ordinarily) be a male head of household, with responsibility to provide for his dependents. He was the sole earner for his household, and his earnings were supposed to be sufficient to meet the needs of the household as a whole. A famous 1907 decision of the Australian Arbitration Court went so far as to stipulate that award wages must be sufficient to meet the needs of a worker, his wife and two children.62 The male breadwinner was supposed to be ‘independent’, in the sense of being able to support himself and his family without public assistance. The female homemaker was expected to be ‘dependent’ upon the income, support and protection of her husband.63 Historically, the position of the wife and children within the house was legally akin to that of ‘servants’ living in the house of their ‘master’: they were obliged to obey his commands, in exchange for the food, clothing and shelter; he in turn was obliged to provide those things.64 Although there has always of course been a certain proportion of wives and mothers working in the paid-labour force across all the countries under study, those proportions were modest until the second half of the twentieth century. The primary responsibility of wives, and especially mothers, was to work at domestic tasks inside the home. In terms of public policy, then, the ideal-typic ‘male-breadwinner regime’ would see ‘married women excluded from the labour market, firmly subordinated to their husbands for the purposes of social security entitlements and tax, and expected to undertake the work of caring for children and other dependants at home without public support’.65

60 61 62

63

64

Korpi 2000, p. 152; Therborn 1995, pp. 104–6. See the respective country surveys in Kamerman and Kahn 1978. Justice Higgins in Ex parte H. V. McKay (the Harvester Judgment), Commonwealth Arbitration Reports 11 (1907–8). The judgment is excerpted in Wilson et al. 1996, pp. 21–5. For discussions see: Hancock 1979; Castles 1985; Macintyre 1985, ch. 3. Fraser and Gordon 1994. See further: Pateman 1988a; Fraser 1994; Young 1995; Goodin 1998. Mill 1869, ch. 2; Pateman 1988b, ch. 5. 65 Lewis 1992, p. 162.

How gender regimes differ

167

The policy instruments typically associated with this model are as follows: *

*

*

*

*

*

In the male breadwinner model . . . the husband is the head of the household, and it is his duty to provide for the members of his family – his wife and children – through full-time employment. The duties of the wife are to make a good home and provide care for her husband and children. This division of labor is codified in family law, social and labor legislation, and the tax system. The unit of benefit is the family, and minimum benefits and pay embody the notion of the family wage. Entitlement is differentiated between husband and wife. Eligibility is based on breadwinner status and the principle of maintenance. Most wives’ rights to benefits are derived from their status as dependents within the family and their husbands’ entitlements. As a result, married women may lack individual entitlements to benefits. In its purest form, the family or household is also the unit of social insurance contributions and taxation. To compensate for the maintenance of his family and offspring the family provider receives tax relief. The division of labor prescribed by familial ideology also affects wage and labor market policies – assigning priority to men’s employment and earnings. The boundary between the private and public sphere is strictly enforced. Caring and reproduction tasks are located in the private sphere, primarily in the home, and this work is unpaid.66

10.3.2 The dual-earner (universal-breadwinner) gender regime In today’s standard set-piece argument, the ‘male-breadwinner’ model is most commonly contrasted with a ‘dual-earner’ model. That ‘dual-earner’ model is based on an ideology of gender equality, and ‘it aims to achieve [that] . . . by promoting women’s employment. The point is to enable women to support themselves and their families through their own wage earning.’ In effect, ‘the breadwinner role is to be universalized . . . so that women too can be citizen workers’.67 As the Swedish government famously declared in the 1968 pronouncement quoted earlier, ‘every individual, irrespective of sex, shall have . . . the same responsibility for his or her own maintenance as well as a shared

66 67

Sainsbury 1996, pp. 41–2. Fraser 1994, p. 601. This became official policy in Sweden in the wake of a 1968 joint task force of the LO and the Social Democratic Party recommendation.

168

Gender regimes matter

responsibility for the upbringing of children and the upkeep of the home’.68 ‘Universal breadwinner’ is perhaps the more apt description, emphasizing as it does that the underlying aspiration is to make everyone (including single women and lone mothers) self-supporting ‘breadwinners’, too. But when (as most commonly happens) this model is being discussed in contrast to the ‘male-breadwinner’ one, attention focuses on couples and the employment status of women within them, whether they are stay-at-home housewives or whether they are also in paid labour. Hence, the most commonly discussed form of the ‘universal-breadwinner’ model is a ‘dual-earner’ model, and that is the name by which the more general model is most commonly (if rather imprecisely) known. Although a minority pattern at mid-century across all the countries under study, by the end of the twentieth century dual-earning had become the most common pattern for couples. Across all the countries under study, virtually half (in several countries more) of mothers even with a child under 6 years old engage in paid employment; and that has been so for all the years straddled by our study.69 True, full gender equality has yet to be achieved: women work fewer hours and typically at lower rates of pay; their annual earnings in most of the countries under discussion average only around 70 per cent of men’s in consequence.70 Still, the ‘dual-earner’ pattern now predominates across all the countries under study. In the ideal-typic ‘dual-earner regime’ there are two partners both of whom are in paid labour. Public policy treats them as wholly separate individuals for purposes of assessing taxes and awarding social benefits. The policy instruments typically associated with a ‘dual-earner regime’ can be summarized as follows: *

*

68

69

Each adult is individually responsible for his or her own maintenance, and the father and mother share the tasks of financial support and care of their children. An essential basis of entitlement is citizenship or residence because it acknowledges that individuals have a variety of useful tasks in life not limited to paid work. It privileges neither earning nor caring and thus accommodates the shared tasks of earner and carer. Sweden 1968, p. 5; quoted in Lewis and A˚stro¨m 1992, p. 71; Liljestro¨m 1978, pp. 33–4. OECD 2001, p. 134; OECD 2005b, table SS4. 70 Orloff 2002b, p. 8.

How gender regimes differ *

* *

169

The unit of benefit, contributions and taxation is the individual with no deductions or allowances for dependents. Labor market policies are aimed at both sexes. The boundary between private and public spheres is fluid. Many reproductive tasks are performed in the public sector. Care, even in the home, can be paid work and provide entitlement to social security benefits.71

10.4 Classifying countries: some standard indicators Let us now present briefly some of the standard sorts of indicators that are typically used to sort countries into those two standard regime types. Some key variables for that purpose are displayed in Table 10.1. We start with an indicator of ‘gender gap in political power’, the difference in the proportions of women and men in the national legislature. This ‘gender gap’ is, as the first line in Table 10.1 shows, substantially lower in Scandinavia than either continental European or Anglo-Saxon countries under study. Those differences may be the cause, or they may be the consequence, of the differences found in the subsequent lines of Table 10.1.72 Table 10.1 suggests the following broad conclusions concerning the gender regimes of the countries under study:

 Sweden and Finland are clearly ‘dual-earner gender regimes’, judging from the high female employment rates found there. Germany and France are clearly ‘male-breadwinner gender regimes’, judging both from the low female employment rates and relevant features of their tax system (joint-taxation being pro-breadwinner, and separate being pro-dual-earner). The US and Australia constitute more ambiguous cases, based on those considerations.73  Germany, France, Finland and Sweden are clearly ‘pro-natalist’, judging from the magnitude of their child benefits and maternity benefits. The US and Australia, by those standards, clearly are not.

71

72 73

Sainsbury 1996, pp. 42–3; note that her own preferred term for this is the ‘Individual Model’. In connection with child-care policy, for example, see Bergman 2004. These conclusions broadly correspond to Korpi’s (2000) influential categorization (reproduced approvingly in Orloff 2002b). He classes Sweden and Finland as ‘dual-earner’ regimes; France and Germany as ‘general family support’ regimes (his term for ‘male-breadwinner’); and the US and Australia as falling into a separate ‘market-oriented’ category.

76 separate 62.9 45.0 0 11.7 58.7

88 optional/joint 70.6 54.0 0 40.3 93.8

18.0 55.6

100

59.6 41.4

60 joint

Australia Germany

9.6 61.7

100

64.6 61.3

88 joint

France

7.2 25.0

70

85.1 49.4

20 separate

Finland

5.6 34.2

63

89.6 85.0

28 separate

Sweden

Source: Gender gap in national legislature, c. 1990: Korpi 2000, pp. 136–7. Tax system c. 1999: OECD 2001, p. 142. Female employment/ population ratio (age 25–54), c. 1990: OECD 2005a, pp. 247–9. Female employment/population ratio (age 25–54) of mothers with youngest child < 6, c. 1990: OECD 2005b, table SS4.1. Maternity benefit as % average wage, c. 1999–2000: OECD 2001, p. 144. Poverty rate of single-parent households, by employment status of head, c. 2000: OECD 2005b, table EQ3.2.

Gender gap in national legislature Tax system Female employment/population ratio (age 25–54) (c. 1990) . of mothers with youngest child < 6 Maternity benefit as % of average wage (1999–2000) Poverty rate of single-parent households . head in paid labour . head not in paid labour

US

Table 10.1. Gender-relevant differences across countries

How gender regimes differ

171

The ‘pro-natalism’ clearly takes a dramatically ‘employment-based’ form in Sweden and to a lesser extent in France, judging from the proportion of mothers with children under 6 who are in paid labour. Judging from that indicator, ‘pro-natalism’ takes a more ‘familybased form’ in Germany and Finland.

10.5 Taking lone mothers into account Preoccupied as they are by the contrast between ‘male-breadwinner’ and ‘dual-earner’ regimes, theorists trying to gender the welfare state forget that both those models pertain purely to intact couples. As we have already remarked, there is no place in those models for single individuals or lone parents. That is deeply ironic, since from both a historical and a policy perspective lone mothers lie at the heart of the state’s gendered approach to welfare.74 As Jane Lewis has observed, ‘historically the modern state has experienced considerable difficulty in deciding whether to treat lone mothers as workers or as mothers. In practice it has pursued first one strategy and then the other.’75 The problem arises particularly within the longtraditional ‘male-breadwinner model: in the absence of a [male] breadwinner, is the mother to be treated primarily as a mother or as a breadwinner? The answer earlier . . . was as a mother,’ who was traditionally excused from paid employment. With the rise of ‘workfare’, however, the answer has increasingly become ‘as a breadwinner’.76

10.5.1 A historical interlude Those contemporary debates have venerable antecedents. The oldfashioned way of assisting the destitute was through the ‘poor law’. Widowed mothers and orphaned children were traditionally among its prime beneficiaries.77 They were ‘deserving poor’ par excellence, whom even most laissez-faire states were prepared to assist publicly if 74

75 76

77

Likewise typologies that distinguish countries that are ‘pro-natalist’ (and interventionist about it) from those that are not ignore the difference between mothers who are partnered and those who are not (Gauthier 1996, pp. 203–5). Lewis 1986, p. 595. Lewis 1997, p. 172. She goes on to say, ‘The answer earlier in the century was as a mother, in part due to the effect of maternalist campaigns.’ See OECD 2005b, table EQ 3.1.

172

Gender regimes matter

they were unable to support themselves through their own paid labour. In earlier periods, that assistance came in workhouses, poor houses and orphanages. Later, particularly for mothers of good character, it came in the form of cash and in-kind assistance in their own homes. Historians of US social policy make much of the fact that it has from the start been ‘fundamentally divided into two channels, one originally designed for white industrial workers and the other designed for impoverished, white, working-class widows with young children . . . [This] can be seen in the first successful benefit programs in the United States: Workmen’s Compensation and Mother’s Aid. These programs had different ideologies, clienteles, principles of entitlement, and administrative styles . . . which . . . greatly reinforced the gender, racial and class organization of the welfare state’ in the US. Specifically, Mothers’ Aid was ideologically ‘linked to the poor law tradition and the administrative practices of the Charity Organization Society movement’.78 In the United States, a 1909 ‘Conference on the Care of Dependent Children’ convened by President Theodore Roosevelt concluded: Home life is the highest and finest product of civilization. It is the great molding force of mind and character. Children should not be deprived of it except for urgent and compelling reasons . . . [C]hildren of reasonably efficient and deserving mothers who are without the support of the normal breadwinner should, as a rule, be kept with their parents, such aid being given as may be necessary to maintain suitable homes for the rearing of children.79

Although private charity was what the Conference itself recommended, a well-orchestrated campaign by Women’s Clubs led to the astonishingly rapid enactment of public programmes of character-tested Mothers’ Pensions which spread across the American states over the next decade.80 These were supplanted in turn by Aid to Dependent Children (subsequently Aid to Families with Dependent Children) provisions in the national Social Security Act of 1935, which remained a prominent feature of the US welfare state until the 1996 reforms.

78

79 80

Nelson 1990, p. 124. The magnum opus on this topic is, of course, Skocpol (1992, esp. ch. 8). See further Weir et al. 1988; Fraser and Gordon 1992; 1994; Gordon 1994. Quoted in Skocpol 1992, p. 425. To forty of the then forty-eight states by 1920 (Skocpol 1992, pp. 424 ff.; Gordon 1994, ch. 3).

How gender regimes differ

173

Across Europe and Australasia, similarly, charitable relief of widows and orphans was replaced by social-insurance schemes with ‘survivor’s’ entitlements in the early twentieth century.81 At mid-century, all but one of the countries under study – and indeed of the OECD (the US being at mid-century the sole exception) – paid a universal ‘child allowance’ or ‘family benefit’.82 Widows and orphans were one thing, mothers of children born out of wedlock were quite another (with divorced mothers somewhere in between). Character-tested regimes, like the original US programmes of Mothers’ Aid, denied any assistance to them and expected them instead to give up their children to adoption or orphanages. Aid to Dependent Children, as originally conceived in the 1935 Social Security Act, envisaged support going to legitimate children who had been orphaned; and the programme lost legitimacy over the 1960s and 1970s as its beneficiaries were increasingly unwed mothers.83 In countries preoccupied by ‘traditional family values’, such as Germany, public policy can be powerfully deployed to sanction lone mothers (anyway, unwed and divorced mothers84) and to encourage women to marry and remain married.85 The ‘marriage and family’ are guaranteed the ‘special protection of the state’ under the 1949 Constitution.86 Although that document ‘stipulated equal legal treatment of all children . . . one-parent families are still viewed as 81

82

83 84

85

86

Gauthier 1996, p. 46. Early in Finland (1897), belatedly in Australia (a nationwide Widow’s Pension being enacted in 1942, generalizing the New South Wales pension enacted in 1925) (US SSA 2006; Gauthier 1996, p. 46). Finland in 1948; France in 1932; Germany in 1936, reintroduced in the FRG in 1954; Sweden in 1947 (Gauthier 1996, p. 70; Wennemo 1994, p. 99); the Australian family allowance, introduced as a universal benefit in 1941, has been (lightly) means-tested since 1988 (Whiteford et al. 2001, pp. 27–8; Gauthier 1996, p. 166). Although ‘universal’ in the sense of not being means-tested or charactertested, the benefits were often limited to families with more than a set number of children (although the number was almost invariably lowered over time). Murray 1984, pp. 17–18, 124–34; Bane 1988. Widows and orphans are supported in Germany, as elsewhere among the countries under study, through survivorship provisions in their husbands’ socialinsurance entitlements. As do so-called ‘communitarians’ in recent US debates (Galston 1991, pp. 283–8; Etzioni 1993, ch. 2). In Article 6, which reads in full: ‘(a) Marriage and family shall enjoy the special protection of the state. (2) The care and upbringing of children shall be a natural right and duty primarily incumbent on the parents. The state shall watch over their endeavours in this respect. (3) Children shall not be separated from their families

174

Gender regimes matter

defective’.87 There is policy detail aplenty to support this conclusion. But let this telling anecdote suffice: in West Germany ‘until 1970 parents as well as landlords were legally prohibited to give or rent rooms or flats to unmarried couples or to allow them to stay overnight. They could be taken to court if caught.’88 Other countries are far more supportive of lone mothers. Being ‘supportive of lone mothers’ does not mean encouraging them to become lone mothers, of course; it merely means providing them with public supports when they do (which may have the unintended but foreseeable consequences of encouraging them, of course). In France, for example, ‘the emphasis in family policy has shifted between a pronatalist and a familist orientation, but progressively its general thrust has moved . . . to a child-centred approach’.89 Lone parents there now receive nearly a quarter of their gross income through a patchwork of public benefits, including the substantial and meanstested Revenu Minimum d’Insertion and Allocation Parent Isole´.90 Gender regimes that support lone mothers come in two forms, however. One is ‘employment-oriented’, the other ‘family-oriented’. In terms of the policy instruments employed, the employment-oriented version would be very similar to the ‘dual-earner’ regime. There would be strong social supports for women engaging in paid work, including subsidized child care to facilitate that; there would be an emphasis upon flexible hours to fit family needs, and so on. Those are the sorts of policies found in Sweden, for example. The policy instruments employed by the ‘family-oriented lone-mother’ regime would resemble

87

88 90

against the will of the persons entitled to bring them up, except, pursuant to a statute, where those so entitled fail in their duties or the children are otherwise threatened with serious neglect. (4) Every mother shall be entitled to the protection and care of the community. (5) Illegitimate children shall be provided by legislation with the same opportunities for their physical and mental development and for their place in society as are enjoyed by legitimate children’ (Germany 1949, Article 6). As Ostner (1993, pp. 104–6) elaborates, there remains a ‘strong emphasis . . . on the married couple and on married parents as the basis of the family’, and ‘deviations from that norm quickly lead to phases of deprivation . . . [N]early 50 per cent of unmarried mothers live on family incomes below the poverty line; 15 per cent have to claim [means-tested social-assistance] benefits.’ While children of lone parents are no longer obligatorily taken into state custody, as they once were, that still happens ‘if the absent parent does not pay alimony or if the caring parent, usually the mother, is deemed not competent’. Ostner 1993, p. 97. 89 Hantrais 1993, p. 132. Chambaz and Martin 2001, pp. 146–7; Hantrais 1993, p. 134. See further Lefaucheur 1996.

How gender regimes differ

175

those of the ‘family-oriented pro-natalist’ one. The emphasis would be upon generous family allowances (perhaps supplemented by a dedicated ‘lone parent’s allowance’ or ‘carer’s allowance’) to enable the lone mother to support her family without having to go out to work. Generous policies of that sort are found in Finland, as discussed in section 10.2.3, and (on a means-tested basis) in France and Australia.91 The 1996 welfare reforms in the US might be seen, most fundamentally, as an attempt to shift between the ‘family-oriented’ and ‘employment-oriented lone-mother gender regimes’. The former Aid to Families with Dependent Children constituted a means-tested ‘entitlement . . . for poor single mothers to care for their children full-time’.92 It was replaced by Temporary Assistance to Needy Families, which in effect requires claimants to engage in paid labour (or anyway labour-market-related activities) except for brief periods. Across all countries, of course, there are strong incentives for lone mothers to secure paid employment, as the last row of Table 10.1 clearly shows. Lone mothers are up to six times more likely to be in poverty if they are not in paid labour than if they are. But just how strong the pressure for lone mothers to secure paid employment is varies from country to country. In female-friendly social-democratic countries, lone mothers stand only around a 30 per cent chance of being in poverty even if they are not in paid labour. In traditionalistcorporatist countries, their chances of being in poverty if not in paid labour are twice that. In Australia they are about the same, but in the arch-liberal-individualist US, lone mothers are virtually certain of being in poverty if they are not in paid labour, and they are 40 per cent likely to be in poverty even if they are in paid labour.

10.5.2 A new classificatory scheme We suggest that a very large proportion of the difference between gender regimes, as they touch upon the welfare state, can be resolved into two principal, distinct components.93 91

92 93

Whiteford et al. 2001, p. 29; Carney and Hanks 1994, pp. 223 ff.; Dapre` 2006, p. 117; Lefaucheur 1996; Chambaz and Martin 2001. Orloff 2002a, p. 99. Gauthier (1996, pp. 203–5) implicitly proposes a two-by-two matrix, with a ‘male-breadwinner/not’ dimension cutting across another ‘interventionistpronatalist/not’ dimension.

176

Gender regimes matter

Table 10.2. Fundamental divides in gender regimes with respect to mothers Treatment of married mothers Encourage to stay at home Supportive (general familyoriented) Treatment of Supportive France? lone mothers (employmentoriented) Sanctioning Germany?

Neutral Australia? (US?)

Encourage into paid labour (Finland?)

Sweden?

Both fundamentally relate to the regime’s attitude towards mothers. But what the regime thinks about the employment of married mothers (whether or not they should engage in paid labour) is, we suggest, independent of what it thinks about lone mothers (and whether they should be supported or sanctioned, and if supported whether in a general family-oriented way or in a specifically employment-oriented way). Those cross-cutting distinctions are mapped in Table 10.2. On the basis of the potted histories and summary statistics presented above, we would ordinarily be inclined to place countries in that taxonomy as indicated in Table 10.2. But those placements are provisional – as indicated by the question marks – and will be revisited in chapter 12 once we have discovered in chapter 11 what the real impact of each country’s gender regime is on these groups’ discretionary time. To foreshadow, the two countries in brackets in Table 10.2 will each be shifted one cell to the left in our revised table in chapter 12.

11

A temporal perspective on gender regimes

In this chapter, we will analyse the impact of gender regimes on discretionary time in the same way we analysed the impact of welfare regimes in chapter 8. The basic measure of impact, here as there, will be the difference that is made to people’s discretionary time by the state’s tax, transfer and child-care subsidy policies. That is far from all that the state does, as we have said in section 8.2, and our apologies for that narrow focus must be redoubled in the case of gender regimes. Still, that is all our data allow us to assess. It is a very partial account of the impact of these gender regimes, but it does nonetheless represent an important part of what these gender regimes do. In this chapter, we will continue to group the countries the same way as in our discussion of welfare regimes – and indeed we will continue labelling them with the same names. There is no clear, consensual alternative nomenclature emerging from the gender-regime literature surveyed in the previous chapter. Sometimes the same basic groups are distinguished, under different (and often distinctly unhelpful) names.1 Other times, the distinction seems to be ‘social democrats’ (sometimes under the label ‘dual-earners’) versus ‘others’. Still other times, the distinction seems to be ‘corporatists’ (under the label ‘male breadwinner’) versus ‘others’.2 By organizing our discussion around the three basic welfare-regime types, we retain the option of collapsing categories either of those latter two ways, according to taste. We organize our discussion of gender regimes around those traditional welfare-regime categories mostly, however, because of what our discussion in the previous chapter leads us to expect. On the basis of 1

2

Korpi’s (2000; Orloff 2002b, p. 15) gender model of ‘general family support’ is basically corporatist, his ‘dual-earner support’ one is basically social-democratic and his ‘market-oriented’ model is basically liberal. Cf. e.g. Lewis 1992; Sainsbury 1996, p. 42; Gauthier 1996, pp. 203–5.

177

178

Gender regimes matter

that discussion, we would expect that ‘liberal’ regimes really should hang together in their ‘individualist’ approach to households and gender. We would expect that ‘social-democratic’ regimes really should hang together in their ‘pro-woman, pro-natalist’ approach to households and gender. We would expect that ‘corporatist’ regimes might well split, however, over their attitudes towards lone mothers. Corporatists’ pro-natalism pulls them towards supporting motherhood in all its forms; their traditionalism pulls them towards supporting traditional-style stay-at-home mothers and punishing out-of-wedlock births. Which attitude dominates within any given corporatist regime is up for grabs, and different ones will dominate in different countries (and perhaps different periods in the same country). We begin our discussion of gender regimes with the same sort of ‘bigpicture’ presentation as opened chapter 8. In section 11.1, we discuss how much discretionary time both men and women have, in each regime and in each country, pre- and post-government. That forms a baseline for our subsequent discussions, which focus upon what difference the state makes (in pre- to post-government discretionary time) for various gender-differentiated groups. The most important such divide is, of course, between people who have children and those who do not. We have discussed this in a more general way in section 8.3. Here, we decompose those more general statistics by gender. Those are discussed in a general way in section 11.2. We then proceed, in section 11.3, to examine the state’s differential impact on mothers’ discretionary time depending on whether they are coupled or lone mothers. The discussion of chapter 10 leads us to expect major differences among (and in the ‘corporatist’ case, within) gender regimes in that regard. And as will be seen, those can indeed be found in our data.

11.1 Gender regimes: the big picture Gender regimes differ in how they treat women: both absolutely, and compared to how they treat men. From Figure 3.1 we know that women have less discretionary time than men, after the effects of government have been taken into account. That pattern is uniform across all six countries under study here, although it is more pronounced in some countries than in others. In France the gender gap in post-government discretionary time is

A temporal perspective on gender regimes

179

100 90

hours per week

80 70 60 50 40 30 20

corporatist Germany France countries men

social- Sweden democratic countries

pre-government post-government

pre-government post-government

pre-government post-government

Australia

pre-government post-government

US

pre-government post-government

pre-government post-government

liberal countries

pre-government post-government

pre-government post-government

0

pre-government post-government

10

Finland

women

Figure 11.1 Pre- and post-government discretionary time, by gender

5.5 hours a week, whereas in Germany it is only a half-an-hour a week. Other places it hovers around 2 hours a week. Those statistics reappear in Figure 11.1 as the second ‘post-government discretionary time’ column for each country. The lightly shaded bar represents the average amount of discretionary time that is available to women; the darker bar extending above that shows how much more discretionary time is on average available to men (which is greater in every case under discussion here). The first column for each country in Figure 11.1 represents ‘pre-government discretionary time’. In addition to country-specific information, Figure 11.1 also provides similar information for ‘regimes’ as a whole, averaging across both exemplars of each type of gender regime. As we have observed in section 8.2, both pre- and post-government discretionary time is highest, on average, in the social-democratic regimes. Breaking the population down by gender, Figure 11.1 confirms that that is true for both men and women. We also noted in section 8.2 that Sweden is the only country among those under study in which the state’s impact on discretionary time overall was actually positive – the only country in which ‘post-government discretionary

180

Gender regimes matter

3.47 3.26

1.97 1.40

0.55

2.63

2.72 2.33

4.56 3.00

2.40 1.46

3.42 1.89

2

–0.21

–0.57

–0.39

–1.03

–2.09

–1.56

–4

–0.94

–2

–1.53

0 –1.23

hours per week

4

2.91 1.68

6

6.49 5.46

8

–6 –8

liberal countries

US

Australia

corporatist Germany France countries

social- Sweden Finland democratic countries

pre-government gender difference (men minus women) post-government gender difference (men minus women) state impact on gender difference

Figure 11.2 State impact on gender gap in discretionary time

time’ is actually higher than ‘pre-government discretionary time’. Figure 11.1 confirms that that is true for both men and women.3 Some of the things that contribute to higher ‘pre-government discretionary time’ might actually be aspects of government policy (active labour-market policies and so on) that simply cannot be detected using these data. When measuring ‘state impact’, all we can do here is look at the effects of government taxes, transfer payments and child-care subsidies on discretionary time. Figure 11.1 shows that all regimes (and indeed all countries under discussion) reduce the ‘gender gap’ in discretionary time, pre- to postgovernment. We see that from the fact that the size of the blackened portion of the bar is smaller, post-government than pre-government, for all regime types and for each country. The reductions are greater in some countries than others, however, and those differences correspond only imperfectly to regime types. This can be seen more clearly in Figure 11.2. There, the first bar for each country and each regime type shows the ‘gender gap in pregovernment discretionary time’; the second column shows the ‘gender 3

It is also true for women, but not men, in Australia – but by only 6 minutes a week, compared to a half-an-hour or an hour a week for men and women respectively in Sweden.

A temporal perspective on gender regimes

181

gap in post-government discretionary time’. A final column for each country and regime shows ‘state impact on the gender gap in discretionary time’, the difference between the gender gap post-government and pre-government. The average gender gap in pre-government discretionary time is generally around 3 hours a week in the countries under study, with two notable exceptions. It is very much higher (6.5) in France, and it is much lower (just under 2) in Sweden. In general, the effect of state taxes, transfers and child-care subsidies reduces that gender gap in discretionary time, pre- to post-government, by an hour or two a week in liberal and corporatist regimes. In socialdemocratic regimes state impact is less: the gender gap is reduced by only just over a half-an-hour a week in Sweden and a mere ten minutes a week in Finland. The upshot is that, post-government, the gender gap in discretionary time remains high in France (5.5 hours a week) and Finland (3.26 hours a week), compared to the other countries under study. Elsewhere, the gender gap is generally between 1 and 2 hours a week. Germany’s postgovernment gender gap is lowest, at just over half-an-hour a week. Thus, there are no easy generalizations across regime types, as regards the gender gap in discretionary time. The picture is particularly mixed in the corporatist cluster, which contains both the best and worst performers: in Germany, the public policies under investigation reduce an already modest pre-government gender gap by the most of any country under study (over 2 hours a week); in France, where the pre-government gender gap is highest of any of the countries by a wide margin, the public policies under investigation make only an average-sized dent in it. The picture is almost as confused in the social-democratic cluster: what is true of both countries there is that the impact of taxes, transfers and child-care subsidies on the gender gap in discretionary time is low (10 to 34 minutes a week); but the smallness of social-democratic-state impact matters more in Finland (where the gender pre-government discretionary time was relatively high) than in Sweden (where it was already the lowest of any country under study). So what is the ‘big-picture’ message? We have seen in Figure 11.1 that both women and men have more discretionary time, post-government, in social-democratic regimes than elsewhere. That is mostly because both men and women started out with high amounts of pre-government discretionary time there; and social-democratic taxes, transfers and

182

Gender regimes matter

child-care subsidies either increase that (in the case of Sweden) or (in the case of Finland) do not decrease it as much as often happens elsewhere. But, as we have seen in Figure 11.2, gender equality in post-government discretionary time is not necessarily higher in social-democratic regimes than elsewhere (Sweden is second-best, but Finland second-worst, among the six countries under study). Still, reverting to Figure 11.1, recall that in terms of absolute amounts of post-government discretionary time women (and men too) are clearly better off in either of the two social-democratic countries than they are in any of the others under study.

11.2 Gendered impacts on parents As we have seen in chapter 10, one of the major differences among various regimes of public policy concerns their treatment of mothers. So let us now disaggregate men and women into those with and without children, and investigate what various regimes do to increase or decrease the discretionary time of each group. Figure 11.3 reports the ‘state impact on discretionary time’ (that is, the difference between ‘pre-’ and ‘post-government discretionary time’) for each of our countries, and for each type of gender regime averaging across the relevant countries. As we see in Figure 11.3, the state impacts 8

0.96 0.83 –2.00 –1.49

–2.90 –2.46

–3.80 –3.44

1.96 1.75

1.38 –4.79 –4.39 –3.64

–6

–3.29 –3.17

–0.41

0.05

0.48

0.28

–1.79

–4

–4.04 –3.78

–2

–3.14 –2.74 –1.98

0

–0.42 –0.33 –1.07

0.11

0.19

2

–1.78 –1.53 –1.52

hours per week

4

2.95 2.67

6

–8 liberal countries

US

Australia

men without kids

corporatist Germany France countries

women without kids

men with kids

social- Sweden Finland democratic countries women with kids

Figure 11.3 State impact on discretionary time, by gender and parental status

A temporal perspective on gender regimes

183

are often negative; that is to say, it is often the case that the state’s taxes cost people more discretionary time, on average, than the state’s transfers and child-care subsidies give to them in discretionary time. Looking across all the countries and all the regime types, that is invariably the case for both childless men and childless women. In all our countries, the impact of state taxes and transfers (child-care subsidies are of no benefit to the childless) is to reduce the discretionary time of childless people of either gender. The reduction is generally around 3 hours a week – a little more in Germany, a little less in Sweden, a lot less in Australia. It is generally a little less for childless men than for childless women. But the state impact on discretionary time of the childless of either gender is consistently negative, wherever we look. The pattern is almost as consistently positive for women with children. (The one exception is Germany, where state impact on mothers is slightly negative, but not nearly so negative as on fathers or the childless.) The magnitudes really matter here, however. Whereas the liberal regimes give mothers only a little more than 10 minutes a week more discretionary time, social-democratic ones give between 50 and 160 minutes a week (in Finland and Sweden respectively). The social-democratic regimes are also distinctive in having a positive impact on the discretionary time of both fathers and mothers. Indeed, in both social-democratic countries the impact of the state policies we are examining here is even more strongly positive on fathers than mothers, by a small margin. In liberal regimes in contrast, while mothers’ discretionary time is increased (very slightly), that of fathers is noticeably decreased as a consequence of these public policies. The corporatist cluster once again presents a mixed picture. France looks as if it is closer to the social-democratic ideal, increasing the discretionary time of mothers by even more than Finland (and even increasing discretionary time of fathers, but by a minuscule amount compared to the two social-democratic countries.) Germany, in contrast, looks like an exaggerated liberal regime in its treatment of parents: its impact on the discretionary time of fathers is almost twice as negative as that of the US; and its impact on the discretionary time of mothers is negative (whereas even in the US it is essentially neutral, but very slightly positive). Decomposing state impact on mothers’ discretionary time allows us to see to what extent different sorts of policy – the state’s tax-andtransfer system or its child-care subsidies – are responsible for producing the overall state impacts reported above. The results of that

184

Gender regimes matter

8 6

1.47

1.44 1.23

0.40 1.35

1.63

1.49

0.03 0.25

1.56

1.71

0.98

2

–0.64

–0.25

–1.91

–4

–1.60

–2

–1.08

0 –0.79

hours per week

4

–6 –8 liberal countries

US

Australia

corporatist Germany France countries

taxes-and-transfers

social- Sweden Finland democratic countries

child-care support

Figure 11.4 State impact (decomposed) on mothers’ discretionary time

decomposition are shown in Figure 11.4. There, the first bar for each country and each regime type represents the impact of state taxes and transfers on mothers’ discretionary time, and the second bar represents the impact of child-care subsidies on mothers’ discretionary time. The net overall impact reported previously (in Figure 11.3) is simply the sum of those two separate impacts. The most striking thing in Figure 11.4 is the consistent pattern, across virtually all the countries under study, of child-care subsidies increasing mothers’ discretionary time by around an hour-and-a-half a week. (Australia is the odd-country-out, increasing it only a quarterhour a week.) Differences in their child-care subsidies are not what distinguish different countries’ treatment of mothers. Those differences arise instead from differences in countries’ tax-and-transfer systems. And, as we see in Figure 11.4, those differences – at least as regards mothers – do not seem to track conventional regime types. The impact of US taxes-and-transfers on mothers’ discretionary time is more similar to Germany’s than Australia’s; France’s is more similar to Finland’s than Germany’s; and the two social-democratic countries’ taxes-and-transfers affect discretionary time of mothers in literally opposite directions (Sweden’s increasing it, while Finland’s decreases it, both by a noticeable amount). Thus, decomposing state impact does not help us see any general patterns very clearly.

A temporal perspective on gender regimes

185

11.3 Gender-regime impact on mothers in different household types

4.73 1.10

0.94 1.29

6.57

5.65

5.38 4.85

2

–0.16

–0.80

–2.59

–0.73

–1.69

–4

–0.42 –0.04

–2

–2.04 –0.16

0 –1.23 –0.10

hours per week

4

2.52

3.95

6

2.03 1.48

6.92

6.18

8

6.55

To better detect regime differences, we need crucially to disaggregate by household type. We continue to focus on the discretionary time of mothers, and we continue to focus on the impact of state taxes, transfers and child-care subsidies on that. But now we will be looking at what the impacts of those policies are on mothers in each of the three principal household types: in other words, we distinguish between mothers in dual-earner couples, mothers in single-earner couples and lone mothers. Figure 11.5 displays the state impact on the discretionary time of mothers in each of those household types. Disaggregating in that way, we see a much more nearly consistent regime picture emerging. Both liberal regimes increase the discretionary time of lone mothers substantially; both are essentially neutral in their treatment of mothers in single-earner households; and both have a negative impact (albeit much larger in the US than Australia) on discretionary time of mothers in dual-earner households. Both social-democratic regimes have a strong positive impact on discretionary time of lone mothers and, to a much lesser extent, of mothers in single-earner households. But only Sweden has the sort of

–6.32

–6 –8 liberal US Australia countries

mothers in dual-earner couples

corporatist Germany France countries

social- Sweden Finland democratic countries

mothers in single-earner couples

lone mothers

Figure 11.5 State impact on mothers’ discretionary time, by household type

186

Gender regimes matter

strong positive impact on mothers in dual-earner couples that regimetype discussions have led us to expect among social democrats. Finland is essentially neutral towards them (state impact there costs such mothers less than 10 minutes a week). True to type, both the corporatist regimes increase the discretionary time of mothers in single-earner couples. Admittedly, that effect is twice as strong in France as in Germany; but Germany still increases such mothers’ discretionary time some 70 per cent more than any non-corporatist country under study. The two corporatist countries are alike, too, in their negative treatment of mothers in dual-earner couples, decreasing their discretionary time (substantially in Germany, but noticeably even in France). Where the two corporatist regimes pull dramatically apart is in their treatment of lone mothers. In France, the impact of state taxes, transfers and child-care subsidies is to increase discretionary time of lone mothers by almost 5 hours a week. In Germany, the impact of such policies is to decrease it by over 6 hours a week. That corporatist regimes might split on this issue should come as no surprise. As foreshadowed in section 10.5, lone mothers pose something of a puzzle for corporatist regimes which sharply distinguish their treatment of ‘breadwinners’ and ‘caregivers’. A lone mother is both at once. How a corporatist regime treats her will crucially depend on whether it sees her as a ‘breadwinner’ or a ‘caregiver’. From the evidence of Figure 11.5, Germany treats lone mothers more like ‘breadwinners’, more like men in Figure 11.3 (in exaggerated fashion). France, in contrast, seems to treat them as ‘caregivers’. All of those discussions of ‘state impacts’ deal exclusively in terms of the impact of taxes-and-transfers and child-care subsidies, of course. Much of what is here coded as ‘pre-government discretionary time’ is the product of state interventions of other sorts. Some states do a better job of enforcing wage equality than others, and some states do a better job of enforcing child-support and alimony obligations on noncustodial parents. Insofar as states do either, much less both, lone mothers have to put fewer hours into paid labour to achieve a minimum income in those places, and they have more discretionary time in consequence.4 4

There is no systematic regime-based pattern to these phenomena however. The proportion of a lone mother’s household’s necessary income to avoid poverty, net of child-care costs, provided by alimony and child-support payments from her ex-partner is: 8.85 per cent in the US, 4.36 in Australia, 9.50 in France, 13.60 in Sweden and 5.73 in Finland (the LIS data do not report alimony and child-support

A temporal perspective on gender regimes

187

taxes-and-transfers

France

social- Sweden democratic countries

employed mothers stay-at-home mothers

–1.50

1.34 0.00 0.00

1.00 1.03 0.00 0.00

1.18 0.00 0.00 –0.25

0.81 0.00 –1.61

0.29 0.00 –2.87

corporatist Germany countries

employed mothers stay-at-home mothers

Australia

employed mothers stay-at-home mothers

US

employed mothers stay-at-home mothers

employed mothers stay-at-home mothers

liberal countries

employed mothers stay-at-home mothers

–2.24

0.55 0.00

0.29 0.00 0.00 –0.71

1.33 0.00 0.00

employed mothers stay-at-home mothers

–8

employed mothers stay-at-home mothers

–6

employed mothers stay-at-home mothers

–4

–3.37

0 –2

0.81 0.00 0.00

2

–2.04

hours per week

4

2.82

4.08

6

5.34

8

Finland

child-care support

Figure 11.6 State impact (decomposed) on coupled mothers’ discretionary time, by employment status

11.3.1 Gender-regime impact on coupled mothers Let us look separately, now, at mothers in couples and (in the next section) lone mothers. Among mothers in couples, we further separate mothers who are employed in paid labour from those who are not (‘stay-at-home mothers’).5 Figure 11.6 disaggregates state impact on the discretionary time of mothers in each of those categories, according to policy instrument (taxes-and-transfers or child-care subsidies). Looking at state impacts in this more disaggregated way gives an even clearer image of regimes at work. We expect social democrats to encourage mothers to enter the paid-labour force. Figure 11.6 shows that both social-democratic countries under study do so, albeit weakly:

5

payments separately in Germany). The mean wage rate for lone mothers, as a percentage of the mean wage rate for all workers, is: 65.88 per cent in the US, 88.45 in Australia, 93.97 in Germany, 81.19 in France, 86.73 in Sweden and 83.59 in Finland. A few mothers in single-earner couples are themselves the earner, of course; so the category of ‘stay-at-home mothers’ is slightly different from that of ‘mothers in single-earner couples’, although the two overlap very substantially.

188

Gender regimes matter

both provide them with child-care subsidies worth over an hour a week of discretionary time. Sweden’s tax-and-transfer system gives them another hour in addition to that, while Finland’s tax-and-transfer system takes a little more than that away from them. But at least the encouragement that regime logic leads us to expect social democrats to provide to dual-earner mothers is present (if perhaps not as strongly as regime logic would lead us to expect) in both countries’ child-care provisions. Conventional regime logic leads us to expect corporatists to be hostile to mothers going out to work. Figure 11.6 shows that both our corporatist countries are: neither gives them much help with child care, and the effect of the tax-and-transfer system in both countries is negative on them, making net state impact negative overall. And, of course, just as regime logic leads us to expect, both corporatist regimes generously support stay-at-home mothers through child-care subsidies, in the form of the kindergarten and the e´cole maternelle. Conventional regime logic leads us to expect liberals to be largely indifferent to gender concerns. And Figure 11.6 suggests they are. They do nothing either way to increase or decrease the discretionary time of stay-at-home mothers. And when mothers go out to work, liberal regimes tax them way more than any child-care subsidies they give them.

11.3.2 Gender-regime impact on lone mothers Next let us perform the same ‘decomposition’ exercise on state impact on the discretionary time of lone mothers. That is done in Figure 11.7. We know from Figure 11.5 that, overall, both liberal and socialdemocratic regimes increase substantially the discretionary time of lone mothers. What Figure 11.7 reveals is that they do so in very different ways. The difference in choice of policy instruments is most dramatic among liberal regimes. Both countries increase discretionary time of lone mothers by a whopping 6 to 7 hours a week, but in the US this is accomplished almost wholly through child-care subsidies, whereas in Australia it is accomplished almost wholly through taxes-and-transfers. Social-democratic countries similarly differ in their choice of policy instruments. The impact of the Swedish state overall is to increase discretionary time of lone mothers by about as much as in the liberal countries, and Finnish impact is only a little less than that. But in Finland that result is accomplished wholly through child-care subsidies (with the

A temporal perspective on gender regimes

189

6.46 3.70 2.87

4.66 0.99

3.54 1.31

0.29

0.80

1.16

2

0.20

0 –2

–1.73

–1.53

hours per week

4

2.98 3.57

6

5.76

5.97

8

–4

–6.60

–6 –8 liberal countries

US

Australia

corporatist Germany France countries

taxes-and-transfers

social- Sweden Finland democratic countries

child-care support

Figure 11.7 State impact (decomposed) on lone mothers’ discretionary time

tax-and-transfer system clawing a little back), whereas in Sweden that result is accomplished by child-care subsidies and taxes-and-transfers in almost equal measure. The corporatist case was a mixed story from the start, with overall state impact being negative in Germany and positive in France. But it is interesting to note that both those countries have child-care subsidies that increase discretionary time of lone mothers modestly (very modestly in Germany, to be sure, and far from dramatically even in France). The difference in the overall impact of these two countries on lone mothers’ discretionary time arises almost wholly from the differing impact of their tax-and-transfer systems. In Germany, that penalizes lone mothers severely, whereas in France it is helpful (fully half as helpful as in Germany it is harmful). In a way, the natural comparators to lone mothers (who in our sample are all in paid labour, remember) are mothers in two-earner couples who are in paid labour. The question is whether the tax-andtransfer system, or the system of child-care subsidies, treats working mothers differently depending on whether they are alone or partnered. The relevant data have already been displayed in Figures 11.6 and 11.7, but for ease of comparison those two cases are plotted alongside one another in Figure 11.8.

liberal countries US Australia corporatist countries Germany

child-care support taxes-and-transfers

France Finland

Sweden

socialdemocratic countries

2.98 3.57 –3.37

1.33 0.20

lone mothers mothers in dual-earner couples

–0.71

lone mothers

mothers in dual-earner couples lone mothers mothers in dual-earner couples

–2.24 –1.53 –2.87

0.29

5.76

0.55 0.80 0.29 0.29

–1.61

lone mothers

mothers in dual-earner couples

5.97

1.16

lone mothers –6.60 mothers in dual-earner couples

8

0.81

lone mothers mothers in dual-earner couples

6

4

2

–2.04

0

–2

–4

–6

–8 mothers in dual-earner couples

0.81 1.31

–0.25

lone mothers

1.18 0.99

mothers in dual-earner couples

1.00 1.03

3.54

4.66

3.70 2.87

lone mothers mothers in dual-earner couples

–1.50

lone mothers

–1.73

1.34 6.46

Figure 11.8 State impact (decomposed) on working mothers’ discretionary time, lone mothers versus partnered mothers

hours per week

A temporal perspective on gender regimes

191

As is clear from Figure 11.8, in virtually no case does the state taxand-transfer or child-care system treat working mothers basically the same, whether they are alone or partnered. Just occasionally they treat lone working mothers worse than coupled working mothers: Germany for example taxes them a lot more, and Finland a little. But usually they treat them better, typically a lot better, either in respect of their taxand-transfer system (in the US, Australia, France and Sweden) or of their child-care system (in the US, Australia, France, Sweden and Finland). In short, countries treat employed lone mothers differently, and typically far better, than employed coupled mothers. Furthermore, it does not look from Figures 11.7 and 11.8 as if the ‘choice of instrument’ – taxes-and-transfers versus child-care support – in itself dictates the overall outcome. Countries that assist lone mothers primarily through the tax-and-transfer system (Australia) assist them to about the same extent as those that assist them primarily through child-care support (US, Finland) and as those that assist them through a mixture of the two (Sweden, France).

12

Gender regimes and temporal autonomy

Gender regimes involve a whole lot more than the state’s policies with respect to taxes, transfer payments and child-care subsidies. Still, the way in which those policies are gendered is an important if incomplete fact about the gender regime of any given country. What we have found is that countries differ significantly in their gendering of those policies, and those differences form recognizable patterns. Those patterns fit broadly, but not perfectly, within traditional ‘regime’ categories. Our findings suggest a new set of cross-cutting dimensions that students of gender regimes might want to add to those conventional ones. In summary form, our findings can be stated as follows: Women have most temporal autonomy in social-democratic regimes. They have 5 or 6 hours a week more discretionary time there than they do in liberal or corporatist regimes, respectively. Once again, however, that is largely due to factors outside the scope of our study. In all regimes, women have less temporal autonomy than do men. Generally their discretionary time is less than men’s by 2 to 3 hours a week. The outliers – in both directions – are countries conventionally categorized as traditional-conservative ‘corporatist’ regimes (5.5 in France, 0.5 in Germany). That gender gap would everywhere be greater if it were not for state taxes, transfers and child-care subsidies. By how much differs, however. Liberal regimes reduce the gender gap by over an hour, social-democratic ones by less than half-an-hour. Within the corporatist cluster, countries once again differ. In France the gender gap pre-government is large, and the state does little to reduce it (by only 1 hour); in Germany the gender gap pre-government is modest, yet the state does far more to reduce it (by fully 2 hours). Across all regimes, the net effect of state tax, transfer and childcare policies is to reduce the discretionary time of the childless of 192

Gender regimes and temporal autonomy

193

both genders. The discretionary time of women with children, however, is generally increased by those same policies – by a reasonable amount in social-democratic countries (an hour-and-threequarters a week), by a minuscule amount (11 minutes a week) in liberal countries. Corporatist countries once again differ, with France increasing mothers’ discretionary time (by over an hour-and-a-half a week) while Germany decreases it (by a little under half-an-hour a week). There is substantial difference across regimes which types of mothers they help, as well as how much they help them. Liberals help lone mothers exclusively, and they help lone mothers in paid labour a lot – increasing their discretionary time by over 6 hours a week. Social democrats help lone mothers in paid labour (by around the same amount in Sweden as in the US, a little less in Finland) and mothers in single-earner couples (to a much lesser extent, a little over 1 hour a week). Corporatists help coupled stay-at-home mothers by around 4 hours a week (more in France, less in Germany); but within the corporatist cluster, there is sharp divergence in the treatment of lone mothers in paid labour, with France helping them greatly (almost 5 hours a week) and Germany hurting them by even more than France helps them. There is also substantial difference across regimes (but also sometimes substantial difference within conventional regime clusters) as regards how they help women. In corporatist countries, using state-provided child care in the form of the kindergarten and e´cole maternelle increases the discretionary time of stay-at-home mothers with male breadwinners by 3 to 5 hours a week (in Germany and France respectively: for once both fit the same pattern). In social-democratic countries, child-care subsidies increase discretionary time for mothers in dual-earner couples by over an hour a week (although in Finland the tax-transfer system costs them a little more than that, whereas in Sweden the tax-transfer system gives them as much yet again). In liberal countries, lone mothers gain around 6.5 hours a week from the state, but from different aspects of state policy (the tax-transfer system in Australia, child-care subsidies in the US). How does all this sit with traditional categorizations of gender regimes? Well, there is some evidence of the regime types that that literature leads us to expect.

194

Gender regimes matter

 The two conservative corporatist regimes are indeed traditional ‘male-breadwinner’ sorts of places, judging from the support they provide stay-at-home mothers in such households.  Social-democratic regimes are supposed to be female-friendly ‘dualearner’ sorts of places, and it is true that both Sweden and Finland help working mothers with child care. But the help they provide them is only around a quarter of what corporatists provide for stayat-home mothers; and only Sweden helps ‘dual-earner’ mothers through its tax-transfer system as well (in Finland, the net effect of that on their discretionary time is actually negative). So only Sweden genuinely fits the ‘dual-earner’ pattern conventionally associated with social-democratic regimes, and even then the help it provides ‘dual-earner’ mothers (around 2.5 hours a week) is only half that given by the most staunchly corporatist regime (Germany) to its stay-at-home mothers.  The liberal regime is supposed to be individualist and marketoriented, and hence pretty much indifferent to gender concerns; true to that form, liberal regimes do both less than other regimes to penalize women without kids and less to help married women with kids. All that broadly fits the conventional regime stereotypes. Our findings do point strongly to the need to disaggregate the corporatist regime type, however. There is a striking division within the corporatist grouping over the treatment of lone mothers. The ‘conservative corporatist’ regime has two underlying values: (a) traditionalism as regards household structure; and (b) pro-natalism as regards population policy. Those two values dovetail nicely in cases of intact families. But they pull in different directions in the case of lone mothers. In Germany, that tension is resolved in favour of traditionalism; in France it is resolved in favour of pro-natalism. So when it comes to lone mothers, Germany treats them as ‘breadwinners’ (taxing them heavily and not giving them much help with child care) while France treats them as ‘caregivers’ (providing them with substantial help both financially and with child care). In terms of our proffered taxonomy of gender regimes in Table 10.2, all this broadly confirms four out of six of our tentative placements of countries in those cells. The evidence of Figure 11.5, however, suggests the need for two of those countries to shift one cell to the left. In light of

Gender regimes and temporal autonomy

195

Table 12.1. Fundamental divides in gender regimes with respect to mothers (revised) Treatment of married mothers Encourage to stay at home Neutral

Treatment of lone mothers

Supportive (general familyoriented) Supportive (employmentoriented) Sanctioning

US

France

Encourage into paid labour

Australia Finland

Sweden

Germany

Figure 11.5, the US looks like it is not ‘neutral’ with respect to married mothers but instead provides incentives for them to stay at home and out of the paid-labour market. And Finland looks as if it is ‘neutral’ with respect to married mothers, rather than encouraging them into paid employment. The revised table appears as Table 12.1.

PART V

Household regimes matter

13

How household regimes differ

Choices that people make within their households can dramatically affect their temporal autonomy. We have already seen how, when two single individuals merge into a common household, each of them gains in discretionary time. We have seen that when they have children, both (but mothers more than fathers) sacrifice much autonomy over their time. We have seen that when they divorce, the partner left with the children suffers even greater losses in discretionary time. The differences that household choices make to discretionary time can often vastly exceed any difference that governments make. To get some sense of that, let us contemplate a ‘toy example’. It is not quite right, in ways that will be discussed shortly and rectified subsequently. But it is close enough to serve the present purpose, which is merely to demonstrate how intra-household arrangements might matter for discretionary time. Here is the example. Imagine a single, childless woman. She lives in the US at the moment, but is contemplating moving to Sweden. Judging from Figure 13.1, such a woman could on average expect to enjoy 2.5 hours more discretionary time a week if she moved to Sweden. But suppose she decided instead to stay in the US, to marry and have children, and then to divorce keeping all responsibility for the kids. Doing so would cost such a woman on average 29 hours of discretionary time. In other words, her choice about the shape of her household might be over ten times as consequential for her temporal autonomy as is her choice of countries. Up to this point in the book, we have been analysing household choices only in terms of people’s choices over what type of household to live in: whether to remain single, to marry or to divorce; whether married partners both continue in paid labour; whether to have children or not; and so on. In the chapters that follow, we focus on choices people face concerning the rules under which their household is run. Those rules – or 199

200

Household regimes matter

‘household regimes’, as we shall sometimes call them – were once fixed in cement, perhaps. But increasingly they are up for negotiation within the household.1 Different ‘household regimes’ represent alternative ways of apportioning responsibility among adults in the household for tasks necessary to the household’s functioning. There are many such tasks, of course.2 Given the orientation of this book and the constraints imposed by the data with which we are working, we focus upon just the two major ones: earning income and homemaking. Shifting from one ‘household rule’ to another alters the amount of ‘necessary time in paid labour’ and ‘necessary time in unpaid household labour’ that each adult in a household will have to do. That, in turn, changes the amount of overall temporal autonomy (‘discretionary time’) that each adult in the household enjoys. In chapter 14 we calibrate the magnitudes of those effects. Public policy might exacerbate the difference that those household choices make in some places, while ameliorating it in others. Chapter 15 explores the possibility that the choice among alternative ‘household rules’ might matter more in some countries than others to people’s discretionary time. To get some preliminary – and again not-quite-right – sense of the possible magnitude of those differences, return to our example of the single, childless American woman. Figure 13.1 shows how much discretionary time she would have, going through the various life changes outlined above, if she lived her life in Sweden rather than the US. The really big change we see in Figure 13.1, of course, comes with divorce. That is bad for the woman’s discretionary time, whether she is in the US or Sweden. But it is much worse for her in the US than Sweden. A Swedish mother who divorces and keeps all responsibility for her children loses on average 18 hours’ discretionary time from her previously partnered dual-earner state, compared to her American counterpart’s 32.3

1 3

Lewis 2001. 2 See, e.g, Lasch 1977. In the other transition too – between being single and becoming a mother in a dual-earner couple – the average woman in Sweden is better off than her American counterpart, albeit by a lesser amount. Swedish women on average gain over 6 hours a week in discretionary time from that transition, American women half that.

How household regimes differ

201

100 90 80

86.14 77.29

80.54

79.82

68.18

hours per week

70 60 48.21

50 40 30 20 10 0

single women

mothers in dual-earner couples US

lone mothers

Sweden

Figure 13.1 Changes in discretionary time through the life cycle, US versus Sweden (first approximation)

This ‘toy example’ gives some sense of how much government might matter to the temporal consequences of one’s choice regarding what ‘type’ of household to form (partnered or single; child or not; etc.). Chapter 15 will show how much government might matter to the temporal consequences of one’s choice among alternative ‘rules’ on which to run one’s household, allocating responsibilities within the household one way rather than another.

13.1 Alternative household rules: a broad overview We shall describe more fully and precisely each of the various alternative household rules we have in mind below. (For more detail see Appendix 1.) But let us begin with a broad overview of the basic options. The most old-fashioned way of allocating responsibilities is to do it on a strictly gendered basis: the man of the house is responsible for bringing home the bacon, the woman is responsible for cooking it. That, in essence, is the ‘male-breadwinner rule’ we discuss below. As an efficiency-oriented economistic variation on this theme, we might allocate breadwinning responsibilities to whichever partner in the

202

Household regimes matter

household can earn the most money in the least time (i.e., has the higher wage rate). That is usually but not invariably the man. Alternatively, we can be ‘modern’ and send both partners out to work, dividing responsibilities for both earnings and housework between them in whatever proportion is standard among dual-earners in their country. As we have noted in chapter 10, those conventional dual-earner arrangements give rise to gender inequalities, because women conventionally continue to do a disproportionate share of the housework even when they also engage in paid labour. If we take offence at that, we might go ‘egalitarian’, thinking about various ways that responsibilities for both income and housework might be shared equally between partners in the household. What ‘egalitarian’ might mean here is a slightly open question. No doubt it involves doing an equal share of unpaid household labour – putting in the same hours over the stove and the sink and the toilet bowls. But in terms of paid labour, it might mean either of two things: bringing in the same amount of money; or spending the same amount of time earning money. Those two interpretations of the egalitarian-household ideal will differ whenever partners’ wage rates differ (as typically they do, somewhat). Finally, households break up. If there are no children involved, dissolving a household is simple (at least in terms of the limited range of things we can here take into account, which conspicuously excludes both financial and human capital). When dissolving a childless marriage, partners simply once again become single individuals, responsible for themselves on the same basis as they would have been had they never married. If there are children of the marriage, however, it crucially matters how responsibility for the children – both for their financial support and for their day-to-day care – is allocated between the divorcing parents. We discuss a range of different Divorce rules for allocating those responsibilities which (as chapter 14 will show) have dramatically different consequences for the temporal autonomy of divorced parents. To some extent, the choice among such alternative household regimes is a choice that is most commonly, naturally and properly left to households themselves. That is the way we most typically think of the choice between being a male-breadwinner or a dual-earner household, for example. Even there, however, the incentives for arranging

How household regimes differ

203

the household one way rather than another are affected by the tax and other policies of one’s government (as we shall discuss in chapter 15). When it comes to the rules governing household breakdown, however, public policy naturally plays a major role. While there is still considerable scope for household-level choice, what responsibility divorced parents have (financially and otherwise) for the children of their previous marriage is shaped more by public law and less by private choice. Even if divorcing partners manage to reach an agreement without going to court, any bargaining that they do will be conducted in the ‘shadow of the law’.

13.2 Breadwinner rules Anyone living alone has to be both breadwinner and homemaker for his or her little one-person household. There is no one else with whom those responsibilities can be shared. The same is true, we assume, for lone-parent households: any tasks that are literally necessary to keep the household functioning are, we assume, entirely the responsibility of the lone adult. In households with two adults, the question of how responsibilities should be shared between them does indeed arise. The first pair of ‘household rules’ we shall discuss imposes a sharp division of household labour, assigning all responsibility for paid labour to one person (the ‘breadwinner’) and virtually all responsibility for unpaid household labour to the other (the ‘homemaker’). The two regimes differ basically just on the basis upon which those responsibilities are assigned: gender, in the case of the Male-Breadwinner rule; the partners’ relative wage rates, in the case of the Most-efficient Breadwinner rule.

13.2.1 Male-Breadwinner rule The Male-Breadwinner model represents a sharply gendered division of labour, with the male doing all of his household’s paid labour and the female doing virtually all of its unpaid household labour. This is often supposed to represent the ‘traditional’ pattern of social life.4 Depending on how far back we go in time, that may not be 4

Cf. Lewis 1992; 2001, ch. 3; Orloff 1993; Fraser 1994; Land 1994; Sainsbury 1996, esp. chs. 3–4.

204

Household regimes matter

completely true. In primitive societies women often share in productive (albeit unpaid) labour, gathering food while the males go hunting; and in agricultural families, women often help with the crops (typically without pay), particularly at harvest time. Still in modern industrialized urban societies, the ‘male-breadwinner’ pattern definitely prevailed. In the US, for example, in 1890 less than 5 per cent of women in intact couples were also in paid labour, and as late as 1950 the figure was still only around 20 per cent.5 Across all the countries under study here, something similar used to be true.6 In specifying the Male-Breadwinner rule for allocating household responsibilities, we assume that people live in shared households, with the male breadwinner bearing complete financial responsibility for keeping the household out of poverty. Others in the household bear no responsibility for meeting the household’s income needs. If the wife or children earn anything at all, it is just ‘pin money’ to be spent on discretionary purchases rather than basic household necessities.7 In our specification of the Male-Breadwinner rule, the stay-at-home wife is assigned the role of ‘homemaker’. She has primary domestic responsibilities and is expected to contribute the bulk of the household’s unpaid labour. The more extreme rule that we will consider next assigns the breadwinner literally no responsibility for unpaid household labour. But our ‘Male-Breadwinner’ rule assigns him a little. It assumes that households allocate chores between husband and wife according to broad social conventions defining some tasks as ‘women’s work’ (e.g. cooking, cleaning and child care) and others as ‘men’s work’ (e.g. mowing and taking out the trash). How much time they spend on them is also, we assume, socially conventional.

5 7

US Bureau of the Census 1965, pp. 15, 72. 6 Esping-Andersen 1999, p. 49. In the words of the Beveridge Committee’s famous report, ‘the housewife’s earnings in general are a means, not of subsistence but of a standard of living above subsistence’ (Beveridge 1942, para. 108; quoted in Land 1994, p. 101). Twenty years later, a sociological study of Bermondsey found that still was true: ‘The way in which her [i.e., a working wife’s] wages were spent showed that the woman worked neither to meet basic economic needs nor to provide personal pleasures for herself. Money was wanted as a means of raising the family’s standard of living. It was used to build up, on a do-it-yourself basis, a more modern and attractive home, to provide more generous food, better footwear and larger wardrobes, to buy durable consumer goods, to give the family a seaside holiday, and to acquire a cheap second-hand car’ (Jephcott 1962, p. 165, quoted in Young and Willmott 1973, p. 102).

How household regimes differ

205

Given that characterization of the Male-Breadwinner rule, we can operationalize our various ‘necessary time’ variables as follows:

 Under this ‘household rule’ – and all others we shall discuss – each person’s ‘necessary personal care time’ is four-fifths of median actual personal care time across the entire population of the country.  The Male-Breadwinner rule assigns complete responsibility to the male for earning an income sufficient to keep the household out of poverty. Hence, ‘necessary time in paid labour’ for the wife is zero; and for the husband it is the poverty line for his household as a whole, divided by his wage rate.  Finally, the Male-Breadwinner rule assigns responsibilities for ‘necessary unpaid household labour’ in whatever way is socially conventional in each country. We surmise those conventions by calculating the mean proportion of the household’s total unpaid household labour that is actually done by men within our restricted sample in each country who actually live in ‘male-breadwinner’ households (that is, households with an employed male head and a non-employed female spouse). Our Male-Breadwinner rule then allocates to each partner that proportionate share of the total unpaid labour time that we deem ‘necessary’.

13.2.2 Most-efficient Breadwinner rule ‘Breadwinner’-style models in general are based on the idea of a sharp division of labour within the household. Mention ‘division of labour’ to economists, and they think immediately of Adam Smith and the efficiency gains in the allocation of resources that can come from a division of labour. The ‘resource’ being allocated by household rules of the sort under discussion here is the labour time of the two adults in the household. An efficient allocation of that labour time would be this: the partner who is best at paid labour (which, to an economist, means ‘is paid most, per hour’) should allocate all of his or her labour time to paid labour; and the partner who is best at unpaid household labour (which, to an economist, means ‘can do the chores with least effort’) should allocate all of his or her labour time to unpaid household labour. At least that is how an economist should think about these things. That is to say, economists ought be concerned with maximizing efficiency of the household’s resource (time) allocation overall. Where, as

206

Household regimes matter

might sometimes happen, the same partner is more efficient at both unpaid as well as paid tasks, then the economist should recommend that the same partner would be assigned entire responsibility for both (up to the point that that person’s efficiency declines below the partner’s).8 That is the way truly to minimize the total amount of time the household as a whole spends in paid plus unpaid household labour. The Most-efficient Breadwinner rule that has actually entered the literature, however, omits any consideration of partners’ relative efficiency at unpaid household tasks.9 It simply assigns all the responsibility for paid labour to whichever partner has the higher wage rate, and all responsibility for unpaid household labour to the other partner. That model is flawed, both logically in ways that we have just noted and sociologically as well: evidence suggests that in couples where the female is the Most-efficient Breadwinner in the couple, she definitely compensates for the embarrassment of the role reversal by not dropping her unpaid household labour.10 Still, flawed though it might be, here we follow Becker in attributing zero responsibility for unpaid household labour to the Most-efficient Breadwinner – partly because Becker’s approach has become influential in the literature, but also partly because empirically there is no way in our data to assess the relative efficiency of partners’ efforts at unpaid household labour. Allocating responsibilities according to this rule allows the household as a whole to maximize its income, within a fixed number of paid labour hours. That is Becker’s primary focus. Allocating responsibilities in that way also allows the household as a whole to minimize the number of hours it needs to put into paid labour to reach the poverty line, and in that way to maximize the discretionary time available to the household as a whole – which is of course our focus here.11 8

9 11

Becker 1981, ch. 2 contemplates this in Theorem 2.3 but concludes that, ‘If commodity production functions have consistent or increasing returns to scale, all members of efficient households would specialize completely in the market or household sectors and would invest only in market or household capital’ (Theorem 2.4, see similarly Theorem 2.5). Becker 1981. 10 Bittman et al. 2002. Again, both the Most-efficient Breadwinner and/or a second earner in the household may, and typically will, put in more hours of paid work than that. But here again, it seems proper to regard their working those ‘extra’ hours, over and above what was strictly necessary for the household to escape poverty, as discretionary acts – one of the ways in which people choose to use their discretionary time.

How household regimes differ 40

37.00

35 30

207

35.86

32.30 29.86 26.81

per cent

25

25.12 22.06

20 15 10 5 0

all countries

US

Australia

Germany France

Sweden Finland

Figure 13.2 Proportion of wives with higher wage rates than their husbands

Virtually no real-world households are run literally on Most-efficient Breadwinner rules. True, role reversals – with ‘house-husbands’ taking over homemaking tasks and ‘working wives’ breadwinner responsibilities – are found increasingly (albeit still relatively infrequently) across the countries here under study. And in many (but certainly not all) those cases, the motivation underlying those role reversals might be to take advantage of the wife’s greater earning power. But in virtually no household will a breadwinning wife do literally no unpaid household labour, as Becker’s model prescribes.12 Still, as we have said, Becker’s proposal is a theoretically influential one and it might become ever more so in an increasingly market-liberalized world. Where the male in the household has a higher wage rate than his wife, the Most-efficient Breadwinner rule is almost the same as the ‘Male-Breadwinner’ one.13 That is typically the case, but far from invariably so. As Figure 13.2 shows, the wife’s wage rate exceeds that

12

13

Such evidence as we have suggests rather the opposite: at least in couples where both partners work in paid labour, the woman reduces her unpaid household labour less for every hour she spends in paid labour, where the woman’s wage rate exceeds her husband’s (Bittman et al. 2002). ‘Almost’, because the Male-Breadwinner rule assigns the breadwinner a little ‘necessary time in unpaid household labour’ whereas the Most-efficient Breadwinner rule assigns him or her none.

208

Household regimes matter

of her husband in around 30% of households, across all six countries. The proportion ranges from a high of 37% in Australia to a low of 22% in Finland. In specifying the Most-efficient Breadwinner rule for allocating household responsibilities, many of our assumptions are similar to those for the ‘Male-Breadwinner’ rule. We assume that people live in shared households, with one person (the ‘breadwinner’) bearing complete financial responsibility for keeping the household out of poverty. We assume, once again, that others in the household bear no responsibility for meeting the household’s income needs. (Any paid labour by the non-breadwinner or children is purely for ‘pin money’ spent on discretionary purchases rather than providing basic household necessities.) The Most-efficient Breadwinner rule differs from the Male-Breadwinner rule, however, in which partner is assigned this ‘breadwinning’ role. On the Most-efficient Breadwinner rule, this role is assigned to whichever of the partners has the higher hourly wage rate. In the Most-efficient Breadwinner rule, the partner with the lower wage rate is assigned the role of ‘homemaker’ with complete responsibility for working the total amount of time in unpaid household labour that is necessary in his or her household. Other members of the household bear no responsibility for any of the unpaid household labour that is strictly necessary in their household. Given that characterization of the Most-efficient Breadwinner rule, we operationalize the ‘Most-efficient Breadwinner’ rule as follows:

 ‘Necessary personal care time’ is as before: four-fifths of median actual personal care time across the entire population of the country.

 The Most-efficient Breadwinner rule assigns all responsibility for paid labour to whichever partner has the higher wage rate. As before, ‘necessary time in paid labour’ for that person is just the poverty line for the household as a whole, divided by his/her wage rate. For the other partner, ‘necessary time in paid labour’ is zero.  ‘Necessary time in unpaid household labour’ is set to zero for the partner who has been assigned responsibility for paid labour. ‘Necessary time in unpaid household labour’ for the other partner equals the entire amount of unpaid household labour time that is necessary for the household as a whole.

How household regimes differ

209

13.3 Conventional Dual-earner rule In the Conventional Dual-earner model, both women and men engage in paid labour. Both also engage in unpaid household labour. How much responsibility each of the partners has in each respect depends upon social conventions in the country in which they live. Whereas the Male-Breadwinner pattern used to be the ‘traditional’ one, this Conventional Dual-earner pattern is an increasingly common one in today’s world. The proportion of intact couples in which the wife works in paid labour is around 70% in the liberal countries under study, the US and Australia. It is 10 or even 20% higher than that in Finland and Sweden respectively;14 it is 10% lower in Germany and France.15 The proportion is also higher for women without young children than with (by how much again varying across countries16). Still, across all of the countries under study the dual-earner pattern is now the most common and is becoming increasingly so.17 In specifying our Conventional Dual-earner rule for allocating household responsibilities, we assume that people live in shared households. Both partners share financial responsibility for raising the income necessary to keep their household out of poverty. For purposes of the Conventional Dual-earner rule, we assume that responsibility is shared between partners in some proportion that is socially conventional in their country. Both partners also share responsibility for the unpaid household labour that is socially necessary in their household; and that too is, in the Conventional Dual-earner rule, assumed to be 14

15

16

17

The higher labour-force participation rates among Scandinavian women, including especially ones with relatively low wage rates, may help to explain the slightly surprising finding in Figure 13.2 that a lower proportion of Scandinavian wives have higher wage rates than their husbands, compared to other countries. Among couples in our restricted sample of households (both husband and wife aged between 25 and 54 years, etc.), dual-earner couples make up the following percentages in the surveys we are using: 69% in the US; 72% in Australia; 62% in Germany; 60% in France; 93% in Sweden; 83% in Finland. The employment rate among mothers of children under 3 is strikingly low in Finland (32.2%) and Australia (45.0%); it is well above the OECD average of 57.5% only in Sweden (72.9%) and France (66.2%) among the countries under study here (OECD 2005b, table SS4). In the US, for example, the proportion of households in which both the husband and wife work in paid labour more than doubled over the middle years of the twentieth century (Esping-Andersen 1999, p. 49, citing Padavic and Reskin 1994, p. 145).

210

Household regimes matter

shared between partners in some proportion that is socially conventional in their country.18 Given that characterization, we operationalize the Conventional Dual-earner rule as follows:

 ‘Necessary personal care time’ is as before: four-fifths of median actual personal care time across the entire population of the country.  The Conventional Dual-earner rule assigns responsibility for earning the amount of money sufficient to keep the household out of poverty to partners in the household in the same proportion as their actual contribution to the household’s actual total earnings. ‘Necessary time in paid labour’ is calculated accordingly for each partner by dividing that amount of money by each partner’s own actual wage rate.  The total amount of time in unpaid labour that we deem is necessary for the household as a whole is divided between the two partners in the household in the same proportions as men and women divide actual time in unpaid household labour, on average, in actual two-earner couples within our restricted sample across the country as a whole.

13.4 Egalitarian rules Next we consider a pair of more ‘egalitarian’ scenarios for allocating responsibilities within the household.19 Partners in a household ‘living as equals’ has long been advocated as a philosophical ideal.20 The ‘symmetrical household’ is an ideal that is rarely implemented perfectly in practice, no doubt, even among those

18

19

20

The variation is less dramatic among the particular countries under study than some others in the OECD, however (Esping-Andersen 2002, p. 92, citing OECD 2001, table 4.5). In the six countries here under study, the ratio of female to male actual time in unpaid household labour ranges from a low of 1.44 in Finland to a high of 1.95 in Australia among dual-earner couples without children in our restricted sample (aged 25 through 54, etc.); and from a low of 1.62 in Finland to a high of 2.23 in Australia among dual-earner couples with children. In explaining why they prefer the term ‘symmetrical household’, Young and Willmott (1973, p. 31) write, ‘The new family could be labeled simply egalitarian but that would not square with the marked differences that still remain in human rights, in the work opportunities and generally in the way of life of the two sexes.’ We take this point. But as a description of what goes on within the household, and the rules on which it is run, ‘egalitarian’ seems to us a perfectly good description. By philosophers ranging from Mill (1869) to Sen (1992; 1996).

How household regimes differ

211

(still far from a majority, in many places) who are trying to practise it.21 Still, there is increasing movement in that direction – again, more so in Scandinavia than in other countries here under study.22 Both of the Egalitarian regimes we shall consider below share many features in common. Both of them presume that adults live in shared households, and that they do so ‘as equals’. Both hold each of the partners equally responsible for the household, financially and otherwise. Spending extra time on those activities, over and above what would be required for them to make equal contributions to meeting their household’s basic needs, is under this ‘household rule’ treated as a discretionary act on their part. In terms of unpaid household labour, both sorts of ‘egalitarian household rules’ would require partners to share all the necessary household tasks. Partners are required, under both forms of ‘egalitarian household rules’, to divide the work time it takes to perform them equally. Each partner has to do strictly half of their household’s total ‘necessary time in unpaid household labour’. That requirement is common across both ‘egalitarian household rules’ discussed below.23 The two sorts of ‘egalitarian household regimes’ differ only as regards how responsibilities for paid labour are shared between partners in the household. Egalitarianism rules of either sort hold each partner ‘equally responsible for meeting the financial needs of the household’, in some sense. But there are two possible ways of defining, and discharging, an ‘equal responsibility’ in that respect. Therein lies the difference between our two different ‘egalitarian household rules’ elaborated below.

13.4.1 Equal Monetary Contribution rule Under the Equal Monetary Contribution model, partners’ ‘equal responsibility’ for the financial needs of the household is understood in monetary terms. Each of the two partners is assigned responsibility for earning exactly half of the income that would be required to keep 21

22 23

For sociological accounts see Young and Willmott 1973 and Goodnow and Bowes 1994. Esping-Andersen 1999, ch. 4; 2002, ch. 3. In common with most feminist discussions of these issues: but an alternative economistic formulation for the Equal Monetary Contribution rule is discussed in a subsequent footnote.

212

Household regimes matter

the household as a whole out of poverty. In this version of the egalitarian model, the cash contributions of each adult are what is equalized.24 Note, however, that equalizing in terms of monetary contributions leads to inequalities between partners in the amount of time the two partners will have to spend in paid labour to raise that amount of money, wherever the two partners have different wage rates (as typically they will). The lower-waged partner will have to spend longer in paid labour than will the higher-waged partner in order to generate the same sum. The Equal Monetary Contribution rule can thus be operationalized as follows:

 ‘Necessary personal care time’ is as before: four-fifths of median actual personal care time across the entire population of the country.

 The Equal Monetary Contribution rule assigns equal responsibility to each partner for the unpaid household labour necessary for their household as a whole. ‘Necessary time in unpaid household labour’ is for each partner exactly half of that total amount.  Each partner is responsible for bringing in half the total income that would be required for their household as a whole to have a povertylevel income. Each partner’s ‘necessary paid labour time’ equals half the poverty line for the household as a whole, divided by that person’s own wage rate.

13.4.2 Equal Temporal Contribution rule The Equal Temporal Contribution rule is identical to the Equal Monetary Contribution rule, except it interprets the equal responsibilities of partners to engage in paid labour in temporal rather than monetary terms. 24

By economistic extension, it might be thought that an Equal Monetary Contribution rule ought to require that each partner’s share of ‘necessary unpaid household labour time’ ought to be equalized, not in the sense of requiring an ‘equal number of hours’ from each, but rather in the sense of ‘equally valuable hours’ from each, where the value of each party’s unpaid-time contribution was fixed by the relative wage rates of the two partners. Requiring the lower-waged partner to work more (perhaps substantially more) hours in unpaid labour than the higher-waged partner would, however, drive this model towards the Mostefficient Breadwinner model. The present way of specifying the egalitarian ‘necessary time in unpaid household labour time’ has the virtue of keeping those two household rules maximally distinct. It also captures the sort of ‘specific egalitarianism’ (Tobin 1970) ways of thinking about separate ‘spheres of justice’ (Walzer 1983): in the ‘cash’ sphere, you equalize cash contributions; in the sphere of ‘housework chores’, you equalize time devoted to such chores.

How household regimes differ

213

On the Equal Temporal Contribution rule, just as members of egalitarian households have a responsibility to work equal amounts of time in necessary unpaid household tasks, so too are they responsible for working equal amounts of time in paid labour to meet their household’s basic needs. Of course, if their wage rates differ (as they typically will), different partners’ working equal numbers of hours will yield unequal amounts of money. But what matters so far as the Equal Temporal Contribution model is concerned is equality of time (and of effort or of sacrifice, more generally). The Equal Temporal Contribution rule can therefore be operationalized as follows:

 ‘Necessary personal care time’ is as before: four-fifths of median actual personal care time across the entire population of the country.  Also as before, the Equal Temporal Contribution rule assigns equal responsibility to each partner for ‘necessary time in unpaid household labour’. For each partner, ‘necessary time in unpaid household labour’ is exactly half of the unpaid labour time necessary for the household as a whole.  The Equal Temporal Contribution rule makes the two partners together responsible for bringing enough income to get their household above poverty, and it makes them each responsible for working the same number of hours in paid labour as one another to do so. Operationally, that implies a ‘necessary paid labour time’ for each partner equal to the poverty line for the household as a whole, divided by the sum of the two partners’ wage rates.

13.5 Withdrawal (Divorce) rules Finally, we offer a suite of alternative ‘withdrawal’ or ‘divorce regimes’. Various Divorce rules here serve to specify what would happen if adult members presently sharing a household divorced and set up separate households. That is an increasingly common phenomenon. As we observed in chapter 4, nearly half of all marriages now end in divorce across all six countries under study.25 The terms on which former partners part matter greatly to their subsequent well-being, in all sorts of ways. 25

OECD 2005b, table GE5.3.

214

Household regimes matter

The consequences of particular concern here are for the temporal autonomy that the former partners will subsequently enjoy. And we shall focus more particularly still on the consequences for temporal autonomy of different ways of allocating post-separation responsibility for the care and financial support of the children of the previous marriage. Historically, divorce was once prohibited in all the countries under study. It was permitted briefly after the Revolution in France, and the French Marshal Bernadotte commonly granted divorce by royal decree when he assumed the throne of Sweden.26 But those were aberrations in the more common historical pattern of prohibiting divorce. Over the course of the nineteenth century laws came to allow husbands or wives to petition for a divorce on grounds that he or she had been wronged by the other in certain ways. Ex-husbands were typically required to pay ex-wives alimony for life.27 Children, particularly of ‘tender age’, were typically left with the ex-wife.28 Divorce proceedings in this period were infused with notions of ‘fault’ and ‘blame’. Both child-custody decisions and decisions concerning the financial settlement typically turned on judgments of how ‘heinously’ the ‘innocent spouse’ had been wronged.29 Those practices changed with the ‘no-fault revolution’. That came earlier in some places than others.30 By the period under study, however, it was an accomplished fact across all the countries we shall be discussing.31 Whereas older-style divorce proceedings were guided by punitive principles organized around notions of ‘guilt’ and ‘innocence’, no-fault divorce proceedings are officially guided by principles of ‘equity’. Feminists rightly complain that the ‘equity’ is more formal

26 27

28

29

30

31

Eekelaar 1991, pp. 11–12, citing Stoljar 1989 and Phillips 1988. In part as a reflection of the view that the lifelong obligation of husbands to support their wives financially was traditionally held to persist even after the marriage was dissolved (Honore´ 1982, pp. 62–3; Weitzman 1985, p. 145). Although they would have been, in a still-earlier era, regarded as literally the ‘property’ of their fathers. Weitzman (1985, p. 12) cites a 1946 judgment of a California appellate court, for example, holding that ‘the greater the offense the larger the proportion of the community property that must be awarded to the innocent spouse’. To Sweden in 1915 (Eekelaar 1991, p. 12); to the US in 1970 with the initial California reform, and spreading to other states over the next decade (Weitzman 1985, p. 20). Therborn 1995, pp. 105–9; Castles and Flood 1993, pp. 300–1.

How household regimes differ

215

than substantive, typically in ways outside our purview in this study.32 One way that is highly relevant to our subsequent discussion, however, is this: even after the presumption of maternal custody of young children was officially replaced by notionally gender-blind rules of one sort or another, a vast majority of young children still ended up living with their mothers (because that is what both parents prefer).33 Under no-fault rules, however, the ex-wife is expected financially to ‘become independent and self-sufficient after divorce’ (so is awarded only ‘transitional’ alimony at most).34 ‘Similarly, the new equality between the spouses’ at the heart of the no-fault reforms ‘is reflected in child support standards: the new law makes both husbands and wives responsible for child support’.35 Strategies for allocating that ‘equal responsibility for child support’ vary. ‘An ‘‘equal living standards’’ approach [would] seek to ensure that the child’s household remains as well off as the absent parent’s.’ A ‘cost-sharing’ approach, as for example in California, ‘attempts to divide the expenses of raising a child between the parents in rough proportion to their financial ability’. An ‘income-sharing’ approach – which is the ‘most widely accepted model’ across the US – sets the support award at a level reflecting a specified percentage’ of the income of either (or both, as in the case of Australia) the non-custodial or the custodial parent.36 The Divorce rules we shall be discussing are highly stylized variations on all these. Or anyway on a subset of them: due to limitations of the data available to us, the specific rules we shall discuss differ from one another in how they allocate responsibility for the children of the previous marriage, and in that respect alone. Hence, where there are no

32

33

34 35 36

The marital assets to be distributed equitably upon divorce ought (but under initial no-fault rules typically did not) to include ‘past and future labor-force earnings, benefits and government entitlements of each spouse . . . Distributive decisions should depend on the parties’ domestic as well as economic contributions to the relationship and on their future needs and earning potential’ (Rhode and Minow 1990, p. 199; see further Weitzman 1985; Fineman 1991). In which circumstances laws that officially give both parents ‘joint custody’ ‘have given men equal authority but not equal responsibility’, by ‘forc[ing the] . . . mothers to consult with and gain the father’s approval on all major decisions about their children’ (Weitzman 1985, p. 361; see also pp. 37, 49). Weitzman 1985, p. 33. Weitzman 1985, p. 36. On the British case, see Lewis 2001, ch. 5. Rhode and Minow 1990, pp. 205–6; Parker 1990; Edwards 2001.

216

Household regimes matter

children involved, all of our various Divorce rules are extensionally equivalent. Different ‘withdrawal’ scenarios can be thought of as differing from one another depending in respect of what (here, what responsibilities for the children of the marriage) adults take with them when they leave a previously joint household.37 A ‘withdrawal rule’, thus construed, constitutes the baseline for bargaining within households. It defines what game theorists call the ‘non-agreement point’: that which each party could get on his or her own, should they be unable to agree on some other arrangement more advantageous to both. How good each party’s fallback position is, compared to the other’s, determines in turn the relative ‘threat advantage’ that each has relative to the other. In standard bargaining models, the benefits that each party can achieve by withdrawing determine the relative bargaining power of each and hence the share that each receives of the benefits of both not withdrawing.38

13.5.1 Atomistic rule The Atomistic rule specifies what would happen if each divorcing partner were truly to ‘go it alone’, leaving the children (if there are any) behind with his or her ex-partner. The Atomistic rule thus also represents what would happen if single individuals had never married or had children: thus it serves both as a model of divorce and (in its ‘Atomistic, no children’ form) as a model of never-marrieds. The Atomistic model invites us, in effect, to contemplate a world of singlemember households. Of course, if everyone did literally withdraw to single-member households there would be no one with any responsibility for the children, who would presumably be left to starve in consequence. Social reproduction would cease; society (indeed, humanity) would end within the space of now-existing generations. Remember, however, the very particular role that these various scenarios are supposed to be serving in our thinking. They are being 37

38

As in Roemer’s (1982) analysis of ‘exploitation’; Goodin (2008) analyses the exploitability of women’s investments in the household in similar terms. Luce and Raiffa 1957, ch. 6. For an application to the family context, see Lundberg and Pollak 1994.

How household regimes differ

217

used to specify how much time it is strictly ‘necessary’ to devote to various tasks, and to what extent the time we spend on tasks ought to be regarded as ‘discretionary’. The ‘Atomistic’ scenario thus is not strictly antithetical to social reproduction. It merely insists that the time devoted to coupling and child-raising ought to be counted as ‘discretionary time’ – as something that people choose to do with their spare time, rather than something that is strictly necessary for them to do, from this ‘Atomistic’ perspective. Children, and coupling more generally, are regarded within the Atomistic model as ‘consumption goods’.39 In Nancy Folbre’s evocative phrase, it invites us to think of ‘children as pets’.40 According to the Atomistic rule’s account of what is ‘necessary’, then, each individual would be responsible for raising just enough money through his or her paid labour to keep himself or herself out of poverty. And each individual would be responsible for spending just the amount of time on unpaid household tasks as would be necessary in a household comprised of himself or herself as its sole member. Operationalizing the Atomistic rule thus characterized:

 ‘Necessary personal care time’ is as before: four-fifths of median actual personal care time across the entire population of the country.

 Under the Atomistic rule each person’s ‘necessary unpaid household labour time’ is just the amount of time necessary for running a oneperson household in his or her country.  Each person’s ‘necessary paid labour time’ is the poverty line for a one-person household, divided by that person’s own wage rate.

13.5.2 Self-reliant Divorce rule Whereas the Atomistic model is at risk of neglecting the kids, the second ‘withdrawal’ scenario might seem, at first blush, to be at risk of counting them twice over. Under the Self-reliant Divorce rule, each adult is deemed fully responsible for taking care of himself or herself, together with all the household’s children. Under this ‘withdrawal rule’, each adult would have to take all the household’s children (if there are any) when he or she leaves. Adults are ‘self-reliant’, in the 39 40

As in Becker 1981. Folbre 2001, ch. 5. Or as she put it in a more sober moment, as ‘consumer durables’ (Folbre 1982, p. 325).

218

Household regimes matter

sense here in view, only insofar as they have the capacity to take care of themselves and all of their dependents.41 Of course, the children cannot actually be full-time in the new, separate households of both their divorced mother and father, simultaneously. Remember what these Divorce rules are mapping, however: what is ‘necessary’ for people to meet their responsibilities, not their actual expenditures of time or money. What the Self-reliant Divorce rule is insisting, in ascribing all the children of the marriage to both divorced parents’ separate households, is that each parent remains fully responsible for his or her children. What each partner needs to be ‘selfreliant’, in the sense of being capable of fully discharging those responsibilities, is thus what it would take to meet the needs not only of himself or herself but also of all the dependents for whom he or she is responsible. In operationalizing this Self-reliant Divorce rule:

 ‘Necessary personal care time’ is as before: four-fifths of median actual personal care time across the entire population of the country.  Under the Self-reliant Divorce rule, each person’s ‘necessary unpaid household labour time’ is the amount of time necessary for running a household comprised of himself or herself and as many children as there are in his or her shared household at present.  Each person’s ‘necessary paid labour time’ is the poverty line for a household of that size, divided by that person’s own wage rate.

13.5.3 Breadwinner Divorce rules When households break down, perhaps the most natural assumption would be that the division of responsibilities governing the separation would mimic the division of responsibilities that had been previously governing the operation of the intact household. In households running on Breadwinner rules, one of the partners is designated ‘breadwinner’ and is responsible for all the necessary 41

This sense of ‘self-reliance’ has infused the rhetoric of welfare reform worldwide: see, e.g.: US Congress 1996; Haveman and Bershadker 1998; cf. Goodin 1998. From a very different political perspective, feminists often speak in favour of a ‘universal breadwinner’ model (Fraser 1994), in which women were able to support all their dependents in ‘an autonomous household’ (Orloff 1993).

How household regimes differ

219

paid labour of the household. The other partner is designated ‘homemaker’ and is primarily responsible for all the necessary unpaid labour of the household. The Male-Breadwinner rule makes that allocation on the basis of gender; the Most-efficient Breadwinner rule makes that allocation on the basis of which of the partners has the higher wage rate. When Breadwinner households break down, the ‘closest continuer’ of that rule would be for the partner who had been the ‘homemaker’ in the intact household to continue in that role. That person would get custody of any children of the marriage; and s/he would be responsible for doing as much unpaid labour as necessary for a household composed of him/herself and as many children as the couple have. The person who had been the ‘breadwinner’ in the intact household would continue in that role; s/he would be responsible for doing as much paid labour as necessary to keep both his/her own household and his/her former partner’s household from falling into poverty (in addition to doing whatever unpaid household labour is necessary for his/her own one-person household). Operationalizing the Breadwinner Divorce rule, then:

 ‘Necessary personal care time’ is as before: four-fifths of median actual personal care time across the entire population of the country.

 Under the Breadwinner Divorce rule, ‘necessary unpaid household labour time’ for the ‘breadwinner’ is the amount of time necessary for running a one-person household. For the ‘homemaker’, ‘necessary unpaid household labour time’ is the amount of time necessary for running a household comprised of the homemaker and as many children as there are in his/her shared household at present.  ‘Necessary paid labour time’ for the ‘homemaker’ is zero. ‘Necessary paid labour time’ for the ‘breadwinner’ is the poverty line for a oneperson household plus the poverty line for a household composed of one person plus as many children there are in his/her shared household at present, all divided by the ‘breadwinner’s’ own wage rate. Breadwinner Divorce rules come in two forms. The Male-Breadwinner Divorce rule assigns the man the role of ‘breadwinner’ and the woman the role of ‘homemaker’. The Most-efficient Breadwinner rule assigns the role of ‘breadwinner’ to the (ex)partner with the higher wage rate and the role of ‘homemaker’ to the (ex)partner with the lower wage rate.

220

Household regimes matter

13.5.4 Gendered Divorce rule The Gendered Divorce rule might be regarded as the ‘traditional’ model of marital breakdown. Under this rule, it is assumed that the husband leaves the marital home and sets up a separate household, leaving his ex-wife and all his children behind, financially and otherwise. The ex-wife is left with full responsibility for both caring and providing financially for herself and her children, while the ex-husband bears responsibility for himself alone. In effect, after a Gendered Divorce, the male lives in an Atomistic household, and the female lives under conditions of Selfreliance. Operationalizing the Gendered Divorce rule, then:

 ‘Necessary personal care time’ is as before: four-fifths of median actual personal care time across the entire population of the country.

 Under the Gendered Divorce rule, ‘necessary unpaid household labour time’ for the male is the amount of time necessary for running a one-person household. For the female, ‘necessary unpaid household labour time’ is the amount of time necessary for running a household comprised of herself and as many children as there are in her shared household at present.  ‘Necessary paid labour time’ for the male is the poverty line for a one-person household, divided by that man’s own wage rate. For the female, ‘necessary paid labour time’ is the poverty line for a household composed of herself and all her children, divided by that woman’s own wage rate.

13.5.5 Financially Egalitarian Divorce rule A more egalitarian variation on the previous model of household breakdown might insist that, even (indeed, especially) if the mother takes full responsibility for caring for the couple’s children, the father should at least shoulder an equal share of financial responsibility for them. The world envisaged by this Financially Egalitarian Divorce rule is thus a world of men living alone, and of children living entirely with their mothers. In this world, the woman performs all the unpaid household labour required to care for herself and her children. But the

How household regimes differ

221

ex-husband and ex-wife split equally the extra costs associated with her caring for the couple’s children in her separate household. As we have seen from our discussion of ‘egalitarian rules’ in section 13.4 above, an ‘equal’ split of those extra costs can be construed in either of two ways. One is in terms of ‘equal financial contributions’: each pays the same amount of money towards the costs of their children. The other is in terms of ‘equal temporal contributions’: each spends an equal amount of time in paid labour to support their children. Insofar as the previous partners’ wage rates differ, those two specifications differ. In Appendix 2 we report the results of both. Here in the main text, however, we will for simplicity’s sake discuss only the ‘equal temporal contributions’ version.42 Operationalizing the Financially Egalitarian Divorce rule, then:

 ‘Necessary personal care time’ is as before: four-fifths of median actual personal care time across the entire population of the country.

 Under the Financially Egalitarian Divorce rule, ‘necessary unpaid household labour time’ for the male is just the amount of time necessary for running a one-person household. For the female, ‘necessary unpaid household labour time’ is the amount of time necessary for running a household composed of herself and as many children as there are in her shared household at present.  To calculate ‘necessary paid labour time’ in a ‘temporally egalitarian’ way, first calculate the difference between the poverty line for a single-person household and the poverty line for a household comprised of the mother and all the couple’s children. That represents the ‘costs of the children’ that the man and woman will share. To calculate the man’s share, divide his wage rate by the sum of both his and her wage rates. Multiply that by the total ‘costs of the children’, to find the amount of money he is to transfer to her. Increase the father’s poverty line (and decrease the mother’s) accordingly. ‘Necessary time in paid labour’ is then that recalibrated poverty line for each, divided by each person’s own wage rate.

42

Child-support obligations, when officially assessed by public agencies (as in Australia for example), are typically specified as some proportion of the parents’ earnings (after some adjustments); they are never, to our knowledge, specified as ‘the same monetary sum from both parents, regardless of their circumstances’.

222

Household regimes matter

13.5.6 Strictly Egalitarian Divorce rule A yet more egalitarian ‘withdrawal rule’ would suppose that both parents should share responsibility for their children, both financially and custodially. Under the Strictly Egalitarian Divorce rule, responsibility would be shared equally, with each being expected to do exactly half the paid labour and exactly half the unpaid household labour required to support the children.43 The simplest way of thinking of this might be as the children spending exactly half their time in each divorced parent’s household.44 Alternatively, we might think in terms of the risk that each parent runs, ex ante in contemplating divorce; from that perspective, each parent might be thought of as running an equal risk of ending up with all the couple’s children in his or her household. Yet another way of thinking of it might be as depicting the equal risk that each ex-partner runs, over the ensuing years, of any or all of the couple’s children coming to live with him or her. Those are ways that the outcome envisaged in the Strictly Egalitarian Divorce scenario might actually come to pass, in the real world.45 But once again, remember that these Divorce rules are first and foremost rules for allocating responsibilities. How children actually 43

44

45

In contrast to the Self-reliant Divorce rule, which assigns each full responsibility, not just half. Or as half the children living full-time with each parent (with some child spending half-time in each, in families with an odd number of children). Note that these real-world ways of implementing the Strictly Egalitarian Divorce ideal might have differing implications for ‘necessity’, in the realm of both money and unpaid labour time. Suppose each of the ex-partners has all the couple’s children in his or her household exactly half the time (alternate weeks, say). In such a scenario, it is unclear where to set the poverty line (and hence how to specify ‘necessary paid labour time’): while the running costs (food, clothing and so on) of having four children half-time might be about the same as having two children full-time, many of the major fixed-costs are very different (you can house two children full-time in one extra bedroom, but to house four even just half-time you probably need two, for example). Furthermore, in many of the MTUS data sets there is no way of estimating how much unpaid household labour time is done (or hence, on our definition, is ‘necessary’ to do) by part-time parents. In the 1992 Australian Time-use Survey, for example, children of separated parents are ascribed to the household in which they are ‘ordinarily’ resident, even if they spend substantial periods of time in the other parent’s household; and on the other parent’s return, the children would appear as mere ‘household visitors’, indistinguishable from any other house guest.

How household regimes differ

223

sort themselves among actual households might be quite another matter. What the Strictly Egalitarian Divorce rule says is that each partner should be seen as responsible for taking care of his or her children halftime, even if that does not (maybe even cannot46) actually happen. The simplest way to operationalize the Strictly Egalitarian Divorce rule is as follows:

 ‘Necessary personal care time’ is as before: four-fifths of median actual personal care time across the entire population of the country.

 Under the Strictly Egalitarian Divorce rule, ‘necessary unpaid household labour time’ is the amount of time necessary for running a oneperson household, plus half the difference between that and the ‘necessary unpaid household labour time’ required for running a household comprised of oneself and as many children as are in the couple’s present shared household.  To calculate ‘necessary paid labour time’, first calculate the difference between the poverty line for a single-person household and the poverty line for a household comprised of a single person plus as many children as there are in the couple’s shared household at present. That represents the ‘costs of the children’ that the two partners will share. The Strictly Egalitarian approach as we are here construing it makes the two partners each responsible for working the same number of hours in paid labour as one another to meet these costs. Operationally, that implies that, for each partner, the time he or she needs to spend in paid labour in order to meet the ‘costs of the children’ is equal to the ‘costs of the children’, divided by the sum of the two partners’ wage rates. To calculate each partner’s total ‘necessary paid labour time’, add to this amount an amount equal to the poverty line for a single-person household, divided by that partner’s own wage rate.47

46 47

Maybe, for example, they live in some other city, now. Once again, this is to construe the egalitarian ideal in a ‘temporally egalitarian’ way. That is the version that will be discussed in the text; for results pertaining to a ‘monetarily egalitarian’ version of the Strictly Egalitarian Divorce rule, see Appendix 2.

14

The difference that household rules make

In this chapter, we will probe the data we have been discussing to determine what difference people can make to their discretionary time through their own household’s choices, by arranging their household on one set of rules rather than another. The statistics we will be reporting in this chapter represent an average across all six countries under study.1 In the next chapter, we will turn to examine differences across these countries in what difference alternative household rules might make.

14.1 Preliminary methodological remarks Up to this point in the book, whenever we have reported on discretionary time for people in certain household types, those reports have been based on the analysis of MTUS and LIS files for people who actually live in that type of household at present. Thus, in reporting the discretionary time available to single mothers in the US, for example, we have been reporting the mean amount of discretionary time among all women who actually are single mothers in the MTUS and LIS data sets for the US in the survey in question. That style of analysis is invaluable for certain purposes but problematic for others. It is problematic, in particular, for answering the sorts of questions with which we are concerned in this part of the book. Next we shall essentially be asking (as we were in our ‘toy examples’ at the beginning of chapter 13) how people’s discretionary time would change if they went from one sort of household to another – if a married mother became a lone mother, for example. 1

Sample sizes differ substantially across the surveys for different countries. To adjust for that, we calculate averages first within each country and then we average those averages across all six countries. The country-specific averages are reported in Appendix 2.

224

The difference that household rules make

225

In our ‘toy examples’ in chapter 13, we offered a preliminary answer to that question by comparing the mean discretionary time of lone mothers with that of married mothers. As we warned, however, that was not quite right. Suppose the average lone mother is importantly different from the average married mother, having a lower wage rate for example. Then just looking at the mean discretionary time of those who are presently lone mothers might be an unreliable guide to how much discretionary time higher-waged married mothers would enjoy were they to divorce. We need to control for all that, if we really want to know – as we do, in this part of the book – how people’s lives would change if they were to arrange their own households differently, with all else being held constant. Here we employ the simplest and most straightforward method for controlling for those confounding effects. We take the whole restricted sample (prime working age, one-family households, etc.) of each country for whom we have all data required for this exercise. We then ask how much discretionary time they would have if they were to (re)arrange their households so that they ran under first one of the alternative household rules, then under another, then under another. In each of those ‘alternative household rules’ projections, it is exactly the same people with exactly the same characteristics who are living in those households.2 Any difference we find among mean discretionary times under alternative household rules really is then a difference due to the difference in ‘household regimes’ and that alone.3 As in our previous calculations of ‘necessary time in paid labour’, so too in our calculations of the consequences of alternative ‘household regimes’: if the whole population really did change simultaneously in the ways envisaged, the macroeconomic consequences would be vast beyond our power to predict. So even though for modelling purposes we are envisaging the whole population changing, the best way to think of our results is as indicating what would happen to the average household in the country if it alone were to make the change. And again, here 2

3

That is to say, in all of those hypothetical projections each person retains the personal characteristics that he or she actually has in the MTUS and LIS data sets: the same wage rate, the same number of children (although in some of our ‘divorce’ scenarios we envisage some people who presently live with children leaving them behind) and so on. At least within any given country: the cross-national comparisons in chapter 15 are still vulnerable to the sorts of concerns raised above.

226

Household regimes matter

as before we will be looking at only the direct and immediate effects of people shifting their way of organizing household responsibilities to another way, rather than any background causes or long-term effects of their doing so.4 Second, note that estimating the effects of some of the alternative ‘household rules’ we describe below requires us to know the wage rate of both adults in the household. We therefore restrict the sample of people under study in this part of the book to couples for whom that information is available in our data sets. Thus, we exclude from the analysis in this part of the book (but it alone) people who are presently living in ‘one-earner couple’ households.5 When we report on the consequences of people living in those sorts of households, what we will be reporting is not what actually happens to people actually living in such households at present. Instead, the report there concerns what would happen, on average, if everyone presently living in dual-earner households were to rearrange their households in that way. We also exclude from the analysis in this part of the book people who are presently living in ‘one-adult’ households (with or without children). Once again, when we report on the consequences of people living in those sorts of households, what is there reported is not what actually happens to people actually living in such households at present. Instead, the report there concerns what would happen, on average, if people presently living as dual-earners (with and without children, respectively) were to rearrange their households in that way. Indeed, that is true of all of our analyses of the consequences of alternative ‘household rules’. Those analyses will always be based on 4

5

In the real world, differences in the rules under which households are run are correlated with differences in their demographic and economic circumstances. Among actual one-earner couples, for example, the earner’s wage rate tends to be lower than the wage rates of people in two-earner couples; and there tend to be more children in households with stay-at-home mothers. But which is cause and which is consequence is unclear. Maybe having higher wages or more children causes households to operate on different rules; or maybe choosing different rules for running the household leads to the household having more or less money or children. Projecting the entire population first into one ‘household rule’, and then into another, brackets out all these selection biases and knock-on consequences of households’ choice of behaviour. Having first noted how people in such households divide ‘time in unpaid household labour’ in each of our countries – a statistic that we will need in operationalizing the Male-Breadwinner rule, in section 14.2.1 below.

The difference that household rules make

227

the mean discretionary time that would be enjoyed by people in our restricted sample of dual-earner couples, were they all to rearrange their households according to each of the various ‘household rules’ in question. Here as before, the in-text reports of our results will typically be in highly aggregated various forms. But in Appendix 2 we give full reports of mean discretionary hours separately for men and women, with and without children, in each country under each of these household regimes. Finally, across all the alternative ‘household regimes’ in which existing households remain intact (i.e., in all the ‘Breadwinner’, ‘Conventional Dual-earner’ and ‘Egalitarian’ scenarios), the main thing that changes from one regime to another is how responsibilities for meeting the household’s overall needs are allocated within the household. ‘Necessary time in personal care’ is the same across all alternative ‘household regimes’. (It is fixed at the amount that is necessary in each country as calculated by the four-fifths median procedure described in section 2.2.4.) Across all those scenarios, the total amount of income that any given household needs remains the same. (That is the amount of income that that household needs to reach the poverty line for money, as calculated in section 2.2.2.) How much time the household needs to spend in paid labour to achieve that differs a little across ‘household regimes’, because taxes-and-transfers may differ depending on whether the household has one earner or two; but those differences are typically not substantial. The one thing that does vary substantially across ‘household regimes’ is the total amount of ‘unpaid household labour time’ that any given household needs to do, which alters (substantially, if children are present) depending on whether the ‘household regime’ involves a stay-at-home adult or not. Our final suite of scenarios involves ‘withdrawal’ rules governing the break-up of households. There, new households are formed as existing ones split. With those changes in the size of the households, there also come changes in the households’ total ‘necessary time in paid labour’ and ‘necessary time in unpaid household labour’. In ‘withdrawal’ scenarios, we adjust both those accordingly. Alternative ‘household rules’ specify how responsibilities for paid labour and unpaid household labour are to be shared between people in two-adult households or, in the case of (most of) the ‘withdrawal’ rules, between the custodial and non-custodial parent now living in separate

228

Household regimes matter

households. The various ‘withdrawal’ rules canvassed below represent what happens in one-adult households. Their implications will all be identical for single-adult households with no children. There, the single adult bears all responsibility for meeting all her own needs for paid labour and unpaid household labour. In single-parent households, what proportion of his or her household’s needs the custodial parent is responsible for depends on how much time and money the noncustodial parent is required to contribute by each of the various ‘withdrawal’ rules.

14.2 Effects of alternative household rules: an overview In assessing the effects of running your household on different rules, let us once again start at a high level of aggregation and work down to some more interesting cases.

14.2.1 Effects on average household discretionary time Figure 14.1 shows the average amount of discretionary time that would be available to adults in a household under each of the major alternative rules for running the household. The statistics reported there average the post-government discretionary time of both adults in the household (or of both of the adults that used to be in the household, for ‘divorce’ scenarios in which households have split into two). The Conventional Dual-earner rule is shown first and set apart in Figure 14.1 as a visual reminder that the people who are being projected first into one and then into another ‘alternative household’ are, at present, all living in that Conventional Dual-earner type of household. Somewhat to our surprise, we find from Figure 14.1 that, among the rules under consideration, the Conventional Dual-earner rule is the one that actually maximizes average discretionary time for the household as a whole. (That is not only true averaging across all six countries: it is true or nearly true in each of the countries separately.6) In the 6

The Conventional Dual-earner rule literally gives households most average discretionary time of all the rules discussed in the US, Australia, Germany and Sweden. In France, it trails the Most-efficient Breadwinner rule but only by less than 2 minutes a week. In Finland it trails both that rule and the Equal Monetary Contribution rule, but only by three-quarters of an hour a week.

81.82 83.44 82.00 84.31

77.60 65.18

70.94 72.02 72.20 72.08 72.07

er os ma t-e le b ffi eq ci read ua w l m ent br inn ea e eq one dw r ua t l t ary in em n c po ontr er ib ra u lc on tion at t r om ib ist utio se ic n lfdi re vo fin lia rc an m m ge nt d e ci al os i n al e v t-e d -b ly o ffi re ered rce str ega ad di lit cien w ic v in tly ar tb ne orc re eg ian e r a al di di dw ita v vo in ria orc r ce e ( ner n di t vo ime divo rc e ( ega rce lit tim ar eeg ian ) al ita ria n)

100 90 84.87 80 70 60 50 40 30 20 10 0

229

m

co n

ve nt

io

na ld

ua l-e ar n

hours per week

The difference that household rules make

Figure 14.1 Effect of alternative household rules on average household discretionary time

Egalitarian rules, discretionary time is of course shared more equally within the household. But even in the better of those rules, the Equal Temporal Contribution rule, the average discretionary time available to the household as a whole decreases by a whisker from that under the Conventional Dual-earner rule. And in the Breadwinner rules (even the so-called ‘Most-efficient Breadwinner’ one), average household discretionary time is another hour less than under the Conventional Dualearner rule. That may seem odd. But remember, the ‘Most-efficient Breadwinner’ is more efficient, in the sense that it yields more discretionary time for the household as a whole, only as compared with assigning the ‘Breadwinner’ role in any other way (on the basis of male gender, for example). Given state tax, transfer and child-care subsidy policies, arranging the household’s affairs on a dual-earner rather than breadwinner basis proves to be an even better way to maximize average discretionary time in the household. And among the dual-earner rules, the Conventional Dual-earner rule (partners sharing responsibilities in the way they do at present) yields the household as a whole a little more average discretionary time than any other way of arranging the dual-earner

230

Household regimes matter

household – again, given existing public policies. This may be little more than a reflection of the fact that governments have arranged their policies in light of the way most people arrange their households. Among intact households (ones not undergoing a divorce), however, the main thing to notice in Figure 14.1 is the similarity in the average amount of discretionary time they enjoy, regardless of the rule on which they are run. The variation among the averages for alternative household rules is only 3 hours a week, on a base of over 80. The best alternative household rule is less than 4 per cent better in terms of the average amount of discretionary time it yields households as a whole. A second and much more important thing to notice in Figure 14.1 is the cost, in terms of average discretionary time across both partners, to households that split. The seven bars to the right in Figure 14.1 (the ‘Divorce’ rules) are all substantially lower than the five (intact household rules) to the left. There are two components to that loss of discretionary time that come from divorcing. The first has to do with the loss of efficiencies of scale that had come from running a joint household. The second has to do with the time-costs associated with children of divorced parents. The first of those components can be assessed by comparing the first ‘Divorce’ rule – Atomistic Divorce – with the bars to its left representing intact households. The Atomistic Divorce scenario, recall, involves both parents going off to set up one-person households with no children present. (Think of this as the ‘put-Moses-in-a-basket’ scenario.) When partners split and form new separate households on this ‘both ditch the kids’ scenario, the average discretionary time that they enjoy is between 4 and 7 hours a week less than they would have enjoyed on average under their previous partnered arrangements (depending on which of the rules had governed their previous joint household). The second, child-related component of the effect of divorce on average discretionary time across the parents can be seen by comparing the Atomistic Divorce bar with the six bars to the right of it. For the moment, let’s ignore the Self-reliant Divorce rule (which in effect double-counts the costs of the children, by assuming that they are in both parents’ households full-time, when in fact they can only be in one place at once). All the Divorce rules other than that yield broadly similar amounts of average discretionary time across the ex-partners taken together (although different rules distribute it very differently indeed, as we shall see shortly). The child-related time-costs of divorce

The difference that household rules make

231

to ex-partners taken together – the shortfall between those five right-hand bars and Atomistic Divorce – amount to between 5 and 7 hours a week. The two components of the time-cost of divorce are thus roughly on a par. One clear implication of that fact is that the time-costs of parents divorcing are twice those of childless couples divorcing. The third ‘big message’ to be derived from Figure 14.1 is how little difference it makes to average discretionary time across the ex-partners which of the (realistic) Divorce rules are applied to them. Ignoring the highly artificial rules that either ignore the children altogether or double-count them, and focusing just on the last five bars of Figure 14.1, the difference that alternative Divorce rules make to average discretionary time across the parents taken together is at most just a couple of hours a week. (Although we add once again, that different rules distribute that discretionary time very differently between the two parents, as will be seen shortly.) The effect of alternative household rules on average discretionary time for the household as a whole can be summarized as follows:

 It makes little difference to a household’s average discretionary time how the household is set up, among intact families.

 It makes a big difference for a household to split up. Part of that difference has to do with the loss of economies of scale, but an equally large part of that difference is due to the impact of children on the discretionary time of divorced parents.  It makes little difference to average discretionary time among divorced people (but, as we shall see, substantial difference to the distribution of that discretionary time among them) which of the withdrawal/divorce rules they employ.

14.2.2 Effects on the distribution of discretionary time within the household So far we have been talking purely at the level of average household discretionary time (or in the case of households that split, of the average discretionary time across partners from the former joint household). We have repeatedly warned that similarities in those averages might mask substantial difference in the distribution of discretionary time within those households (or in the case of divorce, between the households formed by the split). Let us now try to calibrate those differences.

232

Household regimes matter

In Figure 14.2 we disaggregate the statistics in Figure 14.1 by gender. The first bar for each household rule shows the amount of discretionary time that would be enjoyed by men, on average, under that rule; the second shows the amount that would be enjoyed by women. From Figure 14.2 we see that women would actually have more discretionary time than men in only one of the alternative household rules under discussion, the Male-Breadwinner rule. In one other case, the Equal Temporal Contribution rule, they would have equal amounts of discretionary time as men. All other alternative household rules would leave women with less discretionary time. Among all the rules under discussion, the Conventional Dual-earner rule gives men the most discretionary time. That same rule gives women far more discretionary time than any of the ‘Divorce’ scenarios, and noticeably more than a couple of the other rules governing intact households – notably among them the Equal Monetary Contribution rule (by nearly 4 hours a week). The Equal Temporal Contribution rule makes women better off than the Conventional Dual-earner rule, but (surprisingly) only by about an hour a week. Thus, the Conventional Dual-earner rule is not only the best way to maximize average household discretionary time but also comes tolerably close to maximizing the amount of discretionary time of the woman as well as the man in the household. It is in the ‘Divorce’ scenarios where the really big distributional differences emerge. Gendered Divorce – where the divorced mother bears complete responsibility for both the care of the children and their financial support – leads to the greatest inequalities, of course. Under that rule, the divorced mother would have fully 19 hours a week less discretionary time on average than her ex-husband. Other of the ‘Divorce’ rules mitigate that inequality, by making the ex-husband pay half the financial costs of the kids (in the Financially Egalitarian Divorce rule) or all of them (in the Male-Breadwinner Divorce rule). But even under those more ‘progressive’ Divorce rules, the mother still has 8 to 13 hours a week less discretionary time than her ex-partner. (Even if we fully equalize the amount of time both ex-partners spend both caring and financially providing for their children, as in the Strictly Egalitarian Divorce rule, women on average would still have almost 6 hours less discretionary time than their ex-partners purely because their lower wage rates require those women to spend more necessary time in paid labour.)

r ne in w ad re eb

y ar

r ne in w ad e br nt ie

et on m

at

n ic ist om

tio bu tri on lc

n io ut rib nt co

ra po em lt

84.41 79.60

e rc vo di

nt lia re lfse

e rc vo di

ed er nd ge

e rc vo di

rd ne in

ce or iv

rd

it al

n)

tim

e-

ita al

eg

78.41 65.77

ria

ita

e(

rc

vo

di

eg e-

ce

or

iv

n

al

75.75 68.64

m

(ti

ia ar

e rc

ne in

eg

vo di

w ad re n ia ar

tb en

76.23 67.81

tly ic str

lit a eg

ci

dw ea br eal m

ffi t-e os m

80.48

ly al ci an fin

61.40

n) ria

74.94 69.19

Figure 14.2 Effect of alternative household rules on discretionary time, by gender 90 80 70 60 50 40 30 20 0

10

86.46 83.28

r ne

ar

-e

al du

al m

fic ef

86.42 80.46

ua eq

80.48 74.73

100

l na tio en

77.34 86.30

l ua eq

84.31 84.31

women

nv

tos m

68.97 61.40

men

co

hours per week

234

Household regimes matter

14.3 The impact of alternative household rules on gender equality What appear as distributional differences within the household (or between partners who used to share a household) when writ large show up as gender inequalities across society as a whole. We can examine that issue by looking at the ‘gender gap’ – the average difference between post-government discretionary time available to men and to women – under each of the alternative household rules. Figure 14.3 displays that, for each of the household rules under discussion. The gender gap among households operating under Conventional Dual-earner rules (as the households in the subsample upon which we perform these operations all actually do) is just over 3 hours a week, in favour of men. Only one of the alternative rules under discussion would literally reverse the gender gap, leaving men with less discretionary time than women: that is the Male-Breadwinner rule. Another of the alternative rules would eradicate the gender gap, leaving men and women with identical amounts of discretionary time: that is the

40 hours per week (men minus women)

30 19.07

20 10

5.96 4.80

3.18

0 –10

5.74 7.56

8.42 7.11

12.64 5.75

0.00

–8.95

–20 –30 re ffi ad ci w en in tb lm ne r on ea r dw et eq ar in ua y n lt co er em nt rib po ut ra io lc n on at t om rib ist uti o ic se di n lfvo re rc lia e nt fin g d an m m en i v a ci os or de le al t-e ce -b re ly re d ffi eg a d c dw str iv ie al or nt ita ic in ce tly ne br ria ea rd n eg d di i al w v v ita or in ce ne ria orce rd n (ti di iv m vo or erc ce e ( ega lit tim ar eia n) eg al ita ria n)

eb al m

eq ua

t-e os m

co

nv

en tio na

ld

ua

l-e

ar

ne

r

–40

Figure 14.3 Gender gap in discretionary time under alternative household rules

The difference that household rules make

235

Equal Temporal Contribution rule. All the other alternative rules would increase the gender gap. That is true of the two remaining alternative rules for intact households, and all the more true of all the Divorce rules. The much more dramatic differences in gender inequality come with divorce. The ‘worst’ of the intact household rules (the Most-efficient Breadwinner rule) gives rise to a gender gap that is only very slightly worse than the best of the Divorce rules (Atomistic Divorce and Strictly Egalitarian Divorce). The ‘worst’ of the Divorce rules (Gendered Divorce) creates a gender gap that is over six times greater than Conventional Dual-earner households experience at present, and over three times greater than the ‘worst’ of the rules governing intact households. Even the ‘best’ of the Divorce rules (Strictly Egalitarian Divorce) almost doubles the gender gap in discretionary time of Conventional Dual-earners when they are coupled, and is roughly on a par with the ‘worst’ of the rules governing intact households. The bottom line is clear: on whatever terms it occurs, divorce is bad for gender equality.

14.4 Alternative household rules and the paternity penalty Alternative household rules might make more or less difference, depending on whether the household has children or not. In Figure 14.4 we explore that possibility. Having children costs parents around 10 hours a week, pretty much whatever the rules on which their households are run. (Ignore Atomistic Divorce in which neither of the separating parents takes the kids; ignore Self-reliant Divorce where both do; all the others hover around that mark.) The ‘paternity penalty’ is a little lower under most intact household rules than under most rules for governing divorced ones (averaging across both divorced parents, again: we will disaggregate that next). The two Breadwinner rules reduce the ‘paternity penalty’ a little further. But among the ‘realistic’ rules governing divorce – represented by the last five bars in Figure 14.3 – the ‘paternity penalty’ is remarkably constant. How that penalty is distributed between ex-partners varies greatly, of course, as we have shown in section 14.3. But averaging the discretionary time of the two ex-partners, they are pretty consistently 10 hours a week worse off upon divorcing than would be a couple without kids.

236

Household regimes matter 40

hours per week (non-parents minus parents)

30 20.63

20 9.45

10

7.95

10.68

7.29

11.92

9.57

10.27

10.19

9.96

10.21

1.81 0 –10 –20 –30

n)

n) ita

ita

ria

ria

ce

al

al

iv

eg etim

e( rc vo di

di

n

n

ria

ria

ita

ita

al

al str

ic

tly

eg

eg ly al

ci

ee(

rc vo

tb en ci

ffi t-e fin

an

tim

w ad re

ea br eal

m os m

eg

rd ne in

ne in dw

nd ge

or

ce or

e

rd

vo di ed er

lia re lf-

iv

rc

rc

e

vo

rc

di nt

di ic ist

om at

ra po

e

n vo

bu tri

lc

on

nt co y ar

et

em lt ua

eq

ua eq

tio

io ut

w lm

on

ci ffi t-e

os m

rib

in

ne in

ad

w

re

ad

tb

re

en

eb al m

se

r ne

r

r ne ar l-e ua ld na tio en nv co

n

–40

Figure 14.4 Paternity penalty under alternative household rules

14.5 Alternative household rules and custodial versus non-custodial divorced parents In the previous section we have been focusing on the differential impact of alternative household rules on people with and without children. In the alternative rules involving intact households, that is a clean comparison. The statistics we have just reported for them reflect unambiguously the costs (in discretionary time) of having children, under each rule for organizing a household. In some of the alternative rules involving divorce, however, it is ambiguous what ‘having children’ might mean. It might mean ‘having children in your present (separate, post-divorce) household’; or it might mean ‘having children by your previous marriage (albeit living in your ex-partner’s household)’. There are good reasons we might be interested in examining the fate of people who ‘have children’ in either sense, compared to those who do not, under alternative withdrawal/ divorce rules. But they are different comparisons. The previous discussion (section 14.4) has addressed the second question. In it, we have asked how much discretionary time people who are presently in Conventional Dual-earner households with children would have, compared to people in such households that do not

237

40 30

26.71

20

17.21

16.90

10.46

10 0 –10 –20 –30

dw di inn vo e rc r e

or ce iv

re a

rd

tb

ne in

ci

en

dw

ci

os

fin

m

al

t-e

ly

ffi

ea br eal m

an

vo di er ed nd ge

eg (ti alita m ri e- an eg d al iv ita or ria ce n)

–40 rc e

hours per week (non-custodial parents minus custodial parents)

The difference that household rules make

Figure 14.5 Gap in discretionary time between custodial and non-custodial parents under alternative divorce rules

have children, under each of the alternative household rules. But suppose the partners divorce on terms (such as Gendered Divorce, for example) that leave one partner with the kids full-time and the other not at all. The amount of discretionary time that such a rule gives to ‘people with children’ – understood as ‘with children somewhere but not necessarily in my household’ – averages across the two partners in very different situations. One has currently co-resident children, the other does not. We might be, and in the case of Divorce rules we probably typically are, more interested in the other sort of comparison. What we really want to know is the differential impact of alternative Divorce rules on households that actually house children and those that do not. So let us now address that issue. Note that only some of the Divorce rules involve one of the households formed by divorcing parents having children (all the time) while the other household does not (any of the time). On the Atomistic Divorce rule, neither of the new households has any children. On the

238

Household regimes matter

Self-reliant Divorce rule, both do. Likewise on the Strictly Egalitarian Divorce rule both do. Thus, it is only the four Divorce rules shown in Figure 14.5 in which we are interested for purposes of the present comparison. Comparing the gaps in Figure 14.5 to those in 14.4 shows the huge difference it makes in how we conceptualize ‘having children’ for the purposes of these comparisons. If we think of it – as we think it should be thought of – as having co-resident children, the ‘paternity penalty’ for custodial parents under the (historically conventional) Gendered Divorce rule is double that we saw in Figure 14.4. It is over 26 hours a week.7 To put that in perspective, the mother having custody of her children on those terms is like her having another half-time job, while her ex-partner plays golf. Even if her ex-partner pays half the costs of the kids (as under a Financially Egalitarian Divorce rule), being the custodial partner costs the woman over 16 hours more a week: think of it as her working Saturday and Sunday while her ex-partner golfs and sleeps in. (Indeed, even if she got her ex-partner to pay all the costs of the kids, she being the custodial parent would still be almost 10 hours a week worse off than her ex-partner.)

7

Remember that only part of the gap shown in these figures is due to the effect of having children; part of it is also due to the fact that women on average have lower wage rates, and hence lower discretionary time.

15

The difference that states make

To paraphrase Marx, households choose, but they do not choose on terms of their own choosing. What the differences are, between running your household on one set of rules versus some other, depends crucially upon choices made by your government and your larger society. Societies encourage some sorts of household choices and discourage others. States reward some household choices while penalizing others. The nature and magnitude of those rewards and penalties differ from one place to another, reflecting countries’ varying preferences in these matters. In this chapter, we attempt to quantify these effects. We try to assess the ways in which differing state policies either ameliorate or amplify the effects of a household’s own choice over the rules on which it is to be run.

15.1 How differences are made: policy instruments and social norms To some extent, choices among ‘household rules’ are more consequential in some places rather than others – and in any given place, some choices are more consequential than others – simply because of social norms and conventions concerning what people in certain sorts of households are ordinarily expected to do. All of the standards of ‘necessity’ that we have been using here are socially relative. How much income we say you need is specified by reference to the poverty line, which (as we define it) is relative to the incomes of everyone else in your country. How much unpaid household labour or personal care you need to do is here defined in ways that are similarly relative to what others in your country are doing. Social conventions come into our calculations at other points, too: in dividing responsibilities for paid and unpaid household labour within ‘Conventional Dual-earner’ households, for example, and in dividing 239

240

Household regimes matter

up unpaid household labour responsibilities within ‘Male-Breadwinner’ households. These social norms and conventions vary across countries, and they impact more heavily on some of the ‘household rules’ under discussion than others. Thus, to some extent, the differences we find among countries in temporal consequences of opting for one ‘household rule’ rather than another might be the result of ‘society’ rather than the ‘state’. But public policy, too, encourages some sorts of household choices and discourages others. The most conspicuous examples are, of course, welfare benefits and tax breaks explicitly designed to assist households of some specific types. Child benefits, family allowances and tax credits for children are all explicitly designed to assist households with children. Dependent-spouse rebates in the tax code are explicitly designed to assist households with stay-at-home partners. In only slightly more subtle ways, joint tax returns that permit income-splitting between waged and non-waged partners favour ‘breadwinner-style’ households in a way that individualized tax returns do not. So do pension and other welfare arrangements that make the ‘household as a whole’ the unit of assessment, and employment history the basis of entitlement.1 Policies designed to promote gender equality, of the sort discussed in chapter 10 above, all also serve to increase the benefits and reduce the costs of running your household on either of the Egalitarian rules we have been discussing. When the state subsidizes child care, that increases parents’ discretionary time by decreasing ‘necessary time in paid labour’ to pay for it. Likewise, when states pursue labour-market policies designed to promote the employment of women in well-paid jobs, and to close the wage gap between men and women, that too increases women’s discretionary time (by decreasing their ‘necessary time in paid labour’ to get a poverty-level income, again).2 Education and housing policies also clearly impact on people’s choices of whether to form families and have children. Subsidies embedded in both affect the costs faced by households, with flow-on effects to the discretionary time of their members. And again, different 1

2

Sainsbury 1996, ch. 2. For a more extreme example, note that in a wave of pronatalist fervour, the Finns began levying a ‘bachelor’s tax’ in 1935 and explicitly enacted a ‘Law on Family Subsidies’ eight years later (Lindgren 1978, pp. 271–2). Esping-Andersen 2002, ch. 3.

The difference that states make

241

sorts of households are differentially advantaged by such policies – and sometimes in surprising ways. Take ‘early childhood education and care’ policies, for example. We ordinarily think of state-subsidized child-care as something that is of particular importance to women in employment, and of course it is.3 But as we have seen in section 11.3.1, the free public e´cole maternelle and kindergarten give stay-at-home mothers in France and Germany a considerable boost in their discretionary time. Those are the sorts of policies, varying from country to country, that might affect the amount of discretionary time people gain or lose in choosing to run their households in one way rather than another.

15.2 Temporal consequences of changing household types In an attempt to calibrate the magnitude of those effects, we shall now turn to examine what difference each country’s policies make to the temporal consequences of the various ‘life choices’ discussed at the beginning of chapter 13. When two single people join together into a dual-earner couple without children, both partners generally gain discretionary time in the process. But how much discretionary time they gain varies from one country to another. So too do the temporal consequences of transforming your household from a dual-earner one to a male-breadwinner one. And so too, most especially, do the temporal consequences of childbearing and of divorce.

15.2.1 Atomistic person to Conventional Dual-earner When a pair of single individuals get together and decide to form a joint household, there are various possible rules on which they could decide to run that household. We have canvassed five such options in the previous chapters. For present purposes, however, let us focus on the most common pattern: the shift from being single persons in ‘Atomistic’ households to partners in a ‘Conventional Dual-earner’ couple. Figure 15.1 shows the temporal consequences of that transition, separately for men and women, in each of our six countries under study. The bars represent the amount of discretionary time, on average, 3

Esping-Andersen 2002, pp. 73–5.

242

Household regimes matter 40

10.78 12.68

12.10 13.60

10.41 14.03

10.73 13.31

12.92 17.41

10

11.16 8.43

hours per week

20

11.35 13.24

30

0 –10 –20 –30 –40 all countries

US

Australia men

Germany France

Sweden Finland

women

Figure 15.1 Temporal consequences of shifting from single Atomistic individuals to Dual-earner couples without children

people in Conventional Dual-earner households have beyond the amount of discretionary time, on average, people in Atomistic households have.4 From Figure 15.1 we can see that partnering – shifting from an Atomistic single-person household to a Conventional Dual-earner one in which both partners still work, but there are not (yet) any children – increases people’s discretionary time by pretty similar amounts across all our countries. That is certainly true for men, who gain 11 hours a week, give or take. Women on average gain 13 hours (4 hours more in the US, almost 5 hours less in Australia). Across all countries, then, partnering yields a huge gain in discretionary time for both parties.

15.2.2 Conventional Dual-earners have children Having children is the next life change that typically befalls people. Again, there are many ways people could arrange their household once they have children. But again, we will here focus upon the pattern that is most common in most of the countries under study: the parents both 4

We have been talking about it as the Atomistic Divorce rule. But of course if, as under that rule, neither person takes the kids (if there are any), then that is the same as the Atomistic one-person household that existed before they ever partnered or had children.

The difference that states make

243

40 30

10

Germany France

–5.58 –8.64

Australia

–6.57 –8.24

–7.75 –10.13

–9.68 –9.36

–20

–10.99 –11.22

–10

–12.53 –12.73

0 –8.85 –10.05

hours per week

20

–30 –40 all countries

US

men

Sweden Finland

women

Figure 15.2 Temporal consequences of shift from Conventional Dual-earner couples without children to Conventional Dual-earner couples with children

remain in (or soon return to) the paid-labour force. Thus, the transition we are examining here is from a ‘Conventional Dual-earner household without children’ to a ‘Conventional Dual-earner household with children’. Figure 15.2 shows the temporal consequences of that transition, for the average man and the average woman, in each of our six countries under study. From Figure 15.2 we can see that having children – and making no other changes in your household arrangements (i.e., both parents remaining in the workforce) – costs people, on average, 9 hours a week of discretionary time, mothers an hour more, fathers an hour less. That basic pattern, too, is remarkably consistent across countries, although magnitudes vary. The cost to people in liberal regimes is a bit more, elsewhere a bit less; and the costs to people (especially men, ironically) in social-democratic regimes is very much less. But again, the basic pattern is basically constant across all countries. Having kids reduces your discretionary time substantially.

15.2.3 Conventional Dual-earner to Gendered Divorce The final sort of household transition to be considered is, of course, divorce. Again, the alternative possible terms on which marriages might

244

Household regimes matter 40 30

10

all countries

US

Australia men

Germany France

–3.40 –22.96

–29.88

–18.27

–23.80

–30.08

–40

–32.91

–30

–26.32

–20

–6.81

–4.35

–2.41

–1.52

–10

–0.72

0 –3.20

hours per week

20

Sweden Finland

women

Figure 15.3 Temporal consequences of shift from Conventional Dual-earner household with children to Gendered Divorce with children

end are many and varied. Here we focus upon what was once the most traditional, and given failure to collect child support may in many places remain all too common: the ‘Gendered Divorce’ model described in section 13.5.4, where the divorced mother bears all responsibility (and the divorced father none) for financial support and physical care of the children. For purposes of this comparison, we will assume both parents were in paid labour (i.e., in a ‘Conventional Dual-earner household with children’) before their divorce; and per our standard methodology we assume both are in paid labour after divorce. Figure 15.3 shows the temporal consequences of that sort of divorce, for the average man and the average woman, in each of our six countries under study. Divorce is bad for both parents in all sorts of ways, ‘discretionary time’ among them. Figure 15.3 shows that both ex-partners are worse off in those terms after a divorce, when children are involved, even if the divorce is conducted on the Gendered Divorce rule. But, on that rule, men are only a little worse off whereas women are very much worse off. On average across all countries, fathers are around 3 hours a week worse off in consequence of a Gendered Divorce, whereas mothers are over 26 hours a week worse off. The exact magnitudes differ from one

The difference that states make

245

country to another. (Divorcing women lose almost twice as much discretionary time in the US as in France, and in Australia divorcing men lose hardly at all.) Still, the basic pattern is replicated across all countries. Gendered Divorce is seriously bad news for mothers, and very much worse news for mothers than fathers. In the country where the difference between them is smallest (France), Gendered Divorce rules still leave divorced mothers with fully 18 hours a week less discretionary time than they had when partnered (compared to 4 hours less for their ex-partners).

15.3 Temporal consequences of changing household rules The previous transitions all involve changes in ‘household types’: single to married (and back to single parent, in the end); children absent to present; two earners to one. Next we want to explore the effects of changing the ‘rules’ on which the household is run, while the basic composition of the household remains unchanged. For ease of comparison, all of the following discussions are based on the case where there are children present. Some changes to ‘household rules’ would make a big difference to the discretionary time available to each partner (or ex-partner). Were parents to shift from ‘Conventional Dual-earner’ rules to ‘Financially Egalitarian’ ones, women would on average gain considerably (and men lose correspondingly). The same is true if we were to shift from the ‘Gendered Divorce’ rule to either of the egalitarian divorce rules (most especially the ‘Strictly Egalitarian Divorce’ one). Interesting though those difference are in their own right, here we are particularly interested to see what difference the country makes to the magnitude of those differences. As we shall see below, those are generally not very large.

15.3.1 Conventional Dual-earner to Male Breadwinner Let us first ponder what would happen if the mother in a ‘Conventional Dual-earner’ household gives up her paid job to become a full-time homemaker. That is, consider the temporal consequences, across our six countries, of the transition from a ‘Conventional Dual-earner household with children’ to a ‘Male-Breadwinner household with children’.

246

Household regimes matter 40 30

6.49

5.39

2.25

2.55

6.71

3.47

10

–4.34

–8.70 –2.57

–7.19

–10.90

–20

–11.21

–10

–8.96

0 –8.55

hours per week

20

–30 –40 all countries

US

Australia men

Germany France

Sweden Finland

women

Figure 15.4 Temporal consequences of shift from Conventional Dual-earner household with children to Male-Breadwinner household with children

Figure 15.4 shows the temporal consequences of that transition, for the average man and the average woman, in each of our six countries under study. Again, there is a consistent pattern in the results reported in Figure 15.4. When Dual-earner parents become Male-Breadwinner parents, the men are always losers and women (almost) always gainers in terms of discretionary time. (The exception to the rule is Sweden, where both men and women are made worse off by that shift.) The magnitudes differ enough to be genuinely interesting, however. On average across all six countries, men lose over 8.5 hours of discretionary time from this shift, while women gain on average around 3.5 hours. But in some countries (the US, France and Finland) women gain something approaching double as much, and in other countries (Australia and Germany) men lose well over a quarter more.

15.3.2 Male Breadwinner to Most-efficient Breadwinner Our next scenario envisages a pair of parents shifting from one of the ‘Breadwinner’ rules to the other. At present, the household is run on ‘Male-Breadwinner’ rules: the father is the breadwinner, and the

The difference that states make

247

40

8.44

9.22

7.53

10.00

11.11

9.56

10

–7.03

–6.01

–3.54

–7.28

–4.22

–10

–7.45

0 –5.92

hours per week

20

11.07

30

–20 –30 –40 all countries

US

Australia men

Germany France

Sweden Finland

women

Figure 15.5 Temporal consequences of shift from Male Breadwinner with children to Most-efficient Breadwinner with children

mother the homemaker. What would happen if they were to operate according to ‘Most-efficient Breadwinner’ rules instead? Nothing (or nothing much) would change in those households where the father commands a higher hourly wage than the mother, which (as we saw in Figure 13.2) represents over two-thirds of cases. In the other third of households, where the mother’s wage rate is higher than the father’s, shifting from the Male-Breadwinner to the Most-efficient Breadwinner rule would make her the household’s new breadwinner, and the father would become the homemaker. Mothers invariably lose and fathers gain discretionary time in that process. That is partly (but only very partially) due to how the rules are stipulated.5 Mostly, it is due to hard facts about the world.6

5

6

The Male-Breadwinner rule assigns a ‘conventional’ portion of unpaid household responsibilities to men, whereas the Most-efficient Breadwinner rule assigns no responsibility for unpaid household labour to whoever is designated as ‘breadwinner’ on those grounds. i.e., there are some fathers whose wage rates are lower than their partners’. The Male-Breadwinner rule insists that they work whatever hours it takes to earn a poverty-level income for the household, nonetheless; whereas the Most-efficient Breadwinner rule shifts that responsibility onto their wives. In so doing, the Mostefficient Breadwinner rule exempts working mothers from unpaid household

248

Household regimes matter

Figure 15.5 shows the temporal consequences of a shift from the ‘MaleBreadwinner’ to the ‘Most-efficient Breadwinner’ rule, for the average man and the average woman, in each of our six countries under study. In shifting from Male-Breadwinner rules to Most-efficient Breadwinner ones, fathers gain on average just under 10 hours a week in discretionary time and mothers lose on average just under 6. Fathers gain a little more in the US and Germany; mothers lose more in the US, Germany and Finland, and less in Australia and France. But the basic pattern really is pretty similar across all the countries under study.

15.3.3 Conventional Dual-earner to Equal Temporal Contribution Consider next a pair of parents, both in paid labour and who presently run their household according to ‘Conventional Dual-earner’ rules. That is to say, each is responsible for whatever proportion of the household’s income needs as is conventional among dual-earner couples in that country; and each is responsible for whatever proportion of the household’s unpaid household labour as is conventional among dual-earner couples in that country. Suppose they are contemplating a shift from ‘Conventional Dualearner’ rules to some more egalitarian arrangement. Let us assume they decide that, if they are going to be egalitarian, they might as well go all the way and make Equal Temporal Contributions. What difference would that make to the discretionary time of each? Figure 15.6 shows the temporal consequences of that transition, for the average man and the average woman, in each of our six countries under study. As we see from Figure 15.6, for people with children the shift from Conventional Dual-earner rules to Equal Temporal Contribution rules makes little difference. Fathers are a couple of hours a week worse off, mothers a bit over an hour a week better off. And that is broadly true across all the countries under study.7

7

labour duties, to be sure. But it typically imposes even heavier paid-labour obligations on them. In the US women as well as men are made worse off by this shift. That happens because the average wife’s wage rate is so much lower than her husband’s that the extra time she has to engage in paid labour to make an equal contribution is

The difference that states make

249

40 30

3.15

0.70

2.17

0.51

1.90

1.20

10

–2.95

–2.51

–2.33

–1.57

–2.49

–10

–2.56 –1.26

0 –2.40

hours per week

20

–20 –30 –40

all countries

US

Australia men

Germany France

Sweden Finland

women

Figure 15.6 Temporal consequences of shift from Conventional Dual-earner with children to Equal Temporal Contribution with children

15.3.4 Gendered Divorce to Financially Egalitarian Divorce Suppose now the parents decide to divorce. The choice among alternative ‘Divorce rules’ matters greatly. To a large extent, as we have said in section 13.5, the choice of rules governing divorce falls to the government rather than the couple. But ex-partners can always choose to be more generous than the law strictly requires. Suppose the laws of the country would permit Gendered Divorce.8 Such a divorce would leave the mother with all responsibility for taking care of the children, both physically and financially. Suppose the couple is contemplating organizing their own divorce instead on Financially Egalitarian Divorce rules: the mother would provide all the physical care; but both ex-partners share the financial burden on a temporally egalitarian basis. What would be the impact on each partner’s discretionary time of divorcing on Financially Egalitarian Divorce terms, rather than on Gendered Divorce ones?

8

higher than the time she is excused from unpaid household labour as her husband takes on an equal share of that. Or, if they presently require ‘Financially Egalitarian Divorce’, suppose some law reformers are proposing a change back to ‘Gendered Divorce’. What would be the temporal consequence of that for divorcing partners?

250

Household regimes matter

40

4.88

4.63

3.06

8.01

6.64

10

–2.93

–2.86

–1.84

–4.73

–2.68

–10

–3.98

0 –3.17

hours per week

20

7.68

11.59

30

–20 –30 –40 all countries

US

Australia men

Germany France

Sweden Finland

women

Figure 15.7 Temporal consequences of shift from Gendered Divorce with children to Financially Egalitarian Divorce with children

Figure 15.7 shows the temporal consequences of that decision, for the average man and the average woman, in each of our six countries under study. As we see from Figure 15.7, had the Divorce rule been Gendered Divorce – and the alternative being switched to even just Financially Egalitarian Divorce – things would have been very different. In that shift, fathers would be 3 hours a week worse off and mothers 6.5 hours a week better off, on average. And in some countries (especially the US and to a lesser extent Germany) both would be much more so. This is one point at which country differences seem to matter in determining the impact of changes in household rules.

15.3.5 Financially Egalitarian Divorce to Strictly Egalitarian Divorce For our final scenario, imagine again the case of two parents who have decided to divorce. Suppose now that the laws of the country dictate that, at a bare minimum, divorces have to be Financially Egalitarian.9

9

Equivalently, we could think of it in terms of the impact of a move back to ‘Financially Egalitarian Divorce’ rules, if the country in question happened

The difference that states make

251

40 30

5.48

4.75

4.97

4.75

6.32

4.72

5.17

10

–5.48

–4.75

–4.97

–4.76

–6.32

–10

–5.07

0 –5.22

hours per week

20

–20 –30 –40 all countries

US

Australia men

Germany France

Sweden Finland

women

Figure 15.8 Temporal consequences of shift from Financially Egalitarian Divorce with children to Strictly Egalitarian Divorce with children

But divorcing parents are legally permitted to be more generous to one another if they so prefer; and the couple in question is contemplating organizing their own as a Strictly Egalitarian Divorce. That is to say, instead of the mother providing all the physical care of the children, that (as well as financial responsibility) would be split strictly equally between both parents. What would be the impact on each partner’s discretionary time of divorcing on Strictly Egalitarian Divorce terms, rather than on merely Financially Egalitarian Divorce ones? Figure 15.8 shows the temporal consequences of that decision, for the average man and the average woman, in each of our six countries under study. In the shift from a partially (merely Financially) Egalitarian Divorce rule to a fully (Strictly) Egalitarian one, fathers are again the losers and mothers the gainers. The one gains pretty much what the other loses, around 5 hours a week in discretionary time. And that is remarkably constant across all the countries under study.

already to have (as none here do) ‘Strictly Egalitarian Divorce’ rules enshrined in existing law.

252

Household regimes matter

15.4 The major difference states make From everything we have said so far, it would seem that it does not matter much what state you live under, in these connections. Shifting from one sort of household rule to another increases or decreases your discretionary time by pretty similar amounts everywhere under study. In the diagrams we have been looking at, one country or another does occasionally stand out from the rest. But there is no pattern to it. Countries that stand out with respect to one transition are strikingly similar to all the rest with respect to the other transitions. Countries that are very different in their welfare and gender regimes are, very much more often than not, enormously similar in all the figures we have been looking at in this chapter. For purposes of our methodology, that is a useful discovery. The broad similarity of countries in this respect reassures us that we can – as chapter 14 has done – average across all countries without fear that those averages mask any major differences among countries. We truly can talk about ‘what difference running your household on one sort of rule rather than another makes’, without specifying where. The main way your country matters is not in any differential temporal implications of any given rule but, rather, in dictating some ‘Divorce’ rule as the default option. As we have seen, it matters hugely – far more than anything else we have discussed in this book – which rule governs parents’ divorcing. A Gendered Divorce leaves mothers on average with almost 12 hours a week less discretionary time than a Strictly Egalitarian Divorce, and 6.5 hours a week less than a Financially Egalitarian Divorce. To some extent, the choice among ‘Divorce’ rules is a householdlevel choice.10 Even if family law in his country would permit the father to walk out on his family, providing his ex-wife and children with no further support, financial or otherwise, a departing father can always choose to be more generous to his ex-wife and children than that. Even if family law in her country formally requires her ex-husband to pay half (or, indeed, all) the financial costs of their children, a woman can always choose not to pursue her ex-husband through the courts if he defaults on that formal legal obligation. Still, it matters greatly which 10

‘Household-level’, in the sense that it is a choice made by one or the other of the divorcing partners – often unilaterally.

The difference that states make

253

‘Divorce’ rule the state chooses to set as the default option. That constitutes the baseline against which intra-household bargaining occurs among divorcing parents. That specifies an entitlement which courts will enforce, if either of the divorced parents wants. Different states have set different default ‘Divorce’ rules, from time to time. Traditionally, of course, divorce was not legally permitted at all in any of the countries under study; and even once it was, the MaleBreadwinner Divorce rule was typically the default. In the spirit of gender equality, there has typically been movement towards something more like a Financially Egalitarian Divorce rule. Doing so, ironically, costs divorced mothers 3 hours a week of discretionary time.

16

Alternative household rules and temporal autonomy

The lessons of the previous chapters are as follows:

 People can increase or reduce the amount of discretionary time available to them by joining together as a couple, by having children or by divorcing. Let us call that the ‘household form’ component of our conclusions.  People can increase or reduce the amount of discretionary time available to them by operating on different rules. Let us call that the ‘household rules’ component of our conclusions.  The amount of discretionary time people can gain or lose by altering the form or rules of their household can be affected by prevailing public policies and social norms in their country. Let us call that the ‘state and society’ component of our conclusions.  Part of the latter is the way in which discretionary time of divorcing parents can be affected by the rules their state imposes on divorce. Let us call that the specifically ‘divorce rules’ component of our conclusions.

16.1 Private choice matters The first two sets of conclusions speak to the issue of what individuals and couples can do to affect the amount of discretionary time they have. Basically, there are two things they can do. First, they can choose to live in one form of household or another: whether alone or with a partner; whether with children or without them. Second, they can choose the rules on which their household operates: whether partners will share household responsibilities equally, or whether one (and if so, which) will be breadwinner while the other stays at home, or something in between. The first sort of choice is, generally, of much more consequence. Partnering gains each partner around 12 hours a week, becoming 254

Alternative household rules and temporal autonomy

255

parents costs each over 9, divorcing (on Gendered Divorce terms) costs a mother well over 20 hours a week. The other sort of choice, over which rules will govern households that do not change in form, is much less consequential. Running the household on Egalitarian terms, in the Equal Temporal Contribution sense, leaves women with only a little over 1 hour more discretionary time than they enjoyed under the Conventional Dual-earner rule (and men with a couple of hours less). The most dramatic effect of changing ‘household rules’ comes in shifting from a Conventional Dual-earner to a Male-Breadwinner household: that leaves women with under 4 hours more in discretionary time a week, and men with more than twice that less discretionary time. Of course, describing either of these things – what sort of household to form, and on what terms to run it – as ‘choices’ begs certain questions. Fundamentally, it assumes that people had a choice of doing otherwise. And that has to be equally true of both parties involved. Maybe there are not many forced marriages in the countries under study. But it is harder to divorce, not to mention to abort, in some countries (and especially some communities within them) than others. Within households, too, the demands of one partner might powerfully constrain the choices of the other. One partner can severely constrain the other’s ‘choice’ of whether or not to have children by threatening to leave if they do (or do not). More generally, household rules have to be jointly chosen. A wife might wish her husband were the breadwinner and she could stay home with the children. But her wishing it does not make it so; he has to be willing to serve in that capacity. One partner might wish they lived in a more egalitarian household, sharing household tasks more equally. But again, wishing does not make it so unless the other partner is willing to live that way as well. It seems moderately unrealistic to expect much gain in discretionary time from changing household rules. That is in part because householdsharing behaviour has proven powerfully resistant to change. Even where there is dramatically increased lip service paid to gender equality in the domestic division of labour, actual behavioural change has been only relatively modest.1

1

Bittman 1999a. Speculating on the mechanism, Bittman points to a small-scale qualitative project in Sydney suggesting ‘misapprehension and discursive

256

Household regimes matter

More importantly, it is unrealistic to expect much gain in discretionary time from changing household rules because, while household form makes a big difference, running the same household according to one set of rules rather than another turns out to make a moderately small difference to people’s discretionary time.

16.2 Public environment matters 16.2.1 State and society matter As we have seen in chapter 15, it generally does not matter much what country you live in, so far as the temporal consequences of altering household forms or rules are concerned. Partnering is more beneficial in liberal countries than elsewhere, by up to 4 hours a week in the case of women (less for men). Becoming a parent is more costly there for both parents, by around 3 hours a week. Divorcing on Gendered terms is much more costly for women (by around 10 hours a week) in the US, Germany and Sweden than it is in France, Australia or Finland. But in general, that sort of disparity among countries is rare. And only occasionally (as in the case of Gendered Divorce) is it particularly large. It turns out that ‘state and society’ matter only occasionally, and usually only a little, in these connections.

16.2.2 Divorce rules matter What makes the biggest difference to people’s discretionary time is divorce. That is on one side a matter of household choice: the decision to divorce. But it is on the other side a matter of state choice: the terms on which divorce is permitted. Of course, as we have said, the state does not literally dictate the terms of the divorce. For the most part, all it does is to set the ‘default’ divorce rule. Divorcing parties can always choose to do something else. But that default rule remains important. It constitutes the baseline against which divorcing parties negotiate their own arrangements; it redefinition of equality . . . [M]en tend to inflate the size of their own contributions and to diminish the significance of their partner’s contributions’, making them feel as if the domestic division of labour is more equal than it actually is.

Alternative household rules and temporal autonomy

257

constitutes the fallback which dissatisfied partners can demand that the divorce court enforce. If the state were to impose a Gendered Divorce rule, then divorcing mothers in Conventional Dual-earner couples would suffer a loss of over 26 hours a week in discretionary time (more in some countries, less in others, as we have just said). If the rule shifted from Gendered Divorce to Financially Egalitarian – the mother still having full responsibility for the care of the children, but the father now sharing equally the financial costs of the children – that would reduce the mother’s loss by over 6 hours a week on average (and increase the cost to fathers by half that). If divorce were on Strictly Egalitarian terms – both parents sharing equally (in temporal terms) in the caring and financial responsibilities for their children – the average mother would still have a little less discretionary time than the average father because of her lower wage rate. But under the Strictly Egalitarian Divorce rule, divorcing parents would bear the temporal costs of divorce much more nearly equally. Under that rule, the average divorced mother would have over 14 hours a week less discretionary time than the average mother in a Conventional Dual-earner household, while the average divorced father would have over 11 hours a week less. In short, the biggest thing anyone can do to affect their discretionary time is to have kids, then divorce. And the biggest thing any state can do to affect people’s discretionary time is to dictate the terms on which such divorces will be permitted.

PART VI

Conclusions

17

Conclusions

17.1 Major findings In concluding, let us begin by summarizing the main findings of the book in bullet-point fashion:

 People have more temporal autonomy than we might imagine.











‘Spare time’, as conventionally measured (time not actually spent in paid labour, unpaid household labour or personal care) constitutes less than half of people’s true discretionary time (time they did not strictly need to spend in those activities) (see Figure 5.1). People’s temporal autonomy depends mostly on whether or not they have children, and whether or not they have a partner to help with the children. Lone parents have vastly less discretionary time than dual-earner couples without children – more than 40 hours a week less in the most extreme case, the US (see Figure 3.2). Looking just at ‘spare time’, in which respect the two groups are about the same, would mask that fact (see Figure 5.5). In general, women have less – but generally only a little less – temporal autonomy than men. However, stay-at-home wives, even with children, have much more – over 9 hours more – discretionary time than breadwinning fathers, on average (see Figure 14.2). Again, looking just at ‘spare time’, in which respect the two groups are about the same, would mask that fact (see Figure 5.7). Temporal autonomy varies widely from one country to another. In Sweden, where discretionary time is highest, people have over 9 more hours a week than in France, where it is lowest (see Figure 3.1). Temporal autonomy varies with welfare and gender regimes. On average, people have 5 hours more discretionary time in socialdemocratic/female-friendly welfare/gender regimes than in either liberal/individualist or corporatist/traditionalist regimes (see Figure 8.1). The state’s transfer payments and child-care subsidies increase people’s temporal autonomy, with the taxes it takes to pay for those 261

262









Conclusions

reducing it. The net effect is to reduce slightly the temporal autonomy of the average person everywhere except in Sweden (see Figure 8.1). But the net impact is strongly positive (up to 6 hours a week) on specific household types of regime-specific concern: lone mothers in liberal/individualist regimes; mothers in general in socialdemocratic/female-friendly regimes; and stay-at-home mothers in corporatist/traditionalist regimes (Figures 11.3, 11.5 and 11.6). In general, those state policies make only a modestly positive contribution to closing the gender gap in temporal autonomy (see Figure 11.2). But again, those policies impact more substantially on household types of regime-specific concern, as just described, with implications for gender differences. Much the greatest influences on people’s temporal autonomy are their life-cycle choices: to partner; to have children; and to divorce (especially if they have children). Forming a joint household gains each partner around a dozen hours a week extra discretionary time; having children costs each almost as much; divorce can cost a parent twice that, depending upon how the post-divorce costs and care of the children are shared (see Figures 15.1, 15.2 and 15.3). Life-cycle choices have similar effects across all our six countries, except when it comes to divorce. Even if divorce were on starkly gendered terms, with the woman bearing all responsibility financially and otherwise for the children, a divorcing mother would lose over 14 hours more discretionary time a week in the US than in France (see Figure 15.3). And of course if different governments enact different laws imposing different terms of divorce, crossnational differences can be even greater. The choice – whether made by households themselves or imposed by government – of the terms of divorce matter greatly to the relative temporal autonomy of divorced parents. On average across the countries under study, a gendered divorce of the sort just described costs mothers over 26 hours a week in discretionary time, and fathers 3 (see Figure 15.3). Equal sharing of the financial costs of the children, post-divorce, would see the mother recouping a quarter of that loss (and the father doubling his much smaller loss) (see Figure 15.7). Equal sharing of the care of the children as well would drop the divorced mother’s loss in discretionary time to under 15 hours a week, and raise the divorced father’s to over 11 (see Figure 15.8).

Conclusions

263

 There is relatively little that an intact (i.e., non-divorcing) household itself can do to alter the temporal autonomy of its members taken as a whole. Average household discretionary time varies little across the various ways intact households might run their affairs (see Figure 14.1). Even the distribution of discretionary time between partners varies surprisingly little. Running the household in more egalitarian ways than are presently conventional makes little difference to either partner’s discretionary time (see Figure 15.6). Only by reverting to the ‘male-breadwinner’ model would the distribution alter much, to the wife’s advantage and husband’s disadvantage (see Figure 15.4).

17.2 So what? One way of parodying our results might be as follows: To maximize temporal autonomy and discretionary time, people should: * * *

marry but never have children; if they do have children, never divorce; and maybe consider moving to Sweden.

No one should want to maximize discretionary time as an end in itself, however. The value of discretionary time is purely as a resource. And remember, the whole point of resources is to use them in ways you find satisfying. We know that well when it comes to the most familiar resource, money. The reason having lots of money enhances your welfare is that that gives you lots of opportunities to buy things that please you. If you forever hoard your money instead of spending it, those opportunities go unrealized. With the resource of money, clearly, there is no point in having it if you are never going to use it.1 Likewise with discretionary time: it is there to be ‘spent’ in ways that please you. For most people that includes having children; sometimes it includes divorce. For most people that includes doing enough paid labour to earn a comfortable (not just a poverty-level) income. For 1

Saving money might give you a sense of security and contribute to your welfare in that way – but not if you are resolved never to spend it, even come the ‘rainy day’ for which you are saving it.

264

Conclusions

most people that includes spending much more time in taking care of their house, their children and their own bodies than minimally necessary. Those are all ways of using the resource of discretionary time, ways of exercising the discretion that temporal autonomy gives you. There is nothing wrong with using your discretionary time in any of those ways, or indeed in any other way that pleases you. That is what it is there for. The only thing that would be wrong, in our view, would be to complain about the consequences of your choices, afterwards. People who have lots of resources and spend them on things that please them are better off than people who have few resources. They are better off before they spend them: they have more ‘opportunities’. They are better off after they have spent them: they have higher ‘desire satisfaction’. Thus, people who had lots of discretionary resources, and who spent lots of them, should be seen as privileged; they are well off. It would be ludicrous for them to complain that they are as badly off as someone who never had more than the amount of discretionary resources they now have left, after their own spending spree. That is obviously true of money, and it is equally true of the resource of discretionary time. That fundamental principle is commonly contravened, however, in assessing just how ‘time poor’ people are. If you just count how much time childless dual-earners have left, after all the hours they actually spend at the office, it would look like they are about as badly off as a lone mother. But they freely choose to spend most of the time they did at the office, whereas she had to spend very much more of the time she did there just to meet her household’s economic necessities. Or again, if you just counted how much time stay-at-home spouses spend in housework compared to how much time their employed husbands spend at the office, it would look as if they were shouldering roughly equal burdens. But far more of her housework than his paid work is discretionary, something she has chosen to do above and beyond what is strictly necessary. People have choices over how to spend their discretionary time. We should respect those choices. But we should recognize that they are indeed choices – choices that are themselves welfare-enhancing. People are better off, not badly off, the more discretionary time they have and are able to commit to projects of their choosing. Social concern ought instead to focus on the genuinely ‘time poor’, people who genuinely have little discretionary control over how they spend their time. Lone mothers, not double-income-no-kid households,

Conclusions

265

are the ones who should concern us most, in that regard. Judging countries by how they treat those most vulnerable members, Sweden is by far the best and the US by far the worst of the countries we have studied.2 Lone parents in Sweden have over 18 hours a week more discretionary time than they do in the US (see Figure 3.2). That’s like lone parents in the US having to do an extra job half the week, compared to their Swedish counterparts. When people freely choose to use some of their discretionary time to earn an unnecessarily high income, we say that we should respect their choice but not feel sorry for them: they could have chosen otherwise. Ought we, by parity of reasoning, say the same when people make life-cycle choices that have the effect of reducing their discretionary time? Parents have less discretionary time than childless couples, it is true. But (usually) they chose rather than were forced to become parents. Likewise, divorced parents have less discretionary time than married parents; but again (usually) they chose rather than were forced to divorce. Why treat those as events outside the control of people, when they are much more typically the product of their own choices, for which they ought to be held morally responsible?3 Why not treat the reduction in available discretionary time that people experience upon becoming parents (or divorced parents) on a par with the reduction in available discretionary time that people experience when working overtime for pay? All could arguably be seen as consumption acts, as ways of people’s exercising their discretionary control over the discretionary time that is available to them. That is precisely the economistic way of thinking of these things. Choices are choices. Children ought to be thought of as akin to ‘pets’ or ‘consumer durables’.4 Divorce is an expensive life-style option. People should be ‘free to choose’, and they should bear the natural consequences of their choices.5 Of course, not all of those outcomes are freely chosen. Not all children are planned. Not all divorces are by mutual consent. So in at least some cases, the temporal penalties experienced by parents are not

2 3 4 5

Goodin 1985. As a certain sort of ‘luck egalitarian’ might insist (Arneson 1997). As Nancy Folbre (2001, ch. 5; 1982, p. 325) delightfully glosses this view. Friedman and Friedman 1980.

266

Conclusions

something for which they have themselves only, or even principally, to blame. The larger point, however, is that that individualistic, economistic approach elides the crucial social facts of the matter. For society to reproduce itself, it is essential for children to be born and cared for. When producing and rearing children, people are performing a social service, not (certainly not just) a private consumption act.6 In suggesting that society ought to respond sympathetically to the time pressures under which parents (particularly lone parents) labour, we are doing so out of recognition of that social service that is being performed. Of course, when you work long hours in paid employment, you are doing socially useful things as well. But you are also getting rewarded for them through your pay packet. In a perfectly functioning market economy (which nowhere exists, of course) you would be getting fully rewarded for them. Raising children is also rewarding, of course: children do bear some resemblance to pets, in that respect. But you are nowhere nearly as fully rewarded for the long hours spent raising children by the intrinsic rewards of that activity as you are for the long hours spent in paid labour by your pay packet. Children are good for society and everyone in it. In producing and rearing them, you are producing public goods that benefit everyone else as well as yourself.7 So it is only right that we should be more sympathetic to the time plight of overworked parents than overworked childless corporate lawyers. All of that is simply to say that society has an interest in, and derivative from that a responsibility for, its children. Derivative from that, it has a responsibility to care for its children’s carers.8 Sometimes people come to occupy that role by accident or forces largely outside their control; sometimes they come to occupy it by choice. But even when it is a role they have fully chosen, that choice serves to activate – not to cancel – society’s responsibility to care for them as its children’s carers.

6 7

8

Folbre 2008. As is clearly acknowledged by writers as diverse as Friedman (1962, ch. 6) and Etzioni (1993, ch. 2). Policy implications are explored in more detail in Garfinkel et al. 1996 and Esping-Andersen 2002, pp. 26–67. Goodin 1998; 2005; England and Folbre 1999; Folbre 2008.

Conclusions

267

17.3 Implications concerning public policy 17.3.1 Work-time flexibility In this book, we have calibrated temporal autonomy in terms of discretionary time. Those calculations tell us how much temporal autonomy people could have, under various alternative welfare-gender-household regimes. But those calculations specifically presuppose that people could spend any amount of time in paid labour they choose. (Or, more precisely, they presuppose that people could if they wanted spend only as long in paid labour as they strictly need to do to avoid poverty, without suffering any serious penalty in the hourly rate at which their labour is remunerated.) We have offered evidence suggesting that people could do that, oneat-a-time. We infer that from the fact that there are jobs currently being performed in all the economies under study that pay pretty much any wage rate for pretty much any number of hours per week. There are not many people occupying some of those cells, to be sure. In particular there are not many people with high wage rates working minimal numbers of hours; but that may just be because not many people want to live in near-poverty circumstances if they can easily avoid it. The point is merely that there are at least some people in pretty much every cell in that large matrix, suggesting that most sorts of jobs could be arranged in such a way as to allow people to work short (maybe very short) hours, if they wanted to do so. Whether or not people in general are allowed a genuinely free choice in this matter is something that is hotly contested. The evidence seems mixed. Here we have offered no further evidence one way or another on that issue, beyond this minimal reality check: ‘at least a few people do’. What we would emphasize is just this. If your concern is (as ours in this book is) with maximizing people’s temporal autonomy, then one of the most important things policy-makers must do is to ensure worktime flexibility. They must ensure that people have, insofar as possible, a free choice of the number of hours a week they work in paid labour. And they must ensure, insofar as possible, that people do not pay an inordinate penalty in terms of their hourly remuneration for choosing to work different numbers of hours. The evidence to which we have already referred suggests that most jobs can indeed be done, productively, for just about any number of

268

Conclusions

hours a week. Maybe not all jobs are like that. Maybe it is not possible for literally everyone to work in any given job for just a few hours a week; maybe one person’s ability to do that depends upon most people not doing that. So maybe the legislation here in view will have to be hedged in all sorts of ways. Still, if our concern is – as our concern in this book pre-eminently is – with maximizing people’s temporal autonomy, the first thing we ought to recommend is legislation protecting (insofar as is feasible) flexible working hours, at the discretion of the worker involved.9 We should also add – emphatically, oblique though the point may be to our main analysis in this book – that it matters hugely not just how many hours people have of free time but also which hours they have free. It is far more useful to have Monday morning or Friday afternoon free than it is to have free three hours in the middle of Tuesday. It is far more useful to have the same time free as your partner has free.10 The timing and ‘harmonization’ of free time is as important as the amount of such time. Those factors lie outside the present study. But they are nonetheless enormously important. They too ought to be of central concern to policy-makers attempting to promote temporal autonomy.

17.3.2 Equitable divorce rules As we have said, much the most important changes in people’s temporal autonomy come with life-cycle changes, partnering, having children and divorcing. The terms upon which people divorce are partly up to them, but states set the baselines against which those private orderings are internally negotiated. If state policy permits ‘gendered divorce’ as one possible outcome of intra-household negotiations – in which the woman is assigned complete financial and caring responsibilities for the children – that leaves the divorcing mother massively worse off than the divorcing father. If state policy prescribes ‘strictly egalitarian divorce’, instead – requiring both former partners to share equitably in both financial and caring 9

10

A state pursuing such policies systematically might be characterized as ‘postproductivist’; the Netherlands in the 1990s might have been a forerunner of such a state (Goodin 2001; van der Veen and Groot 2006). Boulin and Mu¨ckenberger 1999.

Conclusions

269

responsibilities for the offspring of the former partnership – divorced mothers and fathers have much more nearly equal temporal burdens. Any state that cares about the temporal autonomy of its citizens ought therefore also to legislate equitable divorce rules, as a matter of highest priority.

17.3.3 Culture of equality There are more diffuse ways in which a ‘culture of social equality’ enhances the temporal autonomy of people in a country. That is picked up in the ‘pre-government’ baseline calculations we have reported in chapter 8 above. Partly it is a matter of wage policy – high averages, low variance. Partly it is a matter of moderating social expectations: non-fussy standards of housekeeping and personal grooming; not expecting one another to spend all that much time preening before the mirror or sweeping communal stairs on a daily basis. The salient contrast in this respect, in our data, is between Sweden and France. The former has significantly lower median (and hence, the way we define it, ‘necessary’) time in personal care. One interpretation might of course be simply that Swedes are slobs. Another might be that French standards are simply too obtrusive. We are inclined to frame this discussion in terms of the advantages, in terms of temporal autonomy, that come from a ‘culture of equality’. Sweden has it; France does not. Lots of other countries do not, and would benefit from adopting it.

17.3.4 Public transfers and subsidies Finally, and most familiarly, governments can increase people’s temporal autonomy in more modest ways by providing transfer payments and subsidized services. Those policies increase discretionary time by reducing the amount of time that people would have to work in order to pay for things they need. Governments have to pay for those, however, by levying taxes which decrease discretionary time. The net effects largely cancel. So it is unsurprising that taxes, transfers and child-care subsidies taken together have negligible effects overall on discretionary time, averaging across the whole population.

270

Conclusions

For specifically targeted subgroups of the population, however, transfer payments and subsidized services can make decidedly nonnegligible contributions to discretionary time. Although modest compared to the effects of life-cycle changes, and governments’ potential role in them, these effects are nonetheless well worth achieving. They typically amount to giving people in the targeted subgroups 4 to 6 hours extra discretionary time a week. Any union leaders who brought us a four-and-a-half-day workweek would hardly be castigated for having achieved only such a modest success. Instead, they would be great heroes. So too ought we to regard governments that increase similarly the temporal autonomy of socially important subgroups of their populations.

Appendix 1: Methodology

A1.1 Introductory notes Table A1.1 describes the years and national currencies associated with particular countries. In the other tables in this appendix, all income variables are reported in the national currencies listed in this table. All time-use variables are reported in hours per week. For the purposes of the procedures described in this appendix, adults are taken to be people aged 18 years or over. Children are taken to be people aged under 18 years.

A1.2 Actual households A1.2.1 Actual time in personal care (1) Calculate a new diary day weight (DW), based on how many days of the week a diary day is designed to represent. For example, if each person completed a diary on one random day of the week, then for each diary day, DW is equal to 7. If each person completed diaries on two random days of the week, then for each diary day, DW is equal to 3.5. If each person completed a diary on one random weekday and a diary on one random weekend day, then for each weekday diary day, DW is equal to 5, while for each weekend diary day, DW is equal to 2. (2) For each diary day, calculate PECXX by summing ‘time for dressing/toilet’ (AV13), ‘time spent receiving personal services’ (AV14), ‘time spent eating meals and snacks’ (AV15) and ‘time spent sleeping/napping’ (AV16). (3) For each diary day, calculate PECX by multiplying PECXX by 1440 divided by the difference between 1440 and ‘time in unclassifiable, unrecorded, or unknown activities’ (AV41). (4) Split the data set into groups of diary days associated with each person. 271

272

Appendix 1

Table A1.1. Years and national currencies

US Australia Germany France Sweden Finland

Years to which data relate

National currency

2000 (LIS), 2003 (MTUS) 1989 (LIS), 1992 (MTUS) 1991/2 (MTUS), 1994 (LIS) 1994 (LIS), 1998/9 (MTUS) 1990/1 (MTUS), 1992 (LIS) 1999 (Statistics Finland)

US dollar Australian dollar German deutschmark French franc Swedish krona Finnish markka

(5) Weight the data set by DW. (6) For each of the groups of diary days associated with each person, calculate the mean of PECX among the diary days (MPECX). (7) For each person, calculate his or her actual time in personal care by multiplying MPECX by 7 divided by 60. For each person, the particular value of MPECX used in this calculation should be based on which of the groups of diary days delineated in Step 4 is associated with that person.

A1.2.2 Actual time in unpaid household labour (1) For each diary day, calculate UHLXX by summing ‘time cooking, washing up’ (AV6), ‘time spent doing housework’ (AV7), ‘time spent doing odd jobs’ (AV8), ‘time spent gardening’ (AV9), ‘time spent shopping’ (AV10), ‘time spent in child care’ (AV11) and ‘time spent during domestic travel’ (AV12). (2) For each diary day, calculate UHLX by multiplying UHLXX by 1440 divided by the difference between 1440 and ‘time in unclassifiable, unrecorded or unknown activities’ (AV41). (3) Split the data set into groups of diary days associated with each person. (4) Weight the data set by DW. (5) For each of the groups of diary days associated with each person, calculate the mean of UHLX among the diary days (MUHLX). (6) For each person, calculate his or her actual time in unpaid household labour by multiplying MUHLX by 7 divided by 60. For each person, the particular value of MUHLX used in this calculation

Methodology

273

should be based on which of the groups of diary days delineated in Step 3 is associated with that person.

A1.2.3 Actual time in paid labour, excluding travel to/from work (1) For each diary day, calculate PLXXX by summing ‘time in paid work’ (AV1), ‘time in paid work at home’ (AV2) and ‘time in paid work, second job’ (AV3). (2) For each diary day, calculate PLXX by multiplying PLXXX by 1440 divided by the difference between 1440 and ‘time in unclassifiable, unrecorded or unknown activities’ (AV41). (3) Split the data set into groups of diary days associated with each person. (4) Weight the data set by DW. (5) For each of the groups of diary days associated with each person, calculate the mean of PLXX among the diary days (MPLXX). (6) For each person, calculate his or her actual time in paid labour, excluding travel to/from work, by multiplying MPLXX by 7 divided by 60. For each person, the particular value of MPLXX used in this calculation should be based on which of the groups of diary days delineated in Step 3 is associated with that person.

A1.2.4 Actual time in travel to/from work (1) For each diary day, TRWXX is equal to ‘time in travel to/from work’ (AV5). (2) For each diary day, calculate TRWX by multiplying TRWXX by 1440 divided by the difference between 1440 and ‘time in unclassifiable, unrecorded or unknown activities’ (AV41). (3) Split the data set into groups of diary days associated with each person. (4) Weight the data set by DW. (5) For each of the groups of diary days associated with each person, calculate the mean of TRWX among the diary days (MTRWX). (6) For each person, calculate his or her actual time in travel to/from work by multiplying MTRWX by 7 divided by 60. For each person, the particular value of MTRWX used in this calculation should be based on which of the groups of diary days delineated in Step 3 is associated with that person.

274

Appendix 1

A1.2.5 Actual time in paid labour (1) For each person, calculate his or her actual time in paid labour by summing his or her actual time in paid labour, excluding travel to/from work, and his or her actual time in travel to/ from work.

A1.2.6 Spare time (1) For each diary day, calculate SPTXX by summing ‘time in school, classes’ (AV4), ‘time spent during travel for leisure’ (AV17), ‘time spent on excursions, trips’ (AV18), ‘time spent playing/actively participating in sports’ (AV19), ‘time spent watching/passively participating in sports’ (AV20), ‘time spent walking’ (AV21), ‘time in religious activities’ (AV22), ‘time doing civic duties/civic organizations’ (AV23), ‘time at the cinema or theatre’ (AV24), ‘time at dances or parties, etc.’ (AV25), ‘time at social clubs’ (AV26), ‘time at pubs’ (AV27), ‘time at restaurants’ (AV28), ‘time visiting friends’ (AV29), ‘time listening to radio’ (AV30), ‘time watching the television or video’ (AV31), ‘time listening to records, tapes, cds’ (AV32), ‘time in study’ (AV33), ‘time reading books’ (AV34), ‘time reading papers, magazines’ (AV35), ‘time relaxing’ (AV36), ‘time in conversation’ (AV37), ‘time entertaining friends’ (AV38), ‘time knitting, sewing, etc.’ (AV39) and ‘time in other hobbies or pastimes’ (AV40). (2) For each diary day, calculate SPTX by multiplying SPTXX by 1440 divided by the difference between 1440 and ‘time in unclassifiable, unrecorded or unknown activities’ (AV41). (3) Split the data set into groups of diary days associated with each person. (4) Weight the data set by DW. (5) For each of the groups of diary days associated with each person, calculate the mean of SPTX among the diary days (MSPTX). (6) For each person, calculate his or her spare time by multiplying MSPTX by 7 divided by 60. For each person, the particular value of MSPTX used in this calculation should be based on which of the groups of diary days delineated in Step 3 is associated with that person.

Methodology

275

Equivalently, for each person: spare time ¼ 168  actual time in personal care  actual time in unpaid household labour  actual time in paid labour:

A1.2.7 Necessary time in personal care For the following algorithm, the sample was restricted to men and women who were aged between 20 and 65 years. (1) Weight the data set by a person weight. (2) Calculate the median of actual time in personal care among these people (MPEC). (3) For each man and for each woman, calculate his or her necessary time in person care by multiplying MPEC by 0.8. Estimates of necessary time in personal care in particular countries are given in Table A1.2.

A1.2.8 Necessary time in unpaid household labour (1) For each person, if he or she is aged between 20 and 65 years, his or her actual time in ‘cross-nationally comparable’ unpaid household labour (CUHL) is equal to his or her actual time in unpaid household labour.

Table A1.2. Necessary time in personal care (hours per week) Necessary time in personal care US Australia Germany France Sweden Finland

55.16 59.97 56.47 61.60 53.66 55.22

276

Appendix 1

(2) For each person, if he or she is aged less than 20 years or more than 65 years, his or her actual time in ‘cross-nationally comparable’ unpaid household labour (CUHL) is equal to 0. (3) For each household, calculate the total actual time spent in ‘crossnationally comparable’ unpaid household labour by all the people in the household combined (HCUHL). In order to do this, simply add together the actual time in ‘cross-nationally comparable’ unpaid household labour (CUHL) of all the people in the household. (4) For each household, calculate the household’s ‘equivalent’ actual time in ‘cross-nationally comparable’ unpaid household labour (EHCUHL), using the square root of the number of people in the household as the equivalence scale. In order to do this, divide the total actual time spent in ‘cross-nationally comparable’ unpaid household labour by all the people in the household combined (HCUHL) by the square root of the number of people in the household. (5) Calculate a new household weight (HW) by multiplying a household weight by the number of people in the household. (6) Split the data set into four groups of households based on a crossclassification of the following two dimensions: (a) the age of the youngest child in the household (child between 0 and 4 years in the household; no child between 0 and 4 years in the household); and (b) whether or not there is an adult in the household who is both not employed and not self-employed (i.e., whether or not there is an adult in the household who does not work) (yes; no). (7) Weight the data set by HW. (8) For each of the four groups of households delineated in Step 6, calculate the median of EHCUHL among the households that belong to each of these groups (MEHCUHL). (9) For each household, calculate the household’s necessary time in unpaid household labour (NHUHL) by dividing MEHCUHL by 2, and then multiplying the result by the square root of the number of people in the household. For each household, the particular value of MEHCUHL used in this calculation should be based on which of the four groups of households delineated in Step 6 the household belongs to. Table A1.3 reports – for particular countries and particular household types – estimates of half of median household ‘equivalent’ actual time in ‘cross-nationally comparable’ unpaid household labour.

Methodology

277

Table A1.3. Half of median household ‘equivalent’ actual time in ‘cross-nationally comparable’ unpaid household labour (equivalent hours per week) Child between 0 and 4

US Australia Germany France Sweden Finland

No child between 0 and 4

Adult who does not work

No adult who does not work

Adult who does not work

No adult who does not work

18.69 21.43 21.07 16.10 21.65 20.57

14.60 17.68 16.92 14.29 13.93 17.13

18.14 18.93 20.15 19.05 21.24 16.05

12.84 13.56 13.13 11.79 13.42 12.43

For the following algorithm, the sample was restricted to households that included either: (1) a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and at least one of whom was an earner; or (2) a man or a woman who was single and aged between 25 and 54 years, who did not live with any other men or women, and who was an earner. The households could either include children or not include children. The sample was then further restricted to the men and the women in these households. (10) For each household, calculate the total actual time spent in unpaid household labour by the household head and the household head’s spouse (if a spouse is present) (HSUHL). In order to do this, simply add together the actual time spent in unpaid household labour by the head and the actual time spent in unpaid household labour by the spouse (if present). (11) For each head and for each spouse, calculate his or her proportional responsibility for unpaid household labour (PUHL) by dividing his or her actual time in unpaid household labour by the total actual time spent in unpaid household labour by the head and the spouse (if present) (HSUHL). This will equal the proportion that the head or the spouse contributes to the total actual time spent in unpaid household labour by the head and the spouse (if present) (HSUHL).

278

Appendix 1

(12) For each head and for each spouse, calculate his or her necessary time in unpaid household labour by multiplying the household’s necessary time in unpaid household labour (NHUHL) by PUHL.

A1.2.9 Necessary time in paid labour, excluding travel to/from work A1.2.9.1 From the household’s disposable income gross of child-care costs to the household’s disposable income net of child-care costs For a particular household in the sample, that household’s disposable income net of child-care costs can be estimated by the following equation: d ¼ g  ðC02 n02 þ C35 n35 Þh where, d ¼ disposable income net of child-care costs g ¼ disposable income gross of child-care costs C02 ¼ hourly cost of child care per child aged 2 years or less C35 ¼ hourly cost of child care per child aged between 3 and 5 years n02 ¼ number of children aged 2 years or less in the household n35 ¼ number of children aged between 3 and 5 years in the household h ¼ hours of child care needed We assume that, generally speaking, the hours of child care needed by a particular household is equal to the minimum time in paid labour, excluding travel to/from work, across all the adults in that household. (This assumes that child care takes place very near the workplace of the adult in the household who spends the least time in paid labour, excluding travel to/from work.) Estimates of the hourly costs of child care per child in particular countries are given in Table A1.4. A1.2.9.2 From the household’s disposable income net of child-care costs to the household’s necessary income net of child-care costs (1) For each household, calculate the household’s equivalent income net of child-care costs (EHI), using the square root of the number of people in the household as the equivalence scale. In order to do this, divide the household’s disposable income net of child-care costs by the square root of the number of people in the household.

Methodology

279

Table A1.4. Hourly costs of child care per child (national currency per hour per child) Pre-government

US Australia Germany France Sweden Finland

Post-government

Child 2 or less

Child between 3 and 5

Child 2 or less

Child between 3 and 5

4.16 2.01 4.97 10.30 12.32 12.12

4.16 2.01 4.97 10.30 12.32 12.12

3.75 1.91 3.77 6.10 0.00 0.00

0.59 0.87 0.00 0.00 0.00 0.00

Note: For a discussion of pre-government and post-government discretionary time, see section 2.2.5 in the main text of this book. For a discussion of the costs of child care, see section 2.2.2.

Table A1.5. The poverty line (equivalent national currency per year) Poverty line US Australia Germany France Sweden Finland

(2) (3) (4) (5)

12 027.94 8 325.82 14 540.74 44 620.80 62 867.75 48 304.30

Weight the data set by HW. Calculate the median of EHI among these households (MEHI). Calculate the poverty line by dividing MEHI by 2. For each household, calculate the household’s necessary income net of child-care costs (NHI) by multiplying the poverty line by the square root of the number of people in the household.

Table A1.5 presents estimates of the poverty line in particular countries.

280

Appendix 1

A1.2.9.3 From the household’s necessary income net of child-care costs to the household’s necessary labour income net of child-care costs A1.2.9.3.1 Estimating the household’s expected alimony and child support received For the following algorithm, the sample was restricted to households that included either: (1) a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and at least one of whom was an earner; or (2) a man or a woman who was single and aged between 25 and 54 years, who did not live with any other men or women, and who was an earner. The households could either include children or not include children. The sample was then further restricted to households that received alimony or child support and whose equivalent income net of childcare costs (EHI) was between 50 and 150 per cent of the poverty line. (These further sample restrictions were subsequently lifted.) (1) Weight the data set by a household weight. (2) Calculate the mean of alimony and child support received among these households (MHACS). The further sample restrictions (described above) were lifted in relation to the following algorithm. That is to say, for the following algorithm, the sample was restricted to households that included either: (1) a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and at least one of whom was an earner; or (2) a man or a woman who was single and aged between 25 and 54 years, who did not live with any other men or women, and who was an earner. The households could either include children or not include children. (3) For each household, if the household receives alimony or child support, then the household’s expected alimony and child support received (EHACS) is equal to MHACS. (4) For each household, if the household does not receive alimony or child support, then the household’s expected alimony and child support received (EHACS) is equal to 0.

Methodology

281

Table A1.6. Expected alimony and child support received for households that receive alimony or child support (national currency per year) Expected alimony and child support received US Australia Germany France Sweden Finland

3 633.22 2 224.55 — 20 058.23 15 275.41 10 564.81

Note: The 1994 German LIS data set did not include information specifically on alimony and child support received.

Table A1.6 lists, for particular countries, estimates of expected alimony and child support received for households that receive alimony or child support. A1.2.9.3.2 Estimating the household’s expected ‘taxes-and-transfers’ For the following algorithm, the sample was restricted to households that included either: (1) a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and at least one of whom was an earner; or (2) a man or a woman who was single and aged between 25 and 54 years, who did not live with any other men or women, and who was an earner. The households could either include children or not include children. The sample was then further restricted to households whose equivalent income net of child-care costs (EHI) was between 50 and 150 per cent of the poverty line. (This further sample restriction was subsequently lifted.) (1) Split the data set into the following four groups of households: (a) households without children; (b) two-adult, two-earner households with children; (c) two-adult, one-earner households with children; and (d) one-adult, one-earner households with children. (2) Weight the data set by a household weight.

282

Appendix 1

(3) For each of the four groups of households delineated in Step 1, calculate the mean of ‘taxes-and-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) among the households that belong to each of these groups (MHTT). The further sample restriction (described above) was lifted in relation to the following algorithm. That is to say, for the following algorithm, the sample was restricted to households that included either: (1) a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and at least one of whom was an earner; or (2) a man or a woman who was single and aged between 25 and 54 years, who did not live with any other men or women, and who was an earner. The households could either include children or not include children. (4) For each household, the household’s expected ‘taxes-and-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) (EHTT) is equal to MHTT. For each household, the particular value of MHTT used in this equation should be based on which of the four groups of households delineated in Step 1 the household would belong to. Table A1.7 reports, for particular countries and particular household types, estimates of households’ expected ‘taxes-and-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations). A1.2.9.3.3 Estimating the household’s necessary labour income net of child-care costs For the following algorithm, the sample was restricted to households that included either: (1) a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and at least one of whom was an earner; or (2) a man or a woman who was single and aged between 25 and 54 years, who did not live with any other men or women, and who was an earner. The households could either include children or not include children.

Methodology

283

Table A1.7. Households’ expected ‘taxes-and-transfers’ (national currency per year) Pre-government Post-government US Households without children Two-adult, two-earner households with children Two-adult, one-earner households with children One-adult, one-earner households with children Australia Households without children Two-adult, two-earner households with children Two-adult, one-earner households with children One-adult, one-earner households with children Germany Households without children Two-adult, two-earner households with children Two-adult, one-earner households with children One-adult, one-earner households with children France Households without children Two-adult, two-earner households with children Two-adult, one-earner households with children One-adult, one-earner households with children Sweden Households without children Two-adult, two-earner households with children

53.52

2 673.27

8.51

5 137.48

81.64

1 136.27

73.54

171.34

0.00

320.80

0.00

1 088.54

10.40

1 143.47

4.65

2 607.32

828.57

5 235.82

0.00

7 702.19

89.80

5 240.85

612.15

5 975.81

382.82

10 529.70

0.00

10 080.88

197.13

6 211.66

422.52

8 270.37

0.00

7 929.84

0.00

8 179.12

284

Appendix 1

Table A1.7. (cont.) Pre-government Post-government Two-adult, one-earner households with children One-adult, one-earner households with children Finland Households without children Two-adult, two-earner households with children Two-adult, one-earner households with children One-adult, one-earner households with children

0.00

22 386.77

0.00

12 511.81

410.43

14 676.44

254.79

9 175.64

609.91

13 365.53

298.36

3 197.41

Note: For a discussion of pre-government and post-government discretionary time, see section 2.2.5 in the main text of this book.

(1) For each household, the household’s necessary labour income net of child-care costs (NHLI) = (the household’s necessary income net of child-care costs (NHI)  the household’s expected alimony and child support received (EHACS)  the household’s expected ‘taxes-and-transfers’ (EHTT)  the household’s cash property income  the household’s occupational pensions)  52.

A1.2.9.4 From the household’s necessary labour income net of child-care costs to men and women’s necessary time in paid labour, excluding travel to/from work For the following algorithm, the sample was restricted to households that included either: (1) a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and at least one of whom was an earner; or (2) a man or a woman who was single and aged between 25 and 54 years, who did not live with any other men or women, and who was an earner. The households could

Methodology

285

either include children or not include children. The sample was then further restricted to the men and the women in these households. (1) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then his or her necessary time in paid labour excluding travel to/ from work (NPLX) is equal to 0. (2) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his or her necessary labour income net of child-care costs (NLI) by executing the following algorithm. a) For each household, calculate the total actual labour income earned by the household head and the household head’s spouse (if a spouse is present) (HSLI). In order to do this, simply add together the actual labour income earned by the head and the actual labour income earned by the spouse (if present). b) For each head and for each spouse, calculate his or her proportional responsibility for labour income (PLI) by dividing his or her actual labour income by the total actual labour income earned by the head and the spouse (if present) (HSLI). This will equal the proportion that the head or the spouse contributes to the total actual labour income earned by the head and the spouse (if present) (HSLI). For the one earning adult in one-adult, one-earner and two-adult, oneearner households, PLI should equal 1. For the one non-earning adult in two-adult, one-earner households, PLI should equal 0. For the two earning adults in two-adult, two-earner households, PLI should be greater than 0 but less than 1. c) For each head and for each spouse, calculate his or her necessary labour income net of child-care costs (NLI) by multiplying the household’s necessary labour income net of child-care costs (NHLI) by PLI. As just mentioned, in cases in which the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, men and women’s necessary time in paid labour excluding travel to/from work (NPLX) is also equal to 0. The following algorithms are used to go from men and women’s necessary labour income net of child-care costs (NLI) to men and women’s necessary time in paid labour excluding travel to/from work (NPLX) in those cases in which the household’s necessary labour income net of child-care costs (NHLI) is greater than 0.

286

Appendix 1

For the following algorithms, the sample was restricted to households that included either: (1) a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and at least one of whom was an earner; or (2) a man or a woman who was single and aged between 25 and 54 years, who did not live with any other men or women, and who was an earner. The households could either include children or not include children. The sample was then further restricted to the men and the women in these households. A1.2.9.4.1 Preliminary equations For each man and for each woman, his or her labour income net of child-care costs can be estimated by the following equation: l ¼ wt  ðC02 n02 þ C35 n35 Þhr where, l ¼ the man or the woman’s labour income net of child-care costs w ¼ hourly wage rate t ¼ time in paid labour, excluding travel to/from work r ¼ proportional responsibility for the cost of child care A particular man or woman’s proportional responsibility for the cost of child care can vary between 0 (if the man or woman has no responsibility for funding the cost of child care) and 1 (if the man or woman has total responsibility for funding the cost of child care). For each man and for each woman, the following equation describes the relationship between his or her necessary labour income net of child-care costs (NLI) and his or her necessary time in paid labour, excluding travel to/from work (NPLX): i ¼ wx  ðC02 n02 þ C35 n35 Þhr where, i ¼ the man or the woman’s necessary labour income net of child-care costs (NLI) x ¼ necessary time in paid labour, excluding travel to/from work (NPLX) Rearranging this equation yields: x ¼ ði þ ðC02 n02 þ C35 n35 ÞhrÞ  w

(Equation A1:1)

Methodology

287

If we specify that, for a particular person in the sample, his or her hours of child care needed will equal his or her necessary time in paid labour, excluding travel to/from work, then the relationship between necessary labour income net of child-care costs and necessary time in paid labour, excluding travel to/from work, becomes: i ¼ wx  ðC02 n02 þ C35 n35 Þxr Rearranging this equation yields: x ¼ i  ðw  ðC02 n02 þ C35 n35 ÞrÞ

(Equation A1:2)

For the adult in one-adult, one-earner households, his or her hours of child care needed will equal his or her necessary time in paid labour, excluding travel to/from work. His or her proportional responsibility for the cost of child care will also equal 1. Hence, the following equation describes the relationship between his or her necessary labour income net of child-care costs and his or her necessary time in paid labour, excluding travel to/from work: i ¼ wx  ðC02 n02 þ C35 n35 Þx Rearranging this equation yields: x ¼ i  ðw  ðC02 n02 þ C35 n35 ÞÞ

(Equation A1:3)

For the earning adult in two-adult, one-earner households, his or her hours of child care needed will equal 0. His or her proportional responsibility for the cost of child care will equal 1. Hence, the following equation describes the relationship between his or her necessary labour income net of child-care costs and his or her necessary time in paid labour, excluding travel to/from work: i ¼ wx Rearranging this equation yields: x¼iw

(Equation A1:4)

For the adult in two-adult, two-earner households with the least necessary time in paid labour, excluding travel to/from work, his or her hours of child care needed will equal his or her necessary time in paid labour, excluding travel to/from work. His or her proportional responsibility for the cost of child care will equal his or her proportional responsibility for labour income (PLI). Hence, the following

288

Appendix 1

equation describes the relationship between his or her necessary labour income net of child-care costs and his or her necessary time in paid labour, excluding travel to/from work: ia ¼ wa xa  ðC02 n02 þ C35 n35 Þxa pa where, p ¼ proportional responsibility for labour income (PLI) and the subscript a indicates that the variable refers to the one earning adult in two-adult, two-earner households with the least necessary time in paid labour, excluding travel to/from work. Rearranging this equation yields: xa ¼ ia  ðwa  ðC02 n02 þ C35 n35 Þpa Þ

(Equation A1:5)

For the adult in two-adult, two-earner households with the most necessary time in paid labour, excluding travel to/from work, his or her hours of child care needed will equal xa (as determined by Equation A1.5). His or her proportional responsibility for the cost of child care will equal his or her proportional responsibility for labour income (PLI). Hence, the following equation describes the relationship between his or her necessary labour income net of child-care costs and his or her necessary time in paid labour, excluding travel to/from work: ib ¼ wb xb  ðC02 n02 þ C35 n35 Þxa pb where the subscript b indicates that the variable refers to the one earning adult in two-adult, two-earner households with the most necessary time in paid labour, excluding travel to/from work. Rearranging this equation yields: xb ¼ ðib þ ðC02 n02 þ C35 n35 Þxa pb Þ  wb

(Equation A1:6)

For a two-adult, two-earner household, the total labour income net of child-care costs earned by the man and woman combined can be estimated by the following equation: k ¼ wm tm  ðC02 n02 þ C35 n35 Þhrm þ ww tw  ðC02 n02 þ C35 n35 Þhrw where, k ¼ the total labour income net of child-care costs earned by the man and woman combined

Methodology

289

and where the subscript m indicates that the variable refers to the man, and the subscript w indicates that the variable refers to the woman. Since in this case the sum of rm and rw is 1, this equation simplifies to: k ¼ wm tm þ ww tw  ðC02 n02 þ C35 n35 Þh If we specify that the man and the woman spend the same amount of time in paid labour, excluding travel to/from work – say, tmw ¼ tm ¼ tw – then the hours of child care needed by the man and woman combined will also equal tmw. Substituting tmw into this equation yields: k ¼ wm tmw þ ww tmw  ðC02 n02 þ C35 n35 Þtmw where the subscript mw indicates that the variable refers to both the man and the woman. Rearranging this equation yields: tmw ¼ k  ðwm þ ww  ðC02 n02 þ C35 n35 ÞÞ For a two-adult, two-earner household, if we specify that the man and the woman spend the same amount of time in paid labour, excluding travel to/from work, then the following equation describes the relationship between the total necessary labour income net of childcare costs for the man and woman combined – in other words, the household’s necessary labour income net of child-care costs (NHLI) – and the man and the woman’s necessary time in paid labour, excluding travel to/from work: xmw ¼ j  ðwm þ ww  ðC02 n02 þ C35 n35 ÞÞ

(Equation A1:7)

where, j ¼ the household’s necessary labour income net of child-care costs (NHLI) References will be made to Equations A1.1 through A1.7 in subsequent sections of this appendix. A1.2.9.4.2 An algorithm for one-adult, one-earner households (1) For each man and for each woman, calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) via Equation A1.3.

290

Appendix 1

A1.2.9.4.3 An algorithm for two-adult, one-earner households (1) For each non-earning man and for each non-earning woman, his or her necessary time in paid labour excluding travel to/from work (NPLX) is equal to 0. (2) For each earning man and for each earning woman, calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) via Equation A1.4. A1.2.9.4.4 An algorithm for two-adult, two-earner households under the Conventional Dual-earner rule (1) First, assume that the adult in the household who spends the least time in paid labour, excluding travel to/from work, is the man rather than the woman. This assumption will be referred to as ‘Assumption 1’. For each man, calculate his necessary time in paid labour, excluding travel to/from work under Assumption 1 (M1NPLX) via Equation A1.5. (2) For each woman, calculate her necessary time in paid labour, excluding travel to/from work under Assumption 1 (W1NPLX) via Equation A1.6. (3) Second, assume that the adult in the household who spends the least time in paid labour, excluding travel to/from work, is the woman rather than the man. This assumption will be referred to as ‘Assumption 2’. For each woman, calculate her necessary time in paid labour, excluding travel to/from work, under Assumption 2 (W2NPLX) via Equation A1.5. (4) For each man, calculate his necessary time in paid labour, excluding travel to/from work, under Assumption 2 (M2NPLX) via Equation A1.6. (5) If the man’s necessary time in paid labour, excluding travel to/from work, under Assumption 1 (M1NPLX) is less than the woman’s necessary time in paid labour, excluding travel to/from work, under Assumption 1 (W1NPLX) and the man’s necessary time in paid labour, excluding travel to/from work, under Assumption 2 (M2NPLX) is less than the woman’s necessary time in paid labour, excluding travel to/from work, under Assumption 2 (W2NPLX), then, for each man, his necessary time in paid labour excluding travel to/from work (NPLX) is equal to his necessary time in paid labour, excluding travel to/from work, under Assumption 1 (M1NPLX).

Methodology

291

(6) If the man’s necessary time in paid labour, excluding travel to/from work, under Assumption 1 (M1NPLX) is less than the woman’s necessary time in paid labour, excluding travel to/from work, under Assumption 1 (W1NPLX) and the man’s necessary time in paid labour, excluding travel to/from work, under Assumption 2 (M2NPLX) is less than the woman’s necessary time in paid labour, excluding travel to/from work, under Assumption 2 (W2NPLX), then, for each woman, her necessary time in paid labour excluding travel to/from work (NPLX) is equal to her necessary time in paid labour excluding travel to/from work under Assumption 1 (W1NPLX). (7) If the man’s necessary time in paid labour, excluding travel to/from work, under Assumption 1 (M1NPLX) is greater than the woman’s necessary time in paid labour, excluding travel to/from work, under Assumption 1 (W1NPLX) and the man’s necessary time in paid labour, excluding travel to/from work, under Assumption 2 (M2NPLX) is greater than the woman’s necessary time in paid labour, excluding travel to/from work, under Assumption 2 (W2NPLX), then, for each man, his necessary time in paid labour excluding travel to/from work (NPLX) is equal to his necessary time in paid labour, excluding travel to/from work, under Assumption 2 (M2NPLX). (8) If the man’s necessary time in paid labour, excluding travel to/from work, under Assumption 1 (M1NPLX) is greater than the woman’s necessary time in paid labour, excluding travel to/from work, under Assumption 1 (W1NPLX) and the man’s necessary time in paid labour, excluding travel to/from work, under Assumption 2 (M2NPLX) is greater than the woman’s necessary time in paid labour, excluding travel to/from work, under Assumption 2 (W2NPLX), then, for each woman, her necessary time in paid labour excluding travel to/from work (NPLX) is equal to her necessary time in paid labour, excluding travel to/from work, under Assumption 2 (W2NPLX).

A1.2.10 Necessary time in travel to/from work For the following algorithm, the sample was restricted to diary days during which some actual time was spent in paid labour, excluding travel to/from work. That is to say, for the following algorithm, the sample was restricted to diary days for which the sum of ‘time in paid

292

Appendix 1

work’ (AV1), ‘time in paid work at home’ (AV2) and ‘time in paid work, second job’ (AV3) – i.e., PLXXX – was greater than 0. (This sample restriction was subsequently relaxed.) (1) For each diary day, calculate TRWD by dividing TRWX by 60. (2) Weight the data set by a diary day weight. (3) Calculate the mean of TRWD among these diary days (MTRWD). The sample restriction (described above) was relaxed in relation to the following algorithm. Instead, however, for the following algorithm, the sample was restricted to households that included either: (1) a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and at least one of whom was an earner; or (2) a man or a woman who was single and aged between 25 and 54 years, who did not live with any other men or women, and who was an earner. The households could either include children or not include children. The sample was then further restricted to the men and the women in these households. (4) For each man and for each woman, his or her necessary number of days at work (NDW) is equal to the smallest integer that is greater than or equal to his or her necessary time in paid labour, excluding travel to/from work, divided by 8. (5) For each man and for each woman, calculate his or her necessary time in travel to/from work by multiplying his or her necessary number of days at work (NDW) by MTRWD. Table A1.8 reports estimates of mean actual time in travel to/from work during workdays (i.e., MTRWD) in particular countries. Table A1.8. Mean actual time in travel to/from work during workdays (hours per day) Actual time in travel to/from work during workdays US Australia Germany France Sweden Finland

0.62 0.79 0.70 0.81 0.66 0.72

Methodology

293

A1.2.11 Necessary time in paid labour For the following algorithm, the sample was restricted to households that included either: (1) a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and at least one of whom was an earner; or (2) a man or a woman who was single and aged between 25 and 54 years, who did not live with any other men or women, and who was an earner. The households could either include children or not include children. The sample was then further restricted to the men and the women in these households. (1) For each man and for each woman, calculate his or her necessary time in paid labour by summing his or her necessary time in paid labour, excluding travel to/from work, and his or her necessary time in travel to/from work.

A1.2.12 Discretionary time For the following algorithm, the sample was restricted to households that included either: (1) a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and at least one of whom was an earner; or (2) a man or a woman who was single and aged between 25 and 54 years, who did not live with any other men or women, and who was an earner. The households could either include children or not include children. The sample was then further restricted to the men and the women in these households. For each man and for each woman: discretionary time ¼ 168  necessary time in personal care  necessary time in unpaid household labour  necessary time in paid labour.

A1.3 Alternative households For the following alternative households, necessary time in personal care, necessary time in travel to/from work, necessary time in paid

294

Appendix 1

labour, and discretionary time for the household head and the household head’s spouse (if a spouse is present) are calculated using the same algorithms as are used for actual households (on which, see above). However, necessary time in unpaid household labour and necessary time in paid labour, excluding travel to/from work, for the head and the spouse (if present) are calculated using different algorithms (on which, see below). For the following alternative households, other variables are calculated using the same algorithms as are used for actual households, unless stated otherwise. For the following alternative household rules, the sample was restricted to households that included a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and who were both earners. The households could either include children or not include children. The sample was then further restricted to the men and the women in these households. (These sample restrictions were temporarily relaxed in the process of calculating some figures used in the Male-Breadwinner rule.)

A1.3.1 Male-Breadwinner rule A1.3.1.1 Necessary time in unpaid household labour For the following algorithm, the sample was restricted to households that included a man and a woman who were both married and aged between 25 and 54 years and who did not live with any other men or women – and in which the man was an earner and the woman was not an earner. The households could either include children or not include children. The sample was further restricted to the men and the women in these households. (1) For each household, calculate the total actual time the man and woman combined spend in unpaid household labour (MWUHL). In order to do this, simply add together the man’s actual time in unpaid household labour and the woman’s actual time in unpaid household labour. (2) For each man and for each woman, calculate his or her proportional responsibility for unpaid household labour (PUHL) by dividing his or her actual time in unpaid household labour by the total actual time the man and woman combined spend in unpaid household labour (MWUHL). This will equal the proportion that the

Methodology

(3) (4) (5) (6) (7)

295

man or woman contributes to the total actual time the man and woman combined spend in unpaid household labour (MWUHL). Weight the data set by a person weight. Calculate the mean of PUHL among men without children (MC0MPUHL). Calculate the mean of PUHL among men with children (MC1MPUHL). Calculate the mean of PUHL among women without children (WC0MPUHL). Calculate the mean of PUHL among women with children (WC1MPUHL).

The sample restrictions that were applied in relation to all the alternative household rules (these sample restrictions are described above) were applied in relation to the following algorithm. That is to say, for the following algorithm, the sample was restricted to households that included a man and a woman who were both married and aged between 25 and 54 years, who did not live with any other men or women, and who were both earners. The households could either include children or not include children. The sample was then further restricted to the men and the women in these households. (8) For each man without children, calculate his necessary time in unpaid household labour by multiplying MC0MPUHL by the household’s necessary time in unpaid household labour (NHUHL) for a household in which: (a) the number of people is equal to that in the man’s actual household; (b) there are no children; and (c) there is an adult who is both not employed and not self-employed (i.e., there is an adult who does not work). (9) For each man with children, calculate his necessary time in unpaid household labour by multiplying MC1MPUHL by the household’s necessary time in unpaid household labour (NHUHL) for a household in which: (a) the number of people is equal to that in the man’s actual household; (b) the age of the youngest child is the same as that in the man’s actual household; and (c) there is an adult who is both not employed and not self-employed (i.e., there is an adult who does not work). (10) For each woman without children, calculate her necessary time in unpaid household labour by multiplying WC0MPUHL by the household’s necessary time in unpaid household labour

296

Appendix 1

(NHUHL) for a household in which: (a) the number of people is equal to that in the woman’s actual household; (b) there are no children; and (c) there is an adult who is both not employed and not self-employed (i.e., there is an adult who does not work). (11) For each woman with children, calculate her necessary time in unpaid household labour by multiplying WC1MPUHL by the household’s necessary time in unpaid household labour (NHUHL) for a household in which: (a) the number of people is equal to that in the woman’s actual household; (b) the age of the youngest child is the same as that in the woman’s actual household; and (c) there is an adult who is both not employed and not selfemployed (i.e., there is an adult who does not work). A1.3.1.2 Necessary time in paid labour, excluding travel to/from work (1) For each household, if the household has children, then the household’s expected ‘taxes-and-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) (EHTT) is equal to the household’s expected ‘taxes-andtransfers’ (EHTT) for a two-adult, one-earner household with children. (2) For each household, if the household does not have children, then the household’s expected ‘taxes-and-transfers’ (EHTT) is equal to the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (3) For each household, the household’s necessary labour income net of child-care costs (NHLI) = (the household’s necessary income net of child-care costs (NHI)  the household’s expected alimony and child support received (EHACS)  the household’s expected ‘taxes-and-transfers’ (EHTT)  the household’s cash property income  the household’s occupational pensions)  52.

(4) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then his or her necessary time in paid labour excluding travel to/ from work (NPLX) is equal to 0.

Methodology

297

(5) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. (a) For each woman, her necessary time in paid labour excluding travel to/from work (NPLX) is equal to 0. (b) For each man, his necessary labour income net of child-care costs (NLI) is equal to the household’s necessary labour income net of child-care costs (NHLI). (c) For each man, calculate his necessary time in paid labour excluding travel to/from work (NPLX) via Equation A1.4.

A1.3.2 Most-efficient Breadwinner Rule A1.3.2.1 Necessary time in unpaid household labour (1) Determine which, of the man and the woman, is the partner with the higher wage rate and which is the partner with the lower wage rate. (2) For the partner with the higher wage rate, his or her necessary time in unpaid household labour is equal to 0. (3) For the partner with the lower wage rate, his or her necessary time in unpaid household labour is equal to the household’s necessary time in unpaid household labour (NHUHL) for a household in which: (a) the number of people is equal to that in the man or woman’s actual household; (b) the age of the youngest child is the same as that in the man or woman’s actual household; and (c) there is an adult who is both not employed and not self-employed (i.e., there is an adult who does not work). A1.3.2.2 Necessary time in paid labour, excluding travel to/from work (1) For each household, if the household has children, then the household’s expected ‘taxes-and-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable

298

Appendix 1

organizations) (EHTT) is equal to the household’s expected ‘taxesand-transfers’ (EHTT) for a two-adult, one-earner household with children. (2) For each household, if the household does not have children, then the household’s expected ‘taxes-and-transfers’ (EHTT) is equal to the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (3) For each household, the household’s necessary labour income net of child-care costs (NHLI) = (the household’s necessary income net of child-care costs (NHI)  the household’s expected alimony and child support received (EHACS)  the household’s expected ‘taxes-and-transfers’ (EHTT)  the household’s cash property income  the household’s occupational pensions)  52.

(4) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then his or her necessary time in paid labour excluding travel to/ from work (NPLX) is equal to 0. (5) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) Determine which, of the man and the woman, is the partner with the higher wage rate and which is the partner with the lower wage rate. b) For the partner with the lower wage rate, his or her necessary time in paid labour excluding travel to/from work (NPLX) is equal to 0. c) For the partner with the higher wage rate, his or her necessary labour income net of child-care costs (NLI) is equal to the household’s necessary labour income net of child-care costs (NHLI). d) For the partner with the higher wage rate, calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) via Equation A1.4.

Methodology

299

A1.3.3 Conventional Dual-earner rule A1.3.3.1 Necessary time in unpaid household labour (1) For each household, calculate the total actual time the man and woman combined spend in unpaid household labour (MWUHL). In order to do this, simply add together the man’s actual time in unpaid household labour and the woman’s actual time in unpaid household labour. (2) For each man and for each woman, calculate his or her proportional responsibility for unpaid household labour (PUHL) by dividing his or her actual time in unpaid household labour by the total actual time the man and woman combined spend in unpaid household labour (MWUHL). This will equal the proportion that the man or woman contributes to the total actual time the man and woman combined spend in unpaid household labour (MWUHL). (3) For each man and for each woman, calculate his or her necessary time in unpaid household labour by multiplying the household’s necessary time in unpaid household labour (NHUHL) by PUHL. A1.3.3.2 Necessary time in paid labour, excluding travel to/from work (1) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then his or her necessary time in paid labour excluding travel to/ from work (NPLX) is equal to 0. (2) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) For each household, calculate the total actual labour income earned by the man and woman combined (MWLI). In order to do this, simply add together the man’s actual labour income and the woman’s actual labour income. b) For each man and for each woman, calculate his or her proportional responsibility for labour income (PLI) by dividing his or her actual labour income by the total actual labour income earned by the man and woman combined (MWLI). This will equal the

300

Appendix 1

proportion that the man or woman contributes to the total actual labour income earned by the man and woman combined (MWLI). Given that the man or woman will be one of the two earning adults in a two-adult, two-earner household, for the man or woman, PLI should be greater than 0 but less than 1. c) For each man and for each woman, calculate his or her necessary labour income net of child-care costs (NLI) by multiplying the household’s necessary labour income net of child-care costs (NHLI) by PLI. d) For each man and for each woman, calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) by executing the algorithm for two-adult, two-earner households under the Conventional Dual-earner rule.

A1.3.4 Equal Monetary Contribution rule A1.3.4.1 Necessary time in unpaid household labour (1) For each man and for each woman, calculate his or her necessary time in unpaid household labour by dividing the household’s necessary time in unpaid household labour (NHUHL) by 2. A1.3.4.2 Necessary time in paid labour, excluding travel to/from work (1) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then his or her necessary time in paid labour excluding travel to/ from work (NPLX) is equal to 0. (2) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) For each man and for each woman, set his or her proportional responsibility for labour income (PLI) to 0.5. b) For each man and for each woman, calculate his or her necessary labour income net of child-care costs (NLI) by multiplying the household’s necessary labour income net of child-care costs (NHLI) by PLI.

Methodology

301

c) For each man and for each woman, calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) by executing the algorithm for two-adult, two-earner households under the Conventional Dual-earner rule.

A1.3.5 Equal Temporal Contribution rule A1.3.5.1 Necessary time in unpaid household labour As per the Equal Monetary Contribution rule. A1.3.5.2 Necessary time in paid labour, excluding travel to/from work (1) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then his or her necessary time in paid labour excluding travel to/ from work (NPLX) is equal to 0. (2) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) via Equation A1.7.

A1.3.6 Atomistic Divorce rule A1.3.6.1 Necessary time in unpaid household labour (1) For each man and for each woman, his or her necessary time in unpaid household labour is equal to the household’s necessary time in unpaid household labour (NHUHL) for a household in which there is: (a) one person; (b) no children; and (c) no adult who is both not employed and not self-employed (i.e., there is no adult who does not work). A1.3.6.2 Necessary time in paid labour, excluding travel to/from work (1) For each man and for each woman, the household’s necessary income net of child-care costs (NHI) is equal to the poverty line. (2) For each man and for each woman, if his or her actual household receives alimony or child support and does not have children, and if

302

(3)

(4)

(5)

(6)

Appendix 1

he or she is the partner with the lower wage rate, then the household’s expected alimony and child support received (EHACS) is equal to MHACS. For each man and for each woman, if his or her actual household does not receive alimony or child support or has children, or if he or she is the partner with the higher wage rate, then the household’s expected alimony and child support received (EHACS) is equal to 0. For each man and for each woman, the household’s expected ‘taxes-and-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) (EHTT) is equal to the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. For each man and for each woman, calculate his or her cash property income (PPROP) by dividing the household’s cash property income by 2. For each man and for each woman,

the household’s necessary labour income net of child-care costs (NHLI) = (the household’s necessary income net of child-care costs (NHI)  the household’s expected alimony and child support received (EHACS)  the household’s expected ‘taxes-and-transfers’ (EHTT)  the man or woman’s cash property income (PPROP)  the man or woman’s occupational pensions)  52.

(7) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then his or her necessary time in paid labour excluding travel to/ from work (NPLX) is equal to 0. (8) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) For each man and for each woman, his or her necessary labour income net of child-care costs (NLI) is equal to the household’s necessary labour income net of child-care costs (NHLI). b) For each man and for each woman, calculate his or her necessary time in paid labour excluding travel to/from work (NPLX)

Methodology

303

via Equation A1.3. In Equation A1.3, set the number of children aged 2 years or less in the household and the number of children aged between 3 and 5 years in the household to 0.

A1.3.7 Self-reliant Divorce rule A1.3.7.1 Necessary time in unpaid household labour (1) For each man and for each woman, his or her necessary time in unpaid household labour is equal to the household’s necessary time in unpaid household labour (NHUHL) for a household in which: (a) the number of people is equal to that in the man or woman’s actual household, minus 1; (b) the age of the youngest child is the same as that in the man or woman’s actual household; and (c) there is no adult who is both not employed and not self-employed (i.e., there is no adult who does not work). A1.3.7.2 Necessary time in paid labour, excluding travel to/from work (1) For each man and for each woman, calculate the household’s necessary income net of child-care costs (NHI) by multiplying the poverty line by the square root of the number of people in the household minus 1. (2) For each man and for each woman, if his or her actual household receives alimony or child support and has children, then the household’s expected alimony and child support received (EHACS) is equal to MHACS. (3) For each man and for each woman, if his or her actual household receives alimony or child support and does not have children, and if he or she is the partner with the lower wage rate, then the household’s expected alimony and child support received (EHACS) is equal to MHACS. (4) For each man and for each woman, if his or her actual household receives alimony or child support and does not have children, and if he or she is the partner with the higher wage rate, then the household’s expected alimony and child support received (EHACS) is equal to 0. (5) For each man and for each woman, if his or her actual household does not receive alimony or child support, then the household’s

304

(6)

(7)

(8)

(9)

Appendix 1

expected alimony and child support received (EHACS) is equal to 0. For each man and for each woman, if his or her actual household has children, then the household’s expected ‘taxes-and-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) (EHTT) is equal to the household’s expected ‘taxes-and-transfers’ (EHTT) for a one-adult, one-earner household with children. For each man and for each woman, if his or her actual household does not have children, then the household’s expected ‘taxes-andtransfers’ (EHTT) is equal to the household’s expected ‘taxesand-transfers’ (EHTT) for a household without children. For each man and for each woman, calculate his or her cash property income (PPROP) by dividing the household’s cash property income by 2. For each man and for each woman,

the household’s necessary labour income net of child-care costs (NHLI) = (the household’s necessary income net of child-care costs (NHI)  the household’s expected alimony and child support received (EHACS)  the household’s expected ‘taxes-and-transfers’ (EHTT)  the man or woman’s cash property income (PPROP)  the man or woman’s occupational pensions)  52:

(10) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then his or her necessary time in paid labour excluding travel to/from work (NPLX) is equal to 0. (11) For each man and for each woman, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) For each man and for each woman, his or her necessary labour income net of child-care costs (NLI) is equal to the household’s necessary labour income net of child-care costs (NHLI). b) For each man and for each woman, calculate his or her necessary time in paid labour excluding travel to/from work

Methodology

305

(NPLX) via Equation A1.3. In Equation A1.3, set the number of children aged 2 years or less in the household to the number in the man or woman’s actual household. Set the number of children aged between 3 and 5 years in the household to the number in the man or woman’s actual household.

A1.3.8 Gendered Divorce rule A1.3.8.1 Necessary time in unpaid household labour For each man, as per the Atomistic Divorce rule. For each woman, as per the Self-reliant Divorce rule. A1.3.8.2 Necessary time in paid labour, excluding travel to/from work For each man, as per the Atomistic Divorce rule. For each woman, as per the Self-reliant Divorce rule.

A1.3.9 Male-Breadwinner Divorce rule A1.3.9.1 Necessary time in unpaid household labour For each man, as per the Atomistic Divorce rule. For each woman, as per the Self-reliant Divorce rule. A1.3.9.2 Necessary time in paid labour, excluding travel to/from work (1) For each woman, the household’s necessary income net of childcare costs (NHI) is equal to the poverty line. (2) For each woman, if her actual household receives alimony or child support and does not have children, and if she is the partner with the lower wage rate, then the household’s expected alimony and child support received (EHACS) is equal to MHACS. (3) For each woman, if her actual household does not receive alimony or child support or has children, or if she is the partner with the higher wage rate, then the household’s expected alimony and child support received (EHACS) is equal to 0. (4) For each woman, the household’s expected ‘taxes-and-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) (EHTT) is equal to the

306

Appendix 1

household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (5) For each woman, calculate her cash property income (PPROP) by dividing the household’s cash property income by 2. (6) For each woman, the household’s necessary labour income net of child-care costs (NHLI) = (the household’s necessary income net of child-care costs (NHI)  the household’s expected alimony and child support received (EHACS)  the household’s expected ‘taxes-and-transfers’ (EHTT)  the woman’s cash property income (PPROP)  the woman’s occupational pensions)  52.

(7) For each woman, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then her necessary time in paid labour excluding travel to/from work (NPLX) is equal to 0. (8) For each woman, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate her necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) For each woman, her necessary labour income net of child-care costs (NLI) is equal to the household’s necessary labour income net of child-care costs (NHLI). b) For each woman, calculate her necessary time in paid labour excluding travel to/from work (NPLX) via Equation A1.1. In Equation A1.1, set the proportional responsibility for the cost of child care to 0. (9) For each man, calculate the household’s necessary income net of child-care costs (NHI) by multiplying the poverty line by the square root of the number of people in the household minus 1. (10) For each man, if his actual household receives alimony or child support and has children, then the household’s expected alimony and child support received (EHACS) is equal to MHACS. (11) For each man, if his actual household receives alimony or child support and does not have children, and if he is the partner with

Methodology

(12)

(13)

(14)

(15)

(16) (17)

307

the lower wage rate, then the household’s expected alimony and child support received (EHACS) is equal to MHACS. For each man, if his actual household receives alimony or child support and does not have children, and if he is the partner with the higher wage rate, then the household’s expected alimony and child support received (EHACS) is equal to 0. For each man, if his actual household does not receive alimony or child support, then the household’s expected alimony and child support received (EHACS) is equal to 0. For each man, if his actual household has children, then the household’s expected ‘taxes-and-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) (EHTT) is equal to the household’s expected ‘taxes-and-transfers’ (EHTT) for a one-adult, one-earner household with children. For each man, if his actual household does not have children, then the household’s expected ‘taxes-and-transfers’ (EHTT) is equal to the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. For each man, calculate his cash property income (PPROP) by dividing the household’s cash property income by 2. For each man,

the household’s necessary labour income net of child-care costs (NHLI) = (the household’s necessary income net of child-care costs (NHI)  the household’s expected alimony and child support received (EHACS)  the household’s expected ‘taxes-and-transfers’ (EHTT)  the man’s cash property income (PPROP)  the man’s occupational pensions)  52:

(18) For each man, if the household’s necessary labour income net of childcare costs (NHLI) is less than or equal to 0, then his necessary time in paid labour excluding travel to/from work (NPLX) is equal to 0. (19) For each man, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) For each man, his necessary labour income net of child-care costs (NLI) is equal to the household’s necessary labour income net of child-care costs (NHLI).

308

Appendix 1

b) For each man, calculate his necessary time in paid labour excluding travel to/from work (NPLX) via Equation A1.1. In Equation A1.1, set the number of children aged 2 years or less in the household to the number in the man’s actual household. Set the number of children aged between 3 and 5 years in the household to the number in the man’s actual household. Set the hours of child care needed to the woman’s necessary time in paid labour, excluding travel to/from work (NPLX). Set the proportional responsibility for the cost of child care to 1.

A1.3.10 Most-efficient Breadwinner Divorce rule A1.3.10.1 Necessary time in unpaid household labour For the partner with the higher wage rate, as per the Atomistic Divorce rule. For the partner with the lower wage rate, as per the Self-reliant Divorce rule. A1.3.10.2 Necessary time in paid labour, excluding travel to/from work (1) For the partner with the lower wage rate, the household’s necessary income net of child-care costs (NHI) is equal to the poverty line. (2) For the partner with the lower wage rate, if his or her actual household receives alimony or child support and does not have children, then the household’s expected alimony and child support received (EHACS) is equal to MHACS. (3) For the partner with the lower wage rate, if his or her actual household does not receive alimony or child support or has children, then the household’s expected alimony and child support received (EHACS) is equal to 0. (4) For the partner with the lower wage rate, the household’s expected ‘taxes-and-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) (EHTT) is equal to the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (5) For the partner with the lower wage rate, calculate his or her cash property income (PPROP) by dividing the household’s cash property income by 2.

Methodology

309

(6) For the partner with the lower wage rate, the household’s necessary labour income net of child-care costs (NHLI) = (the household’s necessary income net of child-care costs (NHI)  the household’s expected alimony and child support received (EHACS)  the household’s expected ‘taxes-and-transfers’ (EHTT)  the partner with the lower wage rate’s cash property income (PPROP)  the partner with the lower wage rate’s occupational pensions)  52:

(7) For the partner with the lower wage rate, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then his or her necessary time in paid labour excluding travel to/from work (NPLX) is equal to 0. (8) For the partner with the lower wage rate, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) For the partner with the lower wage rate, his or her necessary labour income net of child-care costs (NLI) is equal to the household’s necessary labour income net of child-care costs (NHLI). b) For the partner with the lower wage rate, calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) via Equation A1.1. In Equation A1.1, set the proportional responsibility for the cost of child care to 0. (9) For the partner with the higher wage rate, calculate the household’s necessary income net of child-care costs (NHI) by multiplying the poverty line by the square root of the number of people in the household minus 1. (10) For the partner with the higher wage rate, if his or her actual household receives alimony or child support and has children, then the household’s expected alimony and child support received (EHACS) is equal to MHACS. (11) For the partner with the higher wage rate, if his or her actual household receives alimony or child support and does not have children, then the household’s expected alimony and child support received (EHACS) is equal to 0.

310

Appendix 1

(12) For the partner with the higher wage rate, if his or her actual household does not receive alimony or child support, then the household’s expected alimony and child support received (EHACS) is equal to 0. (13) For the partner with the higher wage rate, if his or her actual household has children, then the household’s expected ‘taxesand-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) (EHTT) is equal to the household’s expected ‘taxes-and-transfers’ (EHTT) for a one-adult, one-earner household with children. (14) For the partner with the higher wage rate, if his or her actual household does not have children, then the household’s expected ‘taxes-and-transfers’ (EHTT) is equal to the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (15) For the partner with the higher wage rate, calculate his or her cash property income (PPROP) by dividing the household’s cash property income by 2. (16) For the partner with the higher wage rate, the household’s necessary labour income net of child-care costs (NHLI) = (the household’s necessary income net of child-care costs (NHI)  the household’s expected alimony and child support received (EHACS)  the household’s expected ‘taxes-and-transfers’ (EHTT)  the partner with the higher wage rate’s cash property income (PPROP)  the partner with the higher wage rate’s occupational pensions)  52:

(17) For the partner with the higher wage rate, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then his or her necessary time in paid labour excluding travel to/from work (NPLX) is equal to 0. (18) For the partner with the higher wage rate, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) For the partner with the higher wage rate, his or her necessary labour income net of child-care costs (NLI) is equal to the

Methodology

311

household’s necessary labour income net of child-care costs (NHLI). b) For the partner with the higher wage rate, calculate his or her necessary time in paid labour excluding travel to/from work (NPLX) via Equation A1.1. In Equation A1.1, set the number of children aged 2 years or less in the household to the number in the partner with the higher wage rate’s actual household. Set the number of children aged between 3 and 5 years in the household to the number in the partner with the higher wage rate’s actual household. Set the hours of child care needed to the partner with the lower wage rate’s necessary time in paid labour, excluding travel to/from work (NPLX). Set the proportional responsibility for the cost of child care to 1.

A1.3.11 Financially Egalitarian Divorce rule (Money-egalitarian) A1.3.11.1 Necessary time in unpaid household labour For each man, as per the Atomistic Divorce rule. For each woman, as per the Self-reliant Divorce rule. A1.3.11.2 Necessary time in paid labour, excluding travel to/from work (1) For each man and for each woman, the household’s necessary income net of child-care costs (NHI) is equal to half of the sum of: (a) the poverty line multiplied by the square root of the number of people in the household minus 1; and (b) the poverty line. (2) For each man and for each woman, if his or her actual household receives alimony or child support and has children, then the household’s expected alimony and child support received (EHACS) is equal to MHACS divided by 2. (3) For each man and for each woman, if his or her actual household receives alimony or child support and does not have children, and if he or she is the partner with the lower wage rate, then the household’s expected alimony and child support received (EHACS) is equal to MHACS.

312

Appendix 1

(4) For each man and for each woman, if his or her actual household receives alimony or child support and does not have children, and if he or she is the partner with the higher wage rate, then the household’s expected alimony and child support received (EHACS) is equal to 0. (5) For each man and for each woman, if his or her actual household does not receive alimony or child support, then the household’s expected alimony and child support received (EHACS) is equal to 0. (6) For each man and for each woman, if his or her actual household has children, then the household’s expected ‘taxes-and-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) (EHTT) is equal to half of the sum of: (a) the household’s expected ‘taxes-and-transfers’ (EHTT) for a one-adult, one-earner household with children; and (b) the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (7) For each man and for each woman, if his or her actual household does not have children, then the household’s expected ‘taxes-andtransfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) (EHTT) is equal to the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (8) For each man and for each woman, calculate his or her cash property income (PPROP) by dividing the household’s cash property income by 2. (9) For each man and for each woman, the household’s necessary labour income net of child-care costs (NHLI) = (the household’s necessary income net of child-care costs (NHI)  the household’s expected alimony and child support received (EHACS)  the household’s expected ‘taxes-and-transfers’ (EHTT)  the man or woman’s cash property income (PPROP)  the man or woman’s occupational pensions)  52:

(10) For each woman, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then her necessary time in paid labour excluding travel to/from work (NPLX) is equal to 0.

Methodology

313

(11) For each woman, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate her necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) For each woman, her necessary labour income net of child-care costs (NLI) is equal to the household’s necessary labour income net of child-care costs (NHLI). b) For each woman, calculate her necessary time in paid labour, excluding travel to/from work (NPLX) via Equation A1.2. In Equation A1.2, set the number of children aged 2 years or less in the household to the number in the woman’s actual household. Set the number of children aged between 3 and 5 years in the household to the number in the woman’s actual household. Set the proportional responsibility for the cost of child care to 0.5. (12) For each man, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then his necessary time in paid labour excluding travel to/from work (NPLX) is equal to 0. (13) For each man, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) For each man, his necessary labour income net of child-care costs (NLI) is equal to the household’s necessary labour income net of child-care costs (NHLI). b) For each man, calculate his necessary time in paid labour excluding travel to/from work (NPLX) via Equation A1.1. In Equation A1.1, set the number of children aged 2 years or less in the household to the number in the man’s actual household. Set the number of children aged between 3 and 5 years in the household to the number in the man’s actual household. Set the hours of child care needed to the woman’s necessary time in paid labour, excluding travel to/from work (NPLX). Set the proportional responsibility for the cost of child care to 0.5.

314

Appendix 1

A1.3.12 Financially Egalitarian Divorce rule (Time-egalitarian) A1.3.12.1 Necessary time in unpaid household labour For each man, as per the Atomistic Divorce rule. For each woman, as per the Self-reliant Divorce rule. A1.3.12.2 Necessary time in paid labour, excluding travel to/from work (1) For each man and for each woman, the necessary income for the adult net of child-care costs (NAI) is equal to the poverty line. (2) For each man and for each woman, if his or her actual household receives alimony or child support and does not have children, and if he or she is the partner with the lower wage rate, then the expected alimony and child support received for the adult (EAACS) is equal to MHACS. (3) For each man and for each woman, if his or her actual household receives alimony or child support and does not have children, and if he or she is the partner with the higher wage rate, then the expected alimony and child support received for the adult (EAACS) is equal to 0. (4) For each man and for each woman, if his or her actual household receives alimony or child support and has children, then the expected alimony and child support received for the adult (EAACS) is equal to 0. (5) For each man and for each woman, if his or her actual household does not receive alimony or child support, then the expected alimony and child support received for the adult (EAACS) is equal to 0. (6) For each man and for each woman, the expected ‘taxes-andtransfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) for the adult (EATT) is equal to the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (7) For each man and for each woman, calculate his or her cash property income (PPROP) by dividing the household’s cash property income by 2.

Methodology

315

(8) For each man and for each woman, the necessary labour income for the adult net of child-care costs (NALI) = (the necessary income net of child-care costs (NAI)  the expected alimony and child support received for the adult (EAACS)  the expected ‘taxes-and-transfers’ for the adult (EATT)  the man or woman’s cash property income (PPROP)  the man or woman’s occupational pensions)  52:

(9) For each man and for each woman, if the necessary labour income for the adult net of child-care costs (NALI) is less than or equal to 0, then his or her necessary time in paid labour, excluding travel to/from work, for the adult (NAPLX) is equal to 0. (10) For each man and for each woman, if the necessary labour income for the adult net of child-care costs (NALI) is greater than 0, then calculate his or her necessary time in paid labour, excluding travel to/from work, for the adult (NAPLX) via Equation A1.1, with x replaced by NAPLX and i replaced by NALI. In Equation A1.1, set the number of children aged 2 years or less in the household and the number of children aged between 3 and 5 years in the household to 0. (11) For each man, if the necessary labour income for the adult net of child-care costs (NALI) is less than or equal to 0, then the surplus income for the man (SURMI) is equal to the necessary labour income for the adult net of child-care costs (NALI) multiplied by 52. (12) For each man, if the necessary labour income for the adult net of child-care costs (NALI) is greater than 0, then the surplus income for the man (SURMI) is equal to 0. (13) For each woman, if the necessary labour income for the adult net of child-care costs (NALI) is less than or equal to 0, then the surplus income for the woman (SURWI) is equal to the necessary labour income for the adult net of child-care costs (NALI) multiplied by 52. (14) For each woman, if the necessary labour income for the adult net of child-care costs (NALI) is greater than 0, then the surplus income for the woman (SURWI) is equal to 0. (15) For each man and woman combined, the necessary income for the children net of child-care costs (NCI) is equal to the difference between: (a) the poverty line multiplied by the square root of the number of people in the household minus 1; and (b) the poverty line.

316

Appendix 1

(16) For each man and woman combined, if their actual household receives alimony or child support and has children, then the expected alimony and child support received for the children (ECACS) is equal to MHACS. (17) For each man and woman combined, if their actual household receives alimony or child support and does not have children, then the expected alimony and child support received for the children (ECACS) is equal to 0. (18) For each man and woman combined, if their actual household does not receive alimony or child support, then the expected alimony and child support received for the children (ECACS) is equal to 0. (19) For each man and woman combined, if their actual household has children, then the expected ‘taxes-and-transfers’ (i.e., taxesand-transfers, plus regular transfers from relatives or private charitable organizations) for the children (ECTT) is equal to the difference between: (a) the household’s expected ‘taxes-andtransfers’ (EHTT) for a one-adult, one-earner household with children; and (b) the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (20) For each man and woman combined, the necessary labour income for the children net of child-care costs (NCLI) = (the necessary income for the children net of child-care costs (NCI)  the surplus income for the man (SURMI)  the surplus income for the woman (SURWI)  the expected alimony and child support received for the children (ECACS)  the expected ‘taxes-and-transfers’ for the children (ECTT) + the woman’s necessary time in paid labour, excluding travel to/from work, for the adult (NAPLX) multiplied by 52 and by the cost of child care)  52.

(21) For each man and woman combined, if the necessary labour income for the children net of child-care costs (NCLI) is less than or equal to 0, then the man’s necessary time in paid labour, excluding travel to/from work, for the children (NCPLX) is equal to 0. (22) For each man and woman combined, if the necessary labour income for the children net of child-care costs (NCLI) is less than or equal to 0, then the woman’s necessary time in paid labour, excluding travel to/from work, for the children (NCPLX) is equal to 0. (23) For each man and woman combined, if the necessary labour income for the children net of child-care costs (NCLI) is greater

Methodology

317

than 0, then calculate the man’s necessary time in paid labour, excluding travel to/from work, for the children (NCPLX) via Equation A1.7, with xmw replaced by NCPLX and j replaced by NCLI. In Equation A1.7, set the number of children aged 2 years or less in the household to the number in the man’s actual household. Set the number of children aged between 3 and 5 years in the household to the number in the man’s actual household. (24) For each man and woman combined, if the necessary labour income for the children net of child-care costs (NCLI) is greater than 0, then calculate the woman’s necessary time in paid labour, excluding travel to/from work, for the children (NCPLX) via Equation A1.7, with xmw replaced by NCPLX and j replaced by NCLI. In Equation A1.7, set the number of children aged 2 years or less in the household to the number in the woman’s actual household. Set the number of children aged between 3 and 5 years in the household to the number in the woman’s actual household. (25) For each man and for each woman, his or her necessary time in paid labour excluding travel to/from work (NPLX) is equal to the sum of NAPLX and NCPLX.

A1.3.13 Strictly Egalitarian Divorce rule (Money-egalitarian) A1.3.13.1 Necessary time in unpaid household labour (1) For each man and for each woman, his or her necessary time in unpaid household labour is equal to half of the sum of: (a) the household’s necessary time in unpaid household labour (NHUHL) for a household in which (i) the number of people is equal to that in the man or woman’s actual household, minus 1, (ii) the age of the youngest child is the same as that in the man or woman’s actual household and (iii) there is no adult who is both not employed and not self-employed (i.e., there is no adult who does not work); and (b) the household’s necessary time in unpaid household labour (NHUHL) for a household in which there is (i) one person, (ii) no children and (iii) no adult who is both not employed and not selfemployed (i.e., there is no adult who does not work). (In other words, for each man and for each woman, his or her necessary time in unpaid household labour under the Strictly Egalitarian Divorce rule (Money-egalitarian) is equal to half of the sum of his or her

318

Appendix 1

necessary time in unpaid household labour under the Atomistic Divorce rule and his or her necessary time in unpaid household labour under the Self-reliant Divorce rule.) A1.3.13.2 Necessary time in paid labour, excluding travel to/from work (1) For each man and for each woman, the household’s necessary income net of child-care costs (NHI) is equal to half of the sum of: (a) the poverty line multiplied by the square root of the number of people in the household minus 1; and (b) the poverty line. (2) For each man and for each woman, if his or her actual household receives alimony or child support and has children, then the household’s expected alimony and child support received (EHACS) is equal to MHACS divided by 2. (3) For each man and for each woman, if his or her actual household receives alimony or child support and does not have children, and if he or she is the partner with the lower wage rate, then the household’s expected alimony and child support received (EHACS) is equal to MHACS. (4) For each man and for each woman, if his or her actual household receives alimony or child support and does not have children, and if he or she is the partner with the higher wage rate, then the household’s expected alimony and child support received (EHACS) is equal to 0. (5) For each man and for each woman, if his or her actual household does not receive alimony or child support, then the household’s expected alimony and child support received (EHACS) is equal to 0. (6) For each man and for each woman, if his or her actual household has children, then the household’s expected ‘taxes-and-transfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) (EHTT) is equal to half of the sum of: (a) the household’s expected ‘taxes-and-transfers’ (EHTT) for a one-adult, one-earner household with children; and (b) the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (7) For each man and for each woman, if his or her actual household does not have children, then the household’s expected ‘taxes-andtransfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) (EHTT) is equal to

Methodology

319

the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (8) For each man and for each woman, calculate his or her cash property income (PPROP) by dividing the household’s cash property income by 2. (9) For each man and for each woman, the household’s necessary labour income net of child-care costs (NHLI) = (the household’s necessary income net of child-care costs (NHI)  the household’s expected alimony and child support received (EHACS)  the household’s expected ‘taxes-and-transfers’ (EHTT)  the man or woman’s cash property income (PPROP)  the man or woman’s occupational pensions)  52:

(10) For each woman, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then her necessary time in paid labour excluding travel to/from work (NPLX) is equal to 0. (11) For each woman, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate her necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) For each woman, her necessary labour income net of child-care costs (NLI) is equal to the household’s necessary labour income net of child-care costs (NHLI). b) For each woman, calculate her necessary time in paid labour excluding travel to/from work (NPLX) via Equation A1.2. In Equation A1.2, set the number of children aged 2 years or less in the household to the number in the woman’s actual household. Set the number of children aged between 3 and 5 years in the household to the number in the woman’s actual household. Set the proportional responsibility for the cost of child care to 0.5. (12) For each man, if the household’s necessary labour income net of child-care costs (NHLI) is less than or equal to 0, then his necessary time in paid labour excluding travel to/from work (NPLX) is equal to 0. (13) For each man, if the household’s necessary labour income net of child-care costs (NHLI) is greater than 0, then calculate his

320

Appendix 1

necessary time in paid labour excluding travel to/from work (NPLX) by executing the following algorithm. a) For each man, his necessary labour income net of child-care costs (NLI) is equal to the household’s necessary labour income net of child-care costs (NHLI). b) For each man, calculate his necessary time in paid labour excluding travel to/from work (NPLX) via Equation A1.1. In Equation A1.1, set the number of children aged 2 years or less in the household to the number in the man’s actual household. Set the number of children aged between 3 and 5 years in the household to the number in the man’s actual household. Set the hours of child care needed to the woman’s necessary time in paid labour, excluding travel to/from work (NPLX). Set the proportional responsibility for the cost of child care to 0.5. The preceding algorithm was used to calculate, for each man and for each woman, a first estimate of necessary time in paid labour excluding travel to/from work (NPLX) based on the assumption that the children (if children are present) are in the custody of the woman. In order to calculate a second estimate of NPLX based on the assumption that the children (if present) are in the custody of the man, this algorithm – in particular, Steps 10 through 13 of this algorithm – was repeated with ‘woman’ replaced by ‘man’ and ‘man’ replaced by ‘woman’. For each man and for each woman, a final estimate of necessary time in paid labour excluding travel to/from work (NPLX) was calculated by summing these first and second estimates, and then dividing the result by 2.

A1.3.14 Strictly Egalitarian Divorce rule (Time-egalitarian) A1.3.14.1 Necessary time in unpaid household labour As per the Strictly Egalitarian Divorce rule (Money-egalitarian). A1.3.14.2 Necessary time in paid labour, excluding travel to/from work (1) For each man and for each woman, the necessary income for the adult net of child-care costs (NAI) is equal to the poverty line.

Methodology

321

(2) For each man and for each woman, if his or her actual household receives alimony or child support and does not have children, and if he or she is the partner with the lower wage rate, then the expected alimony and child support received for the adult (EAACS) is equal to MHACS. (3) For each man and for each woman, if his or her actual household receives alimony or child support and does not have children, and if he or she is the partner with the higher wage rate, then the expected alimony and child support received for the adult (EAACS) is equal to 0. (4) For each man and for each woman, if his or her actual household receives alimony or child support and has children, then the expected alimony and child support received for the adult (EAACS) is equal to 0. (5) For each man and for each woman, if his or her actual household does not receive alimony or child support, then the expected alimony and child support received for the adult (EAACS) is equal to 0. (6) For each man and for each woman, the expected ‘taxes-andtransfers’ (i.e., taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) for the adult (EATT) is equal to the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (7) For each man and for each woman, calculate his or her cash property income (PPROP) by dividing the household’s cash property income by 2. (8) For each man and for each woman, the necessary labour income for the adult net of child-care costs (NALI) = (the necessary income for the adult net of child-care costs (NAI)  the expected alimony and child support received for the adult (EAACS)  the expected ‘taxes-and-transfers’ for the adult (EATT)  the man or woman’s cash property income (PPROP)  the man or woman’s occupational pensions)  52.

(9) For each man and for each woman, if the necessary labour income for the adult net of child-care costs (NALI) is less than or equal to 0, then his or her necessary time in paid labour excluding travel to/ from work for the adult (NAPLX) is equal to 0.

322

Appendix 1

(10) For each man and for each woman, if the necessary labour income for the adult net of child-care costs (NALI) is greater than 0, then calculate his or her necessary time in paid labour excluding travel to/from work for the adult (NAPLX) via Equation A1.1, with x replaced by NAPLX and i replaced by NALI. In Equation A1.1, set the number of children aged 2 years or less in the household and the number of children aged between 3 and 5 years in the household to 0. (11) For each man, if the necessary labour income for the adult net of child-care costs (NALI) is less than or equal to 0, then the surplus income for the man (SURMI) is equal to the necessary labour income for the adult net of child-care costs (NALI) multiplied by 52. (12) For each man, if the necessary labour income for the adult net of child-care costs (NALI) is greater than 0, then the surplus income for the man (SURMI) is equal to 0. (13) For each woman, if the necessary labour income for the adult net of child-care costs (NALI) is less than or equal to 0, then the surplus income for the woman (SURWI) is equal to the necessary labour income for the adult net of child-care costs (NALI) multiplied by 52. (14) For each woman, if the necessary labour income for the adult net of child-care costs (NALI) is greater than 0, then the surplus income for the woman (SURWI) is equal to 0. (15) For each man and woman combined, the necessary income for the children net of child-care costs (NCI) is equal to the difference between: (a) the poverty line multiplied by the square root of the number of people in the household minus 1; and (b) the poverty line. (16) For each man and woman combined, if their actual household receives alimony or child support and has children, then the expected alimony and child support received for the children (ECACS) is equal to MHACS. (17) For each man and woman combined, if their actual household receives alimony or child support and does not have children, then the expected alimony and child support received for the children (ECACS) is equal to 0. (18) For each man and woman combined, if their actual household does not receive alimony or child support, then the expected alimony and child support received for the children (ECACS) is equal to 0. (19) For each man and woman combined, if their actual household has children, then the expected ‘taxes-and-transfers’ (i.e.,

Methodology

323

taxes-and-transfers, plus regular transfers from relatives or private charitable organizations) for the children (ECTT) is equal to the difference between: (a) the household’s expected ‘taxes-andtransfers’ (EHTT) for a one-adult, one-earner household with children; and (b) the household’s expected ‘taxes-and-transfers’ (EHTT) for a household without children. (20) For each man and woman combined, the necessary labour income for the children net of child-care costs (NCLI) = (the necessary income for the children net of child-care costs (NCI)  the surplus income for the man (SURMI)  the surplus income for the woman (SURWI)  the expected alimony and child support received for the children (ECACS)  the expected ‘taxes-and-transfers’ for the children (ECTT) + the woman’s necessary time in paid labour, excluding travel to/from work, for the adult (NAPLX) multiplied by 52 and by the cost of child care)  52.

(21) For each man and woman combined, if the necessary labour income for the children net of child-care costs (NCLI) is less than or equal to 0, then the man’s necessary time in paid labour, excluding travel to/from work, for the children (NCPLX) is equal to 0. (22) For each man and woman combined, if the necessary labour income for the children net of child-care costs (NCLI) is less than or equal to 0, then the woman’s necessary time in paid labour, excluding travel to/from work, for the children (NCPLX) is equal to 0. (23) For each man and woman combined, if the necessary labour income for the children net of child-care costs (NCLI) is greater than 0, then calculate the man’s necessary time in paid labour, excluding travel to/from work, for the children (NCPLX) via Equation A1.7, with xmw replaced by NCPLX and j replaced by NCLI. In Equation A1.7, set the number of children aged 2 years or less in the household to the number in the man’s actual household. Set the number of children aged between 3 and 5 years in the household to the number in the man’s actual household. (24) For each man and woman combined, if the necessary labour income for the children net of child-care costs (NCLI) is greater than 0, then calculate the woman’s necessary time in paid labour,

324

Appendix 1

excluding travel to/from work, for the children (NCPLX) via Equation A1.7, with xmw replaced by NCPLX and j replaced by NCLI. In Equation A1.7, set the number of children aged 2 years or less in the household to the number in the woman’s actual household. Set the number of children aged between 3 and 5 years in the household to the number in the woman’s actual household. (25) For each man and for each woman, his or her necessary time in paid labour excluding travel to/from work (NPLX) is equal to the sum of NAPLX and NCPLX. The preceding algorithm was used to calculate, for each man and for each woman, a first estimate of necessary time in paid labour excluding travel to/from work (NPLX) based on the assumption that the children (if children are present) are in the custody of the women. In order to calculate a second estimate of NPLX based on the assumption that the children (if present) are in the custody of the man, this algorithm – in particular, Steps 20 through 25 of this algorithm – was repeated with ‘woman’ replaced by ‘man’ and ‘man’ replaced by ‘woman’. For each man and for each woman, a final estimate of necessary time in paid labour excluding travel to/from work (NPLX) was calculated by summing these first and second estimates, and then dividing the result by 2.

A1.4 Decomposing net income into gross income and mandatory social-insurance contributions in the 1994 French LIS data set In the 1994 French LIS data set, all income is recorded net of mandatory social-insurance contributions paid by employees and the selfemployed. This net income was decomposed into two components: (1) mandatory social-insurance contributions paid by employees and the self-employed; and (2) income gross of these contributions. This decomposition was based on the description of the social-security programmes in effect in France at the beginning of 1993 contained in US Social Security Administration (1994). After net income had been decomposed in this way, the ratio of mean pre-government to mean post-government income received by households in the 1994 French LIS data set was equal to 0.82. (Pre-government income received by households was calculated by adding together

Methodology

325

gross wage and salary income, farm self-employment income, non-farm self-employment income, cash property income, occupational pensions, alimony and child support received, and other cash private income. Post-government income received by households was calculated by adding pre-government income to cash sickness-insurance benefits (sick pay), accident pay, disability pay, social-retirement benefits, child and family allowances, unemployment compensation, maternity allowances, military, veteran’s and war benefits, other social insurance, means-tested cash benefits and all near-cash benefits, and then subtracting from this sum mandatory social-insurance contributions paid by employees and the self-employed and income taxes.) In comparison, the ratio of mean pre-government to mean postgovernment income received by households and private unincorporated enterprises in France in 1994 was equal to 0.89, according to the national accounts presented in OECD (1997c, p. 293). (Pre-government income received by households and private unincorporated enterprises was calculated by adding together income received in the form of the compensation of employees (excluding employers’ contributions for social security), the operating surplus of private unincorporated enterprises, property and entrepreneurial income, unfunded employee pension and welfare benefits, casualty insurance claims, transfers from the rest of the world and other transfers not elsewhere classified. Postgovernment income received by households and private unincorporated enterprises was calculated by adding pre-government income to income received in the form of social-security benefits and socialassistance grants, and then subtracting from this sum outlays disbursed in the form of social-security contributions (excluding employers’ contributions for social security) and income taxes.) The figure of 0.82 is reassuringly close to that of 0.89, especially given the general differences between income estimates derived from the 1994 French LIS data set and those derived from national accounts sources. According to information presented on the LIS website (Anonymous, n.d., p. 3), there is a difference of the order to 10 per cent between most income estimates available from the 1994 French LIS data set and those available from national accounts sources.

Appendix 2: Data

326

couples

earner

single-

earners

dual-

total

employed

home

stay-at-

employed

Without kids

482 964

women total

482

total

482

106

men

376

women

482

men

total

376

2 528 106

total men

women

1 264

1 264

women

men

N

Table A2.1. Discretionary time, US

6.24 11.31

16.37

22.61

28.67

20.93

0.00

0.00

10.17 0.00

9.36

10.98

14.06 10.58

7.09

6.48

9.03

5.77

14.67

15.46

7.13 11.84

8.62

5.63

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

care

92.54 90.96

89.37

83.75

75.14

86.14

98.17

97.38

95.54 101.00

94.85

96.23

time

7.29 13.15

19.02

26.31

33.52

24.30

0.00

0.00

11.89 0.00

10.95

12.84

14.06 10.58

7.09

6.48

9.03

5.77

14.67

15.46

7.13 11.84

8.62

5.63

labour labour

Personal Discret’y Paid

household

labour labour

Paid

Unpaid

household

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

care

91.49 89.11

86.73

80.05

70.29

82.77

98.17

97.38

93.82 101.00

93.27

94.37

time

household

Unpaid

7.29 13.15

19.02

26.31

33.52

24.30

0.00

0.00

11.89 0.00

10.95

12.84

5.63

14.06 10.58

7.09

6.48

9.03

5.77

14.67

15.46

7.13 11.84

8.62

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

care

91.49 89.11

86.73

80.05

70.29

82.77

98.17

97.38

93.82 101.00

93.27

94.37

time

Personal Discret’y

Necessary time in

Post-taxes, post-transfers, post-child-care-support

labour labour

Personal Discret’y Paid

Necessary time in

Post-taxes, post-transfers, pre-child-care-support

Unpaid

Necessary time in

Pre-taxes, pre-transfers, pre-child-care-support

stay-at-

total

total

employed

home

employed

one adult

Table A2.1. (cont.)

3 317

total

3 601

3 208

6 809

women

total

2 832 6 327

women total

men

3 495

482

total

men

376

women

106

1 462

women

men

1 855

men

N

14.29

13.16

15.31

14.88 15.37

15.77

0.00

0.00

0.00

18.17

18.49

17.92

10.44

11.39

9.58

10.86 10.12

9.51

14.67

15.46

11.84

12.84

12.84

12.84

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

88.11

88.29

87.95

87.10 87.35

87.56

98.17

97.38

101.00

81.83

81.52

82.09

time

17.25

15.90

18.45

17.98 18.54

19.00

0.00

0.00

0.00

22.34

22.72

22.03

10.44

11.39

9.58

10.86 10.12

9.51

14.67

15.46

11.84

12.84

12.84

12.84

labour labour

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16

care

85.16

85.55

84.81

84.00 84.18

84.32

98.17

97.38

101.00

77.67

77.29

77.97

time

17.25

15.90

18.45

17.98 18.54

19.00

0.00

0.00

0.00

22.34

22.72

22.03

10.44

11.39

9.58

10.86 10.12

9.51

14.67

15.46

11.84

12.84

12.84

12.84

labour labour

hold

care

Personal Discret’y Paid

house-

Personal Discret’y Paid

hold

hold

labour labour

Paid

house-

Unpaid

Unpaid

house-

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16

care

85.16

85.55

84.81

84.00 84.18

84.32

98.17

97.38

101.00

77.67

77.29

77.97

time

Personal Discret’y

Necessary time in

post-child-care-support

Post-taxes, post-transfers,

Unpaid

Necessary time in

pre-child-care-support

pre-child-care-support Necessary time in

Post-taxes, post-transfers,

Pre-taxes, pre-transfers,

stay-at-

total

total

employed

home

employed

total

employed

stay-athome

employed

one adult

couples

singleearner

earners

dual-

With kids

total

1 989

3 978 209

total men

10 468

5 609

total

4 859

women

3 767 8 479

women total

men

4 712

1 989

total

men

1 842

147

1 168

women

men

total

959

1 989

women

1 989

women

1 989

147

1 842

men

total

women

men

147 1 842

5 322

total

men women

2 661

2 661

women

men

19.72

17.14

22.73

25.34 24.28

23.41

0.00

0.00

0.00

45.65

49.20

14.94 27.92

3.17

26.71

29.88

44.83

28.74

0.00

0.00 0.00

17.24

15.09

19.38

8.81

15.23

20.07

9.56

16.88 12.71

9.31

26.10

26.72

17.95

21.37

21.60

17.67 20.21

25.88

9.46

9.24

14.79

8.81

26.10

17.95 26.72

11.99

15.17

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

77.90

75.63

80.55

70.62 75.85

80.12

86.74

86.12

94.89

45.82

42.03

80.23 64.71

83.79

76.67

73.72

53.22

75.28

86.74

94.89 86.12

83.62

82.58

84.65

21.93

18.75

25.65

27.71 26.99

26.41

0.00

0.00

0.00

45.46

49.00

15.69 27.80

3.33

28.05

31.38

47.12

30.18

0.00

0.00 0.00

21.08

18.46

23.71

8.81

15.23

20.07

9.56

16.88 12.71

9.31

26.10

26.72

17.95

21.37

21.60

17.67 20.21

25.88

9.46

9.24

14.79

8.81

26.10

17.95 26.72

11.99

15.17

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

80.32

75.69

74.03

77.63

68.24 73.13

77.12

86.74

86.12

94.89

46.01

42.24

79.48 64.83

83.63

75.33

72.22

50.93

73.84

86.74

94.89 86.12

79.77

79.21

22.19

20.57

17.04

24.71

25.18 25.33

25.44

0.00

0.00

0.00

40.07

43.02

15.69 25.34

3.33

28.05

31.38

47.12

30.18

0.00

0.00 0.00

19.66

17.13

8.81

15.23

20.07

9.56

16.88 12.71

9.31

26.10

26.72

17.95

21.37

21.60

17.67 20.21

25.88

9.46

9.24

14.79

8.81

26.10

17.95 26.72

11.99

15.17

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

81.84

77.04

75.74

78.57

70.77 74.80

78.09

86.74

86.12

94.89

51.40

48.21

79.48 67.28

83.63

75.33

72.22

50.93

73.84

86.74

94.89 86.12

81.19

80.54

couples

earner

single-

earners

dual-

Total

total

employed

home

stay-at-

employed

Table A2.1. (cont.)

2 471

4 942

women

total

2 471

2 471

total

men

2 218 253

2 471

2 218

men women

total

women

253

7 850

total

men

3 925

3 925

women

men

N

14.20

3.79

24.61

28.40

27.36 37.73

0.00

0.00

0.00

14.91

13.20

16.61

7.76

16.23

23.48

8.98

8.68

8.28 12.26

23.78

24.73

15.27

10.39

13.01

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

care

82.41

85.57

79.25

75.76

77.20 62.85

89.06

88.11

97.57

87.55

86.63

88.47

time

15.17

4.14

26.21

30.35

29.14 41.14

0.00

0.00

0.00

18.05

15.98

20.12

16.23

23.48

8.98

8.68

8.28 12.26

23.78

24.73

15.27

10.39

13.01

7.76

labour labour

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

care

81.44

85.23

77.65

73.81

75.42 59.44

89.06

88.11

97.57

84.40

83.85

84.96

time

15.17

4.14

26.21

30.35

29.14 41.14

0.00

0.00

0.00

17.10

15.09

19.10

16.23

23.48

8.98

8.68

8.28 12.26

23.78

24.73

15.27

10.39

13.01

7.76

labour labour

hold

housePersonal Discret’y Paid

hold

hold

Personal Discret’y Paid

house-

labour labour

Paid

Unpaid

Unpaid

house-

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

55.16

55.16

55.16

care

81.44

85.23

77.65

73.81

75.42 59.44

89.06

88.11

97.57

85.36

84.74

85.98

time

Personal Discret’y

Necessary time in

post-child-care-support

Post-taxes, post-transfers,

Unpaid

Necessary time in

pre-child-care-support

pre-child-care-support Necessary time in

Post-taxes, post-transfers,

Pre-taxes, pre-transfers,

stay-at-

total

total

employed

home

employed

one adult

4 485

total

17 277

8 817

women

total

8 460

men

14 806

6 599

total

8 207

women

2 218 2 471

men

women total

253

2 421

women

men

2 064

men

17.52

15.65

19.48

20.36

20.74

20.05

0.00 0.00

0.00

25.25

30.59

18.86

13.28

16.81

9.57

11.57

14.24

9.40

24.73 23.78

15.27

15.04

16.29

13.53

55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

80.44

82.04

80.38

83.79

80.90

77.86

83.39

88.11 89.06

97.57

72.55

65.96

22.58

20.03

17.68

22.49

23.28

23.43

23.16

0.00 0.00

0.00

28.30

33.07

13.53

13.28

16.81

9.57

11.57

14.24

9.40

24.73 23.78

15.27

15.04

16.29

55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

76.73

79.53

78.35

80.78

77.98

75.17

80.28

88.11 89.06

97.57

69.51

63.48

22.35

19.22

16.61

21.96

22.35

22.02

22.61

0.00 0.00

0.00

26.91

30.72

13.53

13.28

16.81

9.57

11.57

14.24

9.40

24.73 23.78

15.27

15.04

16.29

55.16

55.16

55.16

55.16

55.16

55.16

55.16

55.16 55.16

55.16

55.16

55.16

76.96

80.34

79.42

81.31

78.92

76.59

80.82

88.11 89.06

97.57

70.89

65.83

332

Appendix 2

Table A2.2. Spare time, US Actual time in

N

Paid

Unpaid household

Personal

Spare

labour

labour

care

time

Without kids dual-earners

single-earner

employed

stay-at-home

couples employed

total

one adult

total

employed

stay-at-home

employed

total

men

632

47.81

17.42

66.04

36.73

women

732

39.72

24.48

70.82

32.98

1 364 45

43.86 0.00

20.87 31.27

68.38 72.60

34.90 64.14

total men women

150

0.00

35.52

77.38

55.10

total

195

0.00

34.35

76.07

57.58 34.77

men

153

49.76

15.71

67.75

women

114

37.14

24.38

73.45

33.03

total

267

45.11

18.91

69.85

34.13

men

198

38.50

19.23

68.85

41.42

women total

264 462

15.98 26.96

30.73 25.12

75.69 72.35

45.61 43.56

men

781

46.45

15.74

67.51

38.30

women

753

44.57

18.82

70.68

33.93

total

1 534

45.60

17.13

68.94

36.33

men

45

0.00

31.27

72.60

64.14

women

150

0.00

35.52

77.38

55.10

total

195

0.00

34.35

76.07

57.58

men women

1 566 1 599

47.32 41.86

16.43 21.77

66.93 70.93

37.32 33.44

total

3 165

44.81

18.89

68.77

35.53

men

1 611

46.04

16.83

67.09

38.05

women

1 749

38.52

22.87

71.44

35.16

total

3 360

42.47

19.69

69.15

36.68

men

1 545

47.53

23.49

65.96

31.02

women

1 883

32.34

37.48

69.49

28.69

total

3 428

39.86

30.56

67.74

29.84

With kids dual-earners

single-earner

employed

stay-at-home

couples employed

men

90

0.00

43.50

74.83

49.66

women total

773 863

0.00 0.00

57.99 56.24

73.19 73.38

36.83 38.38 30.04

men

786

48.78

21.20

67.98

women

141

35.15

34.75

71.30

26.80

total

927

47.00

22.97

68.42

29.61

Data

333

Table A2.2. (cont.) Actual time in

total

one adult

employed

total

stay-at-home

employed

total

Total dual-earners

single-earner couples

employed

stay-at-home

employed

total

one adult

employed

N

Unpaid Paid household Personal Spare labour labour care time

men women total men women total men women total men women total men women total

876 914 1 790 198 770 968 90 773 863 2 529 2 794 5 323 2 619 3 567 6 186

44.06 5.70 26.54 42.06 37.49 38.54 0.00 0.00 0.00 47.51 33.90 40.97 45.96 26.82 35.81

23.36 54.22 37.45 25.03 29.88 28.77 43.50 57.99 56.24 22.89 35.25 28.83 23.56 40.00 32.28

68.64 72.88 70.58 67.01 70.99 70.08 74.83 73.19 73.38 66.67 69.99 68.27 66.94 70.66 68.91

31.94 35.20 33.43 33.90 29.64 30.62 49.66 36.83 38.38 30.93 28.85 29.93 31.54 30.52 31.00

men women total men women total men women total men women total men women total

2 177 2 615 4 792 135 923 1 058 939 255 1 194 1 074 1 178 2 252 979 1 523 2 502

47.63 34.95 41.31 0.00 0.00 0.00 48.99 36.16 46.49 42.74 8.58 26.65 45.80 41.61 43.52

21.25 32.89 27.06 38.35 53.29 50.93 20.04 29.48 21.88 22.38 47.64 34.28 17.13 23.43 20.57

65.99 69.96 67.97 73.89 74.06 74.04 67.93 72.39 68.80 68.69 73.67 71.04 67.43 70.81 69.27

33.12 30.20 31.67 55.76 40.65 43.04 31.04 29.97 30.83 34.19 38.11 36.04 37.64 32.14 34.65

334

Appendix 2

Table A2.2. (cont.) Actual time in

total

stay-at-home

employed

total

men women total men women total men women total

N

Unpaid Paid household Personal Spare labour labour care time

135 923 1 058 4 095 4 393 8 488 4 230 5 316 9 546

0.00 0.00 0.00 47.42 37.45 42.72 46.00 31.61 38.72

38.35 53.29 50.93 19.88 29.24 24.29 20.43 32.99 26.79

73.89 74.06 74.04 66.79 70.41 68.50 67.01 70.98 69.02

55.76 40.65 43.04 33.91 30.90 32.49 34.56 32.42 33.48

Data

335

Table A2.3. The state impact on discretionary time and the time-pressure illusion, US

Impact of state taxesand-transfers on discret’y time Without kids dual-earners employed

single-earner stay-atcouples home employed

total

one adult

employed

total

stay-athome employed

total

With kids dual-earners employed

Impact of state child-care support on discret’y time

Total state impact on Timediscret’y pressure time illusion

men women total men women total men women total men women total men women total men women total men women total men women total

1.86 1.58 1.72 0.00 0.00 0.00 3.38 4.85 3.70 2.64 1.05 1.85 4.12 4.23 4.17 0.00 0.00 0.00 3.23 3.10 3.17 3.14 2.74 2.95

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

1.86 1.58 1.72 0.00 0.00 0.00 3.38 4.85 3.70 2.64 1.05 1.85 4.12 4.23 4.17 0.00 0.00 0.00 3.23 3.10 3.17 3.14 2.74 2.95

57.64 60.29 58.92 36.86 42.28 40.59 47.99 37.26 45.92 45.32 45.88 45.55 39.67 43.35 41.34 36.86 42.28 40.59 47.00 50.57 48.65 46.75 50.39 48.48

men women total

4.32 3.37 3.85

1.52 1.33 1.42

2.80 2.04 2.42

50.83 51.85 51.35

336

Appendix 2

Table A2.3. (cont.)

Impact of state taxesand-transfers on discret’y time single-earner stay-atcouples home employed

total

one adult

employed

total

stay-athome employed

total

Total dual-earners employed

single-earner stay-atcouples home employed

Impact of state child-care support on discret’y time

Total state impact on Timediscret’y pressure time illusion

men women total men women total men women total men women total men women total men women total men women total

0.00 0.00 0.00 1.44 2.29 1.50 1.34 0.16 0.75 0.13 0.20 0.19 0.00 0.00 0.00 3.00 2.37 2.72 2.92 1.60 2.21

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2.45 5.97 5.39 0.00 0.00 0.00 0.97 2.53 1.67 0.94 1.71 1.36

0.00 0.00 0.00 1.44 2.29 1.50 1.34 0.16 0.75 2.58 6.18 5.58 0.00 0.00 0.00 2.03 0.16 1.05 1.98 0.11 0.85

45.22 49.29 48.36 43.81 24.13 42.61 43.39 48.43 46.05 33.38 18.57 20.78 45.22 49.29 48.36 47.16 41.92 44.87 47.03 45.22 46.05

men women total men women total men women total

3.51 2.78 3.15 0.00 0.00 0.00 1.78 3.41 1.94

1.02 0.89 0.95 0.00 0.00 0.00 0.00 0.00 0.00

2.49 1.89 2.19 0.00 0.00 0.00 1.78 3.41 1.94

52.85 54.54 53.69 41.82 47.46 46.02 44.38 29.47 42.99

Data

337

Table A2.3. (cont.)

Impact of state taxesand-transfers on discret’y time total

one adult

employed

total

stay-athome employed

total

men women total men women total men women total men women total men women total

1.60 0.34 0.97 3.71 2.48 3.04 0.00 0.00 0.00 3.10 2.69 2.92 3.02 2.03 2.51

Impact of state child-care support on discret’y time 0.00 0.00 0.00 0.23 2.35 1.39 0.00 0.00 0.00 0.54 1.42 0.94 0.53 1.07 0.81

Total state impact on Timediscret’y pressure time illusion 1.60 0.34 0.97 3.48 0.13 1.65 0.00 0.00 0.00 2.56 1.27 1.98 2.49 0.96 1.71

43.46 47.11 45.40 39.32 33.69 36.25 41.82 47.46 46.02 46.92 45.69 46.43 46.74 47.00 46.86

338

Appendix 2

Table A2.4. Discretionary time (post-taxes, post-transfers, post-child-caresupport) under alternative household rules, US Necessary time in

N Conventional dual-earner without kids men 1 264 women 1 264 total 2 528 with kids men 2 661 women 2 661 total 5 322 total men 3 925 women 3 925 total 7 850 Male breadwinner without kids men women total with kids men women total total men women total

Paid labour

Unpaid household labour

Personal care

Discret’y time

12.84 10.95 11.89

5.63 8.62 7.13

55.16 55.16 55.16

94.37 93.27 93.82

22.19 17.13 19.66

8.81 15.17 11.99

55.16 55.16 55.16

81.84 80.54 81.19

19.10 15.09 17.10

7.76 13.01 10.39

55.16 55.16 55.16

85.98 84.74 85.36

1 264 1 264 2 528

23.73 0.00 11.87

5.77 15.46 10.62

55.16 55.16 55.16

83.34 97.38 90.36

2 661 2 661 5 322

31.52 0.00 15.76

8.44 25.59 17.01

55.16 55.16 55.16

72.88 87.25 80.07

3 925 3 925 7 850

28.95 0.00 14.48

7.56 22.25 14.90

55.16 55.16 55.16

76.33 90.59 83.46

12.56 6.74 9.65

8.04 17.61 12.82

55.16 55.16 55.16

92.24 88.50 90.37

Most-efficient breadwinner without kids men 1 258 women 1 258 total 2 516

Data

339

Table A2.4. (cont.) Necessary time in

with kids men women total total men women total

N

Paid labour

Unpaid household labour

2 647 2 647 5 294

17.01 8.76 12.89

11.84 24.28 18.06

55.16 55.16 55.16

83.99 79.80 81.89

3 905 3 905 7 810

15.54 8.09 11.82

10.58 22.07 16.33

55.16 55.16 55.16

86.72 82.67 84.70

12.52 15.86 14.19

9.08 9.08 9.08

55.16 55.16 55.16

91.24 87.90 89.57

19.51 26.83 23.17

13.38 13.38 13.38

55.16 55.16 55.16

79.95 72.64 76.30

17.19 23.18 20.18

11.95 11.95 11.95

55.16 55.16 55.16

83.71 77.71 80.71

11.93 11.93 11.93

9.08 9.08 9.08

55.16 55.16 55.16

91.84 91.84 91.84

20.18 20.18 20.18

13.38 13.38 13.38

55.16 55.16 55.16

79.28 79.28 79.28

17.46 17.46 17.46

11.96 11.96 11.96

55.16 55.16 55.16

83.42 83.42 83.42

Equal monetary contribution without kids men 1 263 women 1 263 total 2 526 with kids men 2 626 women 2 626 total 5 252 total men 3 889 women 3 889 total 7 778 Equal temporal contribution without kids men 1 264 women 1 264 total 2 528 with kids men 2 661 women 2 661 total 5 322 total men 3 925 women 3 925 total 7 850

Personal care

Discret’y time

340

Appendix 2

Table A2.4. (cont.) Necessary time in

Atomistic divorce without kids men women total with kids men women total total men women total Self-reliant divorce without kids men women total with kids men women total total men women total Gendered divorce without kids men women total

N

Paid labour

Unpaid household labour

1 264 1 264 2 528

18.56 24.15 21.35

12.84 12.84 12.84

55.16 55.16 55.16

81.45 75.86 78.65

2 661 2 661 5 322

19.68 26.78 23.23

12.84 12.84 12.84

55.16 55.16 55.16

80.33 73.22 76.77

3 925 3 925 7 850

19.31 25.91 22.61

12.84 12.84 12.84

55.16 55.16 55.16

80.70 74.09 77.39

1 264 1 264 2 528

18.56 24.15 21.35

12.84 12.84 12.84

55.16 55.16 55.16

81.45 75.86 78.65

2 661 2 661 5 322

30.65 42.23 36.44

22.98 22.98 22.98

55.16 55.16 55.16

59.21 47.63 53.42

3 925 3 925 7 850

26.66 36.27 31.47

19.63 19.63 19.63

55.16 55.16 55.16

66.54 56.94 61.74

1 264 1 264 2 528

18.56 24.15 21.35

12.84 12.84 12.84

55.16 55.16 55.16

81.45 75.86 78.65

Personal care

Discret’y time

Data

341

Table A2.4. (cont.) Necessary time in

with kids men women total total men women total

N

Paid labour

Unpaid household labour

2 661 2 661 5 322

19.68 42.23 30.95

12.84 22.98 17.91

55.16 55.16 55.16

80.33 47.63 63.98

3 925 3 925 7 850

19.31 36.27 27.79

12.84 19.63 16.23

55.16 55.16 55.16

80.70 56.94 68.82

18.56 24.15 21.35

12.84 12.84 12.84

55.16 55.16 55.16

81.45 75.86 78.65

28.86 26.78 27.82

12.84 22.98 17.91

55.16 55.16 55.16

71.14 63.08 67.11

25.46 25.91 25.69

12.84 19.63 16.23

55.16 55.16 55.16

74.54 67.29 70.92

12.84 12.84 12.84

55.16 55.16 55.16

81.42 75.82 78.62

16.18 19.64 17.91

55.16 55.16 55.16

71.53 63.33 67.43

15.07 17.39 16.23

55.16 55.16 55.16

74.80 67.46 71.13

Male-breadwinner divorce without kids men 1 264 women 1 264 total 2 528 with kids men 2 661 women 2 661 total 5 322 total men 3 925 women 3 925 total 7 850

Most-efficient breadwinner divorce without kids men 1 258 18.59 women 1 258 24.19 total 2 516 21.39 with kids men 2 647 25.14 women 2 647 29.87 total 5 294 27.50 total men 3 905 22.97 women 3 905 27.99 total 7 810 25.48

Personal care

Discret’y time

342

Appendix 2

Table A2.4. (cont.) Necessary time in

N

Paid labour

Unpaid household labour

Financially egalitarian divorce (money-egalitarian) without kids men 1 264 18.56 12.84 women 1 264 24.15 12.84 total 2 528 21.35 12.84 with kids men 2 661 25.53 12.84 women 2 661 34.02 22.98 total 5 322 29.78 17.91 total men 3 925 23.23 12.84 women 3 925 30.77 19.63 total 7 850 27.00 16.23 Financially egalitarian divorce (time-egalitarian) without kids men 1 264 18.56 12.84 women 1 264 24.15 12.84 total 2 528 21.35 12.84 with kids men 2 661 23.66 12.84 women 2 661 30.64 22.98 total 5 322 27.15 17.91 total men 3 925 21.98 12.84 women 3 925 28.50 19.63 total 7 850 25.24 16.23 Strictly egalitarian divorce (money-egalitarian) without kids men 1 264 18.56 12.84 women 1 264 24.15 12.84 total 2 528 21.35 12.84

Personal care

Discret’y time

55.16 55.16 55.16

81.45 75.86 78.65

55.16 55.16 55.16

74.48 55.84 65.16

55.16 55.16 55.16

76.78 62.44 69.61

55.16 55.16 55.16

81.45 75.86 78.65

55.16 55.16 55.16

76.35 59.22 67.78

55.16 55.16 55.16

78.03 64.71 71.37

55.16 55.16 55.16

81.45 75.86 78.65

Data

343

Table A2.4. (cont.) Necessary time in

with kids men women total total men women total

N

Paid labour

Unpaid household labour

2 661 2 661 5 322

25.11 33.65 29.38

17.91 17.91 17.91

55.16 55.16 55.16

69.82 61.29 65.55

3 925 3 925 7 850

22.95 30.51 26.73

16.23 16.23 16.23

55.16 55.16 55.16

73.65 66.09 69.87

12.84 12.84 12.84

55.16 55.16 55.16

81.45 75.86 78.65

17.91 17.91 17.91

55.16 55.16 55.16

71.28 63.94 67.61

16.23 16.23 16.23

55.16 55.16 55.16

74.63 67.87 71.25

Strictly egalitarian divorce (time-egalitarian) without kids men 1 264 18.56 women 1 264 24.15 total 2 528 21.35 with kids men 2 661 23.65 women 2 661 30.99 total 5 322 27.32 total men 3 925 21.97 women 3 925 28.74 total 7 850 25.35

Personal care

Discret’y time

singleearner couples

total

employed

stay-athome

Without kids dualemployed earners

men women total men women total men women total men women total

526 526 1 052 12 100 112 100 12 112 112 112 224

N

Table A2.5. Discretionary time, Australia

10.09 8.12 9.10 0.00 0.00 0.00 20.73 16.31 20.17 18.10 2.07 10.08

6.49 12.68 9.59 14.44 19.98 19.28 6.79 12.34 7.49 7.76 19.01 13.39

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

91.45 87.23 89.34 93.59 88.05 88.75 80.51 79.39 80.37 82.17 86.95 84.56

10.39 8.37 9.38 0.00 0.00 0.00 21.32 16.77 20.74 18.62 2.12 10.37

6.49 12.68 9.59 14.44 19.98 19.28 6.79 12.34 7.49 7.76 19.01 13.39

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

91.15 86.99 89.07 93.59 88.05 88.75 79.92 78.93 79.80 81.65 86.89 84.27

10.39 8.37 9.38 0.00 0.00 0.00 21.32 16.77 20.74 18.62 2.12 10.37

6.49 12.68 9.59 14.44 19.98 19.28 6.79 12.34 7.49 7.76 19.01 13.39

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

91.15 86.99 89.07 93.59 88.05 88.75 79.92 78.93 79.80 81.65 86.89 84.27

Unpaid household Personal Discret’y labour care time

Necessary time in

Post-taxes, post-transfers, post-child-care-support

Unpaid household Personal Discret’y Paid labour care time labour

Necessary time in

Post-taxes, post-transfers, pre-child-care-support

Unpaid housePaid hold Personal Discret’y Paid labour labour care time labour

Necessary time in

Pre-taxes, pre-transfers, pre-child-care-support

singleearner couples

With kids dualearners

stay-athome

total

employed

stay-athome

employed

total

employed

employed

one adult

men women total men women total men women total

men women total men women total men women total men women total

1 195 1 195 2 390 13 749 762 749 13 762

464 285 749 12 100 112 1 090 823 1 913 1 102 923 2 025

17.73 10.92 14.32 0.00 0.00 0.00 27.58 35.68 27.71

14.20 16.80 15.17 0.00 0.00 0.00 12.76 11.14 12.06 12.60 9.96 11.40

9.40 20.93 15.16 20.41 31.82 31.64 10.80 18.77 10.93

13.56 13.56 13.56 14.44 19.98 19.28 9.48 12.96 10.98 9.54 13.71 11.44

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

80.91 76.19 78.55 87.62 76.21 76.40 69.66 53.58 69.40

80.28 77.68 79.31 93.59 88.05 88.75 85.79 83.93 84.99 85.89 84.36 85.20

18.91 11.63 15.27 0.00 0.00 0.00 29.47 38.04 29.61

14.73 17.39 15.73 0.00 0.00 0.00 13.18 11.51 12.46 13.02 10.29 11.78

9.40 20.93 15.16 20.41 31.82 31.64 10.80 18.77 10.93

13.56 13.56 13.56 14.44 19.98 19.28 9.48 12.96 10.98 9.54 13.71 11.44

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

79.73 75.48 77.61 87.62 76.21 76.40 67.76 51.22 67.49

79.75 77.09 78.75 93.59 88.05 88.75 85.37 83.56 84.59 85.48 84.04 84.82

18.48 11.34 14.91 0.00 0.00 0.00 29.47 38.04 29.61

14.73 17.39 15.73 0.00 0.00 0.00 13.18 11.51 12.46 13.02 10.29 11.78

9.40 20.93 15.16 20.41 31.82 31.64 10.80 18.77 10.93

13.56 13.56 13.56 14.44 19.98 19.28 9.48 12.96 10.98 9.54 13.71 11.44

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

80.16 75.77 77.97 87.62 76.21 76.40 67.76 51.22 67.49

79.75 77.09 78.75 93.59 88.05 88.75 85.37 83.56 84.59 85.48 84.04 84.82

stay-athome

total

total

employed

employed

one adult

total

Table A2.5. (cont.)

men women total men women total men women total men women total men

762 762 1 524 42 178 220 13 749 762 1 986 1 386 3 372 1 999

N

27.13 0.57 13.85 22.23 30.84 29.24 0.00 0.00 0.00 21.56 13.50 18.27 21.42

10.96 31.61 21.28 22.01 23.67 23.36 20.41 31.82 31.64 10.17 21.23 14.68 10.23

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

69.94 75.85 72.90 63.79 53.52 55.43 87.62 76.21 76.40 76.31 73.30 75.08 76.38

29.00 0.61 14.81 17.42 25.08 23.66 0.00 0.00 0.00 22.90 13.46 19.05 22.76

10.96 31.61 21.28 22.01 23.67 23.36 20.41 31.82 31.64 10.17 21.23 14.68 10.23

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

68.08 75.81 71.94 68.60 59.28 61.01 87.62 76.21 76.40 74.97 73.34 74.30 75.05

29.00 0.61 14.81 17.12 23.92 22.66 0.00 0.00 0.00 22.63 13.07 18.73 22.49

10.96 31.61 21.28 22.01 23.67 23.36 20.41 31.82 31.64 10.17 21.23 14.68 10.23

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

68.08 75.81 71.94 68.90 60.44 62.01 87.62 76.21 76.40 75.23 73.73 74.62 75.31

Unpaid household Personal Discret’y labour care time

Necessary time in

Post-taxes, post-transfers, post-child-care-support

Unpaid household Personal Discret’y Paid labour care time labour

Necessary time in

Post-taxes, post-transfers, pre-child-care-support

Unpaid housePaid hold Personal Discret’y Paid labour labour care time labour

Necessary time in

Pre-taxes, pre-transfers, pre-child-care-support

stay-athome

total

employed

employed

total

employed

stay-athome

employed

one adult

singleearner couples

Total dualearners

men women total men women total men women total men women total men women total men women total men

women total

1 721 1 721 3 442 25 849 874 849 25 874 874 874 1 748 506 463 969 25 849 874 3 076

2 135 4 134

15.38 10.06 12.72 0.00 0.00 0.00 26.80 25.38 26.76 25.99 0.76 13.38 14.81 22.06 18.21 0.00 0.00 0.00 18.46

8.70 14.87

8.50 18.39 13.44 17.24 30.48 30.08 10.35 15.35 10.50 10.55 30.02 20.29 14.20 17.35 15.68 17.24 30.48 30.08 9.92

25.00 17.84

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

59.97 59.97

84.16 79.59 81.87 90.80 77.55 77.95 70.89 67.30 70.78 71.48 77.25 74.37 79.03 68.62 74.15 90.80 77.55 77.95 79.65

74.33 75.33

16.28 10.62 13.45 0.00 0.00 0.00 28.55 26.73 28.49 27.69 0.80 14.25 14.93 20.27 17.44 0.00 0.00 0.00 19.48

8.67 15.50

8.50 18.39 13.44 17.24 30.48 30.08 10.35 15.35 10.50 10.55 30.02 20.29 14.20 17.35 15.68 17.24 30.48 30.08 9.92

25.00 17.84

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

59.97 59.97

83.25 79.02 81.14 90.80 77.55 77.95 69.14 65.95 69.04 69.79 77.21 73.50 78.90 70.41 74.92 90.80 77.55 77.95 78.63

74.36 74.69

15.99 10.42 13.20 0.00 0.00 0.00 28.55 26.73 28.49 27.69 0.80 14.25 14.91 19.84 17.22 0.00 0.00 0.00 19.30

8.42 15.24

8.50 18.39 13.44 17.24 30.48 30.08 10.35 15.35 10.50 10.55 30.02 20.29 14.20 17.35 15.68 17.24 30.48 30.08 9.92

25.00 17.84

59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97 59.97

59.97 59.97

83.55 79.22 81.38 90.80 77.55 77.95 69.14 65.95 69.04 69.79 77.21 73.50 78.93 70.85 75.13 90.80 77.55 77.95 78.80

74.61 74.95

total

Table A2.5. (cont.)

women total men women total

2 209 5 285 3 101 3 058 6 159

N

12.62 16.03 18.30 9.08 13.73

18.14 13.34 9.99 21.61 15.74

59.97 59.97 59.97 59.97 59.97

77.27 78.66 79.75 77.35 78.56

12.73 16.67 19.31 9.16 14.28

18.14 13.34 9.99 21.61 15.74

59.97 59.97 59.97 59.97 59.97

77.16 78.02 78.74 77.27 78.01

12.49 16.47 19.14 8.98 14.11

18.14 13.34 9.99 21.61 15.74

59.97 59.97 59.97 59.97 59.97

77.40 78.22 78.91 77.45 78.18

Unpaid household Personal Discret’y labour care time

Necessary time in

Post-taxes, post-transfers, post-child-care-support

Unpaid household Personal Discret’y Paid labour care time labour

Necessary time in

Post-taxes, post-transfers, pre-child-care-support

Unpaid housePaid hold Personal Discret’y Paid labour labour care time labour

Necessary time in

Pre-taxes, pre-transfers, pre-child-care-support

Data

349

Table A2.6. Spare time, Australia Actual time in

N Without kids dual-earners employed

single-earner couples

stay-at-home

employed

total

one adult

employed

total

stay-at-home

employed

total

With kids dual-earners

single-earner couples

employed

stay-at-home

Unpaid housePaid hold labour labour

Personal Spare care time

men 468 51.31 women 464 37.43 total 932 44.39 men 11 0.00 women 57 0.00 total 68 0.00 men 57 36.65 women 10 14.29 total 67 33.34 men 68 30.81 women 67 2.12 total 135 16.56 men 325 45.79 women 234 36.22 total 559 41.80 men 11 0.00 women 57 0.00 total 68 0.00 men 850 48.47 women 708 36.76 total 1 558 43.11 men 861 47.80 women 765 33.80 total 1 626 41.15

14.06 25.74 19.88 21.96 40.78 37.78 15.35 23.30 16.53 16.40 38.19 27.23 16.24 20.52 18.03 21.96 40.78 37.78 14.85 24.30 19.18 14.95 25.62 20.02

72.67 75.96 74.31 94.66 82.66 84.57 77.83 87.74 79.29 80.51 83.41 81.95 71.82 77.45 74.17 94.66 82.66 84.57 72.78 76.54 74.50 73.08 77.03 74.96

29.96 28.87 29.42 51.38 44.56 45.65 38.18 42.66 38.84 40.28 44.28 42.27 34.15 33.81 34.01 51.38 44.56 45.65 31.90 30.41 31.21 32.17 31.54 31.87

men 789 47.73 women 800 22.77 total 1 589 35.15 men 24 0.00 women 431 0.00 total 455 0.00

20.98 44.80 32.98 40.02 59.81 58.78

71.38 73.95 72.67 79.57 76.57 76.73

27.91 26.48 27.19 48.41 31.62 32.49

350

Appendix 2

Table A2.6. (cont.) Actual time in

employed

total

one adult

employed

total

stay-at-home

employed

total

Total dual-earners

single-earner couples

employed

stay-at-home

employed

total

one adult

employed

N

Unpaid housePaid hold labour labour

Personal Spare care time

men women total men women total men women total men women total men women total men women total

434 23 457 458 454 912 8 103 111 24 431 455 1 231 926 2 157 1 255 1 357 2 612

44.71 20.83 43.52 42.41 1.04 21.78 34.42 23.03 23.75 0.00 0.00 0.00 46.60 22.75 36.45 45.72 15.42 30.04

22.27 36.51 22.97 23.18 58.64 40.86 33.76 39.26 38.91 40.02 59.81 58.78 21.49 44.07 31.10 21.84 49.13 35.97

73.52 76.29 73.66 73.84 76.56 75.19 70.95 75.27 75.00 79.57 76.57 76.73 72.14 74.13 72.99 72.28 74.92 73.64

27.50 34.36 27.84 28.58 31.76 30.16 28.88 30.45 30.35 48.41 31.62 32.49 27.77 27.05 27.47 28.16 28.52 28.35

men women total men women total men women total men women total men women total

1 257 1 264 2 521 35 488 523 491 33 524 526 521 1 047 333 337 670

49.05 28.10 38.54 0.00 0.00 0.00 43.79 18.87 42.24 40.93 1.17 21.12 45.52 31.78 38.48

18.43 37.87 28.18 34.42 57.63 56.11 21.48 32.56 22.16 22.32 56.07 39.14 16.66 26.82 21.87

71.86 74.68 73.27 84.24 77.27 77.73 74.01 79.72 74.37 74.68 77.42 76.05 71.80 76.72 74.32

28.67 27.35 28.00 49.33 33.10 34.16 28.72 36.85 29.22 30.06 33.33 31.69 34.03 32.68 33.34

Data

351

Table A2.6. (cont.) Actual time in

total

stay-at-home

employed

total

men women total men women total men women total

N

Unpaid housePaid hold labour labour

Personal Spare care time

35 488 523 2 081 1 634 3 715 2 116 2 122 4 238

0.00 0.00 0.00 47.31 28.53 39.08 46.51 21.69 34.05

84.24 77.27 77.73 72.38 75.13 73.58 72.58 75.64 74.12

34.42 57.63 56.11 18.96 35.91 26.39 19.22 41.11 30.22

49.33 33.10 34.16 29.34 28.44 28.95 29.68 29.55 29.62

352

Appendix 2

Table A2.7. The state impact on discretionary time and the time-pressure illusion, Australia

Impact of state taxesand-transfers on discret’y time Without kids dual-earners employed

single-earner stay-atcouples home employed

total

one adult

employed

total

stay-athome employed

total

With kids dual-earners employed

Impact of state child-care support on discret’y time

Total state impact on Timediscret’y pressure time illusion

men women total men women total men women total men women total men women total men women total men women total men women total

0.30 0.25 0.27 0.00 0.00 0.00 0.59 0.46 0.57 0.52 0.06 0.29 0.53 0.59 0.56 0.00 0.00 0.00 0.42 0.37 0.40 0.42 0.33 0.38

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

0.30 0.25 0.27 0.00 0.00 0.00 0.59 0.46 0.57 0.52 0.06 0.29 0.53 0.59 0.56 0.00 0.00 0.00 0.42 0.37 0.40 0.42 0.33 0.38

61.19 58.12 59.66 42.21 43.49 43.11 41.74 36.27 40.96 41.37 42.62 42.01 45.60 43.28 44.74 42.21 43.49 43.11 53.47 53.16 53.38 53.31 52.50 52.95

men women total

1.18 0.71 0.95

0.43 0.29 0.36

0.75 0.42 0.58

52.25 49.29 50.78

Data

353

Table A2.7. (cont.)

Impact of state taxesand-transfers on discret’y time single-earner stay-atcouples home employed

total

one adult

employed

total

stay-athome employed

total

Total dual-earners employed

single-earner stay-atcouples home employed

Impact of state child-care support on discret’y time

Total state impact on Timediscret’y pressure time illusion

men women total men women total men women total men women total men women total men women total men women total

0.00 0.00 0.00 1.90 2.36 1.91 1.87 0.04 0.95 4.81 5.76 5.58 0.00 0.00 0.00 1.34 0.04 0.78 1.33 0.03 0.63

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.30 1.16 1.00 0.00 0.00 0.00 0.26 0.39 0.32 0.26 0.25 0.26

0.00 0.00 0.00 1.90 2.36 1.91 1.87 0.04 0.95 5.11 6.92 6.58 0.00 0.00 0.00 1.08 0.44 0.46 1.07 0.28 0.37

39.21 44.59 43.91 40.26 16.86 39.65 39.50 44.05 41.78 40.02 29.99 31.67 39.21 44.59 43.91 47.46 46.68 47.15 47.15 46.09 46.60

men women total men women total men women total

0.91 0.57 0.74 0.00 0.00 0.00 1.75 1.35 1.74

0.30 0.20 0.25 0.00 0.00 0.00 0.00 0.00 0.00

0.61 0.37 0.49 0.00 0.00 0.00 1.75 1.35 1.74

54.88 51.88 53.38 41.46 44.45 43.79 40.42 29.10 39.82

354

Appendix 2

Table A2.7. (cont.)

Impact of state taxesand-transfers on discret’y time total

one adult

employed

total

stay-athome employed

total

men women total men women total men women total men women total men women total

1.70 0.04 0.87 0.13 1.79 0.77 0.00 0.00 0.00 1.02 0.11 0.64 1.01 0.08 0.55

Impact of state child-care support on discret’y time 0.00 0.00 0.00 0.02 0.44 0.22 0.00 0.00 0.00 0.17 0.25 0.20 0.17 0.18 0.17

Total state impact on Timediscret’y pressure time illusion 1.70 0.04 0.87 0.11 2.22 0.99 0.00 0.00 0.00 0.85 0.14 0.44 0.84 0.10 0.38

39.72 43.87 41.80 44.90 38.17 41.80 41.46 44.45 43.79 49.46 48.97 49.27 49.22 47.89 48.56

Data

355

Table A2.8. Discretionary time (post-taxes, post-transfers, post-childcare-support) under alternative household rules, Australia Necessary time in

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

526 526 1 052

10.39 8.37 9.38

6.49 12.68 9.59

59.97 59.97 59.97

91.15 86.99 89.07

1 195 1 195 2 390

18.48 11.34 14.91

9.40 20.93 15.16

59.97 59.97 59.97

80.16 75.77 77.97

1 721 1 721 3 442

15.99 10.42 13.20

8.50 18.39 13.44

59.97 59.97 59.97

83.55 79.22 81.39

526 526 1 052

19.41 0.00 9.70

6.79 19.98 13.39

59.97 59.97 59.97

81.83 88.05 84.94

1 195 1 195 2 390

29.00 0.00 14.50

10.09 29.71 19.90

59.97 59.97 59.97

68.95 78.32 73.64

1 721 1 721 3 442

26.04 0.00 13.02

9.07 26.72 17.90

59.97 59.97 59.97

72.92 81.32 77.12

Most-efficient breadwinner without kids men 522 women 522 total 1 044

9.03 6.69 7.86

10.88 15.90 13.39

59.97 59.97 59.97

88.13 85.44 86.79

Conventional dual-earner without kids men women total with kids men women total total men women total Male breadwinner without kids men women total with kids men women total total men women total

356

Appendix 2

Table A2.8. (cont.) Necessary time in

with kids men women total total men women total

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

1 190 1 190 2 380

15.13 8.08 11.61

13.95 25.85 19.90

59.97 59.97 59.97

78.95 74.10 76.53

1 712 1 712 3 424

13.26 7.66 10.46

13.01 22.79 17.90

59.97 59.97 59.97

81.77 77.58 79.68

9.87 10.87 10.37

9.59 9.59 9.59

59.97 59.97 59.97

88.57 87.58 88.08

15.41 19.61 17.51

15.13 15.13 15.13

59.97 59.97 59.97

77.49 73.30 75.39

13.70 16.90 15.30

13.41 13.41 13.41

59.97 59.97 59.97

80.92 77.72 79.32

9.30 9.30 9.30

9.59 9.59 9.59

59.97 59.97 59.97

89.15 89.15 89.15

15.20 15.20 15.20

15.16 15.16 15.16

59.97 59.97 59.97

77.67 77.67 77.67

13.39 13.39 13.39

13.44 13.44 13.44

59.97 59.97 59.97

81.20 81.20 81.20

Equal monetary contribution without kids men 526 women 526 total 1 052 with kids men 1 186 women 1 186 total 2 372 total men 1 712 women 1 712 total 3 424 Equal temporal contribution without kids men 526 women 526 total 1 052 with kids men 1 195 women 1 195 total 2 390 total men 1 721 women 1 721 total 3 442

Data

357

Table A2.8. (cont.) Necessary time in

Atomistic divorce without kids men women total with kids men women total total men women total Self-reliant divorce without kids men women total with kids men women total total men women total Gendered divorce without kids men women total with kids men women total

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

526 526 1 052

14.49 15.92 15.20

13.56 13.56 13.56

59.97 59.97 59.97

79.99 78.56 79.28

1 195 1 195 2 390

15.04 19.51 17.27

13.56 13.56 13.56

59.97 59.97 59.97

79.44 74.97 77.20

1 721 1 721 3 442

14.87 18.40 16.64

13.56 13.56 13.56

59.97 59.97 59.97

79.61 76.07 77.84

526 526 1 052

14.49 15.92 15.20

13.56 13.56 13.56

59.97 59.97 59.97

79.99 78.56 79.28

1 194 1 194 2 388

22.35 29.87 26.11

26.19 26.19 26.19

59.97 59.97 59.97

59.49 51.97 55.73

1 720 1 720 3 440

19.93 25.58 22.75

22.30 22.30 22.30

59.97 59.97 59.97

65.81 60.16 62.98

526 526 1 052

14.49 15.92 15.20

13.56 13.56 13.56

59.97 59.97 59.97

79.99 78.56 79.28

1 195 1 194 2 389

15.04 29.87 22.45

13.56 26.19 19.87

59.97 59.97 59.97

79.44 51.97 65.71

358

Appendix 2

Table A2.8. (cont.) Necessary time in

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

1 721 1 720 3 441

14.87 25.58 20.22

13.56 22.30 17.93

59.97 59.97 59.97

79.61 60.16 69.89

526 526 1 052

14.49 15.92 15.20

13.56 13.56 13.56

59.97 59.97 59.97

79.99 78.56 79.28

1 194 1 195 2 389

21.37 19.51 20.44

13.56 26.19 19.88

59.97 59.97 59.97

73.10 62.33 67.71

1 720 1 721 3 441

19.25 18.40 18.83

13.56 22.30 17.93

59.97 59.97 59.97

75.23 67.33 71.27

14.59 16.04 15.31

13.56 13.56 13.56

59.97 59.97 59.97

79.88 78.44 79.16

18.93 21.40 20.17

17.96 21.79 19.88

59.97 59.97 59.97

71.13 64.84 67.99

17.60 19.75 18.68

16.61 19.26 17.94

59.97 59.97 59.97

73.82 69.02 71.42

Financially egalitarian divorce (money-egalitarian) without kids men 526 14.49 13.56 women 526 15.92 13.56 total 1 052 15.20 13.56

59.97 59.97 59.97

79.99 78.56 79.28

total men women total Male-breadwinner divorce without kids men women total with kids men women total total men women total

Most-efficient breadwinner divorce without kids men 522 women 522 total 1 044 with kids men 1 190 women 1 190 total 2 380 total men 1 712 women 1 712 total 3 424

Data

359

Table A2.8. (cont.) Necessary time in

with kids men women total total men women total

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

1 195 1 195 2 390

19.42 24.36 21.89

13.56 26.19 19.88

59.97 59.97 59.97

75.06 57.48 66.27

1 721 1 721 3 442

17.90 21.76 19.83

13.56 22.30 17.93

59.97 59.97 59.97

76.58 63.97 70.27

59.97 59.97 59.97

79.99 78.56 79.28

59.97 59.97 59.97

76.75 59.65 68.20

59.97 59.97 59.97

77.75 65.47 71.61

59.97 59.97 59.97

79.99 78.56 79.28

59.97 59.97 59.97

68.88 63.93 66.40

59.97 59.97 59.97

72.30 68.43 70.37

Financially egalitarian divorce (time-egalitarian) without kids men 526 14.49 13.56 women 526 15.92 13.56 total 1 052 15.20 13.56 with kids men 1 195 17.72 13.56 women 1 195 22.19 26.19 total 2 390 19.96 19.88 total men 1 721 16.73 13.56 women 1 721 20.26 22.30 total 3 442 18.49 17.93 Strictly egalitarian divorce (money-egalitarian) without kids men 526 14.49 13.56 women 526 15.92 13.56 total 1 052 15.20 13.56 with kids men 1 195 19.28 19.88 women 1 195 24.23 19.88 total 2 390 21.75 19.88 total men 1 721 17.80 17.93 women 1 721 21.67 17.93 total 3 442 19.74 17.93

360

Appendix 2

Table A2.8. (cont.) Necessary time in

N

Paid labour

Strictly egalitarian divorce (time-egalitarian) without kids men 526 14.49 women 526 15.92 total 1 052 15.20 with kids men 1 195 17.72 women 1 195 22.19 total 2 390 19.96 total men 1 721 16.72 women 1 721 20.26 total 3 442 18.49

Unpaid household labour

Personal care

Discret’y time

13.56 13.56 13.56

59.97 59.97 59.97

79.99 78.56 79.28

19.88 19.88 19.88

59.97 59.97 59.97

70.44 65.96 68.20

17.93 17.93 17.93

59.97 59.97 59.97

73.38 69.84 71.61

one adult

couples

singleearner

earners

dual-

employed

total

employed

stay-athome

employed

Without kids

0.00

0.00 0.00

8.13

7.35

8.91

53

105 7.86 208 12.68

155 14.76

363 13.48

total men

women

total

2.27

52 13.56

women

52 15.86

9 18.10

43 15.54

53

Unpaid

Unpaid

Necessary time in

Post-taxes, post-transfers, pre-child-care-support

Unpaid

Necessary time in

Post-taxes, post-transfers, post-child-care-support

13.13

13.13

14.32 13.13

20.66

7.86

7.61

13.24

6.78

20.91

15.26 21.72

9.29

11.49

7.07

56.47

56.47

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

56.47

56.47

56.47

84.93

83.65

89.36 85.73

88.61

90.11

88.06

80.20

89.22

90.62

96.27 89.82

94.12

92.69

95.55

19.67

21.54

10.36 18.50

2.97

17.88

20.90

23.72

20.49

0.00

0.00 0.00

10.79

9.77

11.80

13.13

13.13

14.32 13.13

20.66

7.86

7.61

13.24

6.78

20.91

15.26 21.72

9.29

11.49

7.07

56.47

56.47

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

56.47

56.47

56.47

78.74

76.86

86.86 79.90

87.91

85.79

83.03

74.57

84.26

90.62

96.27 89.82

91.46

90.27

92.66

19.67

21.54

10.36 18.50

2.97

17.88

20.90

23.72

20.49

0.00

0.00 0.00

10.79

9.77

11.80

13.13

13.13

14.32 13.13

20.66

7.86

7.61

13.24

6.78

20.91

15.26 21.72

9.29

11.49

7.07

56.47

56.47

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

56.47

56.47

56.47

78.74

76.86

86.86 79.90

87.91

85.79

83.03

74.57

84.26

90.62

96.27 89.82

91.46

90.27

92.66

Paid household Personal Discret’y Paid household Personal Discret’y Paid household Personal Discret’y labour labour care time labour labour care time labour labour care time

men

total

women

men

total

9 44

478

total

men women

239

239

women

men

N

Pre-taxes, pre-transfers, pre-child-care-support

Necessary time in

Table A2.9. Discretionary time, Germany

earners

dual-

With kids

total

employed

total

employed

stay-athome

Table A2.9. (cont.)

946 11.09

total

women total

426 10.04 852 12.88

426 15.74

447 10.47

women

men

499 11.57

893 11.60

total

men

403 11.50

0.00

women

53

490 11.68

total

0.00 0.00

17.99 13.35

8.68

11.87

13.23

10.82

11.45

12.40

10.78

20.91

15.26 21.72

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

care

83.50 85.30

87.11

88.57

87.84

89.15

88.48

87.64

89.08

90.62

96.27 89.82

time

12.92 16.60

20.31

15.70

14.86

16.36

16.44

16.31

16.52

0.00

0.00 0.00

17.99 13.35

8.68

11.87

13.23

10.82

11.45

12.40

10.78

20.91

15.26 21.72

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

care

80.62 81.58

82.54

83.96

83.45

84.35

83.65

82.82

84.23

90.62

96.27 89.82

time

12.63 16.24

19.88

15.70

14.86

16.36

16.44

16.31

16.52

0.00

0.00 0.00

17.99 13.35

8.68

11.87

13.23

10.82

11.45

12.40

10.78

20.91

15.26 21.72

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

care

80.91 81.94

82.97

83.96

83.45

84.35

83.65

82.82

84.23

90.62

96.27 89.82

time

household Personal Discret’y

Unpaid

Necessary time in

post-child-care-support

Post-taxes, post-transfers,

labour labour

household Personal Discret’y Paid

Unpaid labour labour

household Personal Discret’y Paid

Unpaid

labour labour

men

9 44

men women

N

Paid

Necessary time in

pre-child-care-support

pre-child-care-support Necessary time in

Post-taxes, post-transfers,

Pre-taxes, pre-transfers,

stay-at-

total

total

employed

home

employed

total

employed

home

stay-at-

one adult

couples

earner

single-

20

0.00

829 12.12

total

0.00

22.37

10.70 25.07 18.13

women 894 7.31 total 1 738 13.00

13.47

18.53

10.40

32.60

33.14

19.92 22.37

20.11

18.56

22.73

32.72

12.56

12.69

12.05 24.86

32.60

33.14

844 19.08

men

1 320 17.18

496 13.24

total

women

0.00

824 19.57

total

men

398

418

women

57 25.88 20 0.00

total men

7 15.25

50 27.32

women

men

418

women

1.93

411 22.48

411 24.44

men

total

0.00

0.00

391 23.67 20 38.87

418

398

men women

total

women

men

56.47

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

56.47

56.47

89.17

79.15 80.41

81.75

80.88

79.77

81.56

78.93

78.40

65.73 89.17

64.11

77.72

76.68

76.88

76.49

74.40

75.82 47.80

78.93

78.40

9.22 16.12

23.50

21.31

16.70

24.10

0.00

0.00

32.31 0.00

33.92

20.39

14.40

2.26

26.74

29.02

28.15 45.42

0.00

0.00

0.00

25.07 18.13

10.70

13.47

18.53

10.40

32.60

33.14

19.92 22.37

20.11

18.56

22.73

32.72

12.56

12.69

12.05 24.86

32.60

33.14

22.37

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

56.47

56.47

56.47

77.24 77.29

77.34

76.76

76.31

77.03

78.93

78.40

59.30 89.17

57.50

72.58

74.41

76.55

72.23

69.82

71.34 41.25

78.93

78.40

89.17

9.07 15.94

23.29

21.07

16.42

23.88

0.00

0.00

32.06 0.00

33.64

20.39

14.40

2.26

26.74

29.02

28.15 45.42

0.00

0.00

0.00

23.73 17.16

10.13

13.10

18.38

9.90

29.77

30.32

19.92 19.30

20.11

18.56

20.73

29.88

11.43

11.54

11.01 21.45

29.77

30.32

19.30

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

56.47

56.47

56.47

78.74 78.44

78.11

77.37

76.73

77.75

81.76

81.21

59.55 92.24

57.79

72.58

76.41

79.40

73.37

70.97

72.37 44.67

81.76

81.21

92.24

couples

earner

single-

earners

Total dual-

total

employed

home

stay-at-

employed

Table A2.9. (cont.)

8.96

471

934 11.51

women

total

1.98

463 21.21

men

29 32.71 463 23.21

women total

0.00

0.00

0.00

434 22.59

471

442

29

1 330 10.97

665

665 12.99

8.03

21.52

30.99

11.89

21.42 11.97

11.35

30.93

31.61

20.25

11.71

15.37

56.47

56.47

56.47

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47

care

78.50

78.56

78.44

57.41 76.35

77.60

80.61

79.92

91.28

88.85

87.20

90.52

time

13.82

2.36

25.47

38.98 27.86

27.13

0.00

0.00

0.00

14.26

11.65

16.88

8.03

21.52

30.99

11.89

21.42 11.97

11.35

30.93

31.61

20.25

11.71

15.37

56.47

56.47

56.47

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47

care

76.19

78.18

74.17

51.13 71.71

73.06

80.61

79.92

91.28

85.56

84.51

86.62

time

13.82

2.36

25.47

38.98 27.86

27.13

0.00

0.00

0.00

14.04

11.48

16.62

8.03

19.81

28.56

10.92

19.01 10.97

10.45

28.50

29.17

18.10

11.71

15.37

56.47

56.47

56.47

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47

care

77.90

80.62

75.15

53.53 72.70

73.95

83.03

82.36

93.44

85.78

84.68

86.88

time

household Personal Discret’y

Unpaid

Necessary time in

post-child-care-support

Post-taxes, post-transfers,

labour labour

household Personal Discret’y Paid

Unpaid labour labour

household Personal Discret’y Paid

Unpaid

labour labour

men

total

women

men

total

women

men

N

Paid

Necessary time in

pre-child-care-support

pre-child-care-support Necessary time in

Post-taxes, post-transfers,

Pre-taxes, pre-transfers,

stay-at-

total

total

employed

home

employed

one adult

0.00

31.61

19.91 15.14

women 1 341 8.69 total 2 684 12.08

12.38

15.09

10.60

30.93

10.76

2 213 14.19

899 12.26

1 314 15.47

0.00

13.58 20.25

14.12

13.20

1 343 15.20

men

total

women

men

442

471

420 14.32 29 0.00

total men

total

205 16.55

women

women

215 12.71

men

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

56.47

56.47

82.94 84.31

85.57

84.96

84.18

85.47

80.61

79.92

83.63 91.28

80.86

85.62

11.68 15.92

19.81

18.70

16.48

20.16

0.00

0.00

20.53 0.00

23.31

18.53

19.91 15.14

10.76

12.38

15.09

10.60

30.93

31.61

13.58 20.25

14.12

13.20

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

56.47

56.47

79.95 80.48

80.96

80.45

79.96

80.78

80.61

79.92

77.42 91.28

74.10

79.80

18.53

11.59 15.83

19.71

18.58

16.36

20.05

0.00

0.00

20.51 0.00

23.27

13.20

19.15 14.63

10.49

12.21

15.03

10.36

28.50

29.17

13.58 18.10

14.12

56.47

56.47 56.47

56.47

56.47

56.47

56.47

56.47

56.47

56.47 56.47

56.47

79.80

80.79 81.08

81.34

80.73

80.14

81.13

83.03

82.36

77.44 93.44

74.14

366

Appendix 2

Table A2.10. Spare time, Germany Actual time in

Without kids dual-earners employed

single-earner stay-atcouples home employed

total

one adult

employed

total

stay-athome employed

total

With kids dual-earners employed

single-earner stay-atcouples home

N

Paid labour

Unpaid household Personal labour care

Spare time

614 614 1 228 24 110 134 110 24 134 134 134 268 370 444 814 24 110 134 1 094 1 082 2 176 1 118 1 192 2 310

47.16 38.41 42.78 0.00 0.00 0.00 42.23 25.11 38.91 34.03 4.87 19.45 42.04 42.62 42.33 0.00 0.00 0.00 44.46 40.04 42.35 43.51 36.44 40.01

16.79 26.01 21.40 29.33 43.40 40.67 15.45 28.98 18.07 18.14 40.60 29.37 16.08 22.18 19.07 29.33 43.40 40.67 16.36 24.31 20.15 16.63 26.02 21.29

68.85 71.00 69.93 76.38 79.35 78.77 70.34 71.24 70.51 71.51 77.77 74.64 65.70 67.89 66.77 76.38 79.35 78.77 67.59 69.57 68.54 67.78 70.45 69.10

35.20 32.58 33.89 62.29 45.25 48.56 39.99 42.67 40.51 44.32 44.75 44.53 44.18 35.31 39.84 62.29 45.25 48.56 39.59 34.08 36.96 40.07 35.09 37.60

men 1 560 44.38 women 1 559 24.93 total 3 119 34.66 men 26 0.00 women 734 0.00 total 760 0.00

20.71 40.97 30.84 38.88 58.72 58.17

68.48 70.70 69.59 80.58 73.48 73.68

34.44 31.40 32.92 48.54 35.79 36.15

men women total men women total men women total men women total men women total men women total men women total men women total

Data

367

Table A2.10. (cont.) Actual time in

employed

total

one adult

employed

total

stay-athome employed

total

Total dual-earners employed

single-earner stay-atcouples home employed

total

one adult

employed

N

Paid labour

Unpaid household Personal labour care

Spare time

men women total men women total men women total men women total men women total men women total

734 26 760 760 760 1 520 34 290 324 26 734 760 2 328 1 875 4 203 2 354 2 609 4 963

41.54 19.95 40.94 40.38 0.56 20.47 43.92 33.32 34.61 0.00 0.00 0.00 43.26 26.01 36.14 42.78 16.71 29.23

22.18 41.83 22.74 22.65 58.25 40.45 29.19 35.37 34.62 38.88 58.72 58.17 21.40 40.21 29.17 21.60 46.83 34.71

68.97 78.78 69.24 69.29 73.63 71.46 60.57 68.67 67.69 80.58 73.48 73.68 68.56 70.55 69.38 68.70 71.60 70.21

35.31 27.45 35.09 35.68 35.56 35.62 34.32 30.64 31.08 48.54 35.79 36.15 34.78 31.23 33.31 34.93 32.86 33.86

men women total men women total men women total men women total men women total

2 174 2 173 4 347 50 844 894 844 50 894 894 894 1 788 404 734 1 138

45.63 31.02 38.33 0.00 0.00 0.00 41.68 23.39 40.48 38.95 1.53 20.24 42.10 40.95 41.48

18.94 34.21 26.58 32.50 55.75 54.23 20.88 33.25 21.69 21.64 54.28 37.96 16.46 24.55 20.76

68.65 70.84 69.74 77.78 74.62 74.83 69.23 73.75 69.53 69.79 74.56 72.18 65.55 68.03 66.87

34.78 31.93 33.36 57.72 37.63 38.95 36.22 37.61 36.31 37.62 37.63 37.63 43.90 34.47 38.88

368

Appendix 2

Table A2.10. (cont.) Actual time in

total

stay-athome employed

total

men women total men women total men women total

N

Paid labour

Unpaid household Personal labour care

Spare time

50 844 894 3 422 2 957 6 379 3 472 3 801 7 273

0.00 0.00 0.00 43.87 34.09 39.49 43.16 26.37 34.63

32.50 55.75 54.23 18.82 31.05 24.30 19.05 36.65 27.98

57.72 37.63 38.95 37.23 32.87 35.28 37.57 33.95 35.73

77.78 74.62 74.83 68.07 69.98 68.93 68.23 71.04 69.65

Data

369

Table A2.11. The state impact on discretionary time and the time-pressure illusion, Germany Impact of state taxesand-transfers on discret’y time Without kids dual-earners employed

single-earner stay-atcouples home employed

total

one adult

employed

total

stay-athome employed

total

With kids dual-earners

employed

Impact of state childcare support on discret’y time

Total state impact on Timediscret’y pressure time illusion

men women total men women total men women total men women total men women total men women total men women total men women total

2.89 2.43 2.66 0.00 0.00 0.00 4.95 5.63 5.04 4.32 0.71 2.50 5.83 6.78 6.19 0.00 0.00 0.00 4.84 4.82 4.83 4.79 4.39 4.62

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

2.89 2.43 2.66 0.00 0.00 0.00 4.95 5.63 5.04 4.32 0.71 2.50 5.83 6.78 6.19 0.00 0.00 0.00 4.84 4.82 4.83 4.79 4.39 4.62

57.46 57.69 57.57 33.98 44.56 42.06 44.28 31.90 42.52 41.48 43.15 42.33 35.73 41.55 38.90 33.98 44.56 42.06 44.65 48.74 46.69 44.28 48.36 46.36

men women total

4.57 2.87 3.72

0.43 0.29 0.36

4.14 2.59 3.36

48.54 49.51 49.02

370

Appendix 2

Table A2.11. (cont.) Impact of state taxesand-transfers on discret’y time single-earner stay-atcouples home employed

total

one adult

employed

total

stay-athome employed

total

Total dual-earners

employed

single-earner stay-atcouples home employed

Impact of state childcare support on discret’y time

Total state impact on Timediscret’y pressure time illusion

men women total men women total men women total men women total men women total men women total men women total

0.00 0.00 0.00 4.48 6.55 4.58 4.25 0.33 2.27 5.14 6.60 6.43 0.00 0.00 0.00 4.53 3.46 4.12 4.42 1.91 3.12

3.07 2.82 2.83 1.04 3.42 1.16 1.14 2.85 2.00 0.00 0.29 0.25 3.07 2.82 2.83 0.72 0.42 0.61 0.78 1.49 1.15

3.07 2.82 2.83 3.44 3.13 3.43 3.11 2.52 0.27 5.14 6.32 6.18 3.07 2.82 2.83 3.81 3.03 3.52 3.64 0.41 1.97

43.70 45.42 45.61 37.07 17.22 35.89 37.69 43.84 40.79 38.26 27.15 28.47 43.70 45.42 45.61 42.98 45.50 44.05 43.18 45.88 44.58

men women total men women total men women total

3.89 2.69 3.29 0.00 0.00 0.00 4.54 6.27 4.65

0.26 0.17 0.21 2.16 2.44 2.42 0.90 2.40 0.99

3.63 2.52 3.08 2.16 2.44 2.42 3.64 3.87 3.66

52.10 52.75 52.42 35.72 44.73 44.08 37.74 15.92 36.39

Data

371

Table A2.11. (cont.) Impact of state taxesand-transfers on discret’y time total

one adult

employed

total

stay-athome employed

total

men women total men women total men women total men women total men women total

4.26 0.38 2.31 5.82 6.76 6.21 0.00 0.00 0.00 4.69 4.22 4.50 4.61 2.99 3.84

Impact of state childcare support on discret’y time 0.98 2.44 1.71 0.00 0.04 0.02 2.16 2.44 2.42 0.35 0.19 0.28 0.38 0.84 0.60

Total state impact on Timediscret’y pressure time illusion 3.29 2.06 0.59 5.82 6.72 6.19 2.16 2.44 2.42 4.35 4.03 4.22 4.23 2.15 3.24

37.52 42.99 40.28 35.90 39.67 38.56 35.72 44.73 44.08 43.89 47.27 45.45 43.77 46.84 45.34

372

Appendix 2

Table A2.12. Discretionary time (post-taxes, post-transfers, post-childcare-support) under alternative household rules, Germany Necessary time in Paid labour

Unpaid household Personal labour care

Discret’y time

11.80 9.77 10.79

7.07 11.49 9.29

56.47 56.47 56.47

92.66 90.27 91.46

19.88 12.63 16.24

8.68 17.99 13.35

56.47 56.47 56.47

82.97 80.91 81.94

16.62 11.48 14.04

8.03 15.37 11.71

56.47 56.47 56.47

86.88 84.68 85.78

20.35 0.00 10.15

6.78 21.72 14.27

56.47 56.47 56.47

84.40 89.82 87.12

29.16 0.00 14.53

10.30 28.37 19.37

56.47 56.47 56.47

72.07 83.16 77.64

25.60 0.00 12.76

8.88 25.69 17.31

56.47 56.47 56.47

77.05 85.84 81.46

Most-efficient breadwinner without kids men 238 12.29 women 238 5.96 total 476 9.12

7.95 20.46 14.22

56.47 56.47 56.47

91.29 85.11 88.19

N Conventional dual-earner without kids men 239 women 239 total 478 with kids men 426 women 426 total 852 total men 665 women 665 total 1 330 Male breadwinner without kids men 239 women 239 total 478 with kids men 426 women 426 total 852 total men 665 women 665 total 1 330

Data

373

Table A2.12. (cont.) Necessary time in Paid labour

Unpaid household Personal labour care

Discret’y time

425 425 850

18.50 6.86 12.66

9.89 28.80 19.38

56.47 56.47 56.47

83.14 75.88 79.50

663 663 1 326

16.02 6.50 11.24

9.11 25.46 17.31

56.47 56.47 56.47

86.41 79.57 82.98

9.28 9.28 9.28

56.47 56.47 56.47

91.54 88.61 90.07

13.34 13.33 13.34

56.47 56.47 56.47

82.30 75.55 78.91

11.70 11.70 11.70

56.47 56.47 56.47

86.02 80.81 83.41

9.28 9.28 9.28

56.47 56.47 56.47

91.36 91.36 91.36

13.34 13.33 13.34

56.47 56.47 56.47

81.41 81.42 81.42

N with kids men women total total men women total

Equal monetary contribution without kids men 239 10.71 women 239 13.64 total 478 12.18 with kids men 426 15.90 women 426 22.65 total 852 19.29 total men 665 13.81 women 665 19.02 total 1 330 16.42 Equal temporal contribution without kids men 239 10.89 women 239 10.89 total 478 10.89 with kids men 426 16.79 women 426 16.78 total 852 16.78

374

Appendix 2

Table A2.12. (cont.) Necessary time in

total men women total

N

Paid labour

Unpaid household Personal labour care

Discret’y time

665 665 1 330

14.41 14.41 14.41

11.70 11.70 11.70

56.47 56.47 56.47

85.42 85.43 85.43

16.48 21.45 18.97

13.13 13.13 13.13

56.47 56.47 56.47

81.93 76.96 79.44

17.84 25.35 21.61

13.13 13.13 13.13

56.47 56.47 56.47

80.57 73.05 76.80

17.30 23.79 20.56

13.13 13.13 13.13

56.47 56.47 56.47

81.11 74.62 77.85

16.48 21.45 18.97

13.13 13.13 13.13

56.47 56.47 56.47

81.93 76.96 79.44

26.77 38.08 32.44

22.64 22.63 22.64

56.47 56.47 56.47

62.12 50.83 56.45

22.65 31.43 27.06

18.83 18.83 18.83

56.47 56.47 56.47

70.05 61.27 65.65

16.48 21.45 18.97

13.13 13.13 13.13

56.47 56.47 56.47

81.93 76.96 79.44

Atomistic divorce without kids men 238 women 238 total 476 with kids men 426 women 426 total 852 total men 664 women 664 total 1 328 Self-reliant divorce without kids men 238 women 238 total 476 with kids men 426 women 426 total 852 total men 664 women 664 total 1 328 Gendered divorce without kids men 238 women 238 total 476

Data

375

Table A2.12. (cont.) Necessary time in Paid labour

Unpaid household Personal labour care

Discret’y time

426 426 852

17.84 38.08 28.00

13.13 22.63 17.89

56.47 56.47 56.47

80.57 50.83 65.64

664 664 1 328

17.30 31.43 24.39

13.13 18.83 15.99

56.47 56.47 56.47

81.11 61.27 71.16

16.48 21.45 18.97

13.13 13.13 13.13

56.47 56.47 56.47

81.93 76.96 79.44

26.74 25.35 26.04

13.13 22.63 17.89

56.47 56.47 56.47

71.67 63.55 67.60

22.63 23.79 23.21

13.13 18.83 15.99

56.47 56.47 56.47

75.77 68.91 72.33

13.13 13.13 13.13

56.47 56.47 56.47

81.93 76.96 79.44

15.49 20.27 17.89

56.47 56.47 56.47

72.46 63.80 68.12

14.54 17.41 15.98

56.47 56.47 56.47

76.25 69.07 72.65

N with kids men women total total men women total

Male-breadwinner divorce without kids men 238 women 238 total 476 with kids men 426 women 426 total 852 total men 664 women 664 total 1 328

Most-efficient breadwinner divorce without kids men 238 16.48 women 238 21.45 total 476 18.97 with kids men 425 23.58 women 425 27.45 total 850 25.52 total men 663 20.74 women 663 25.05 total 1 326 22.90

376

Appendix 2

Table A2.12. (cont.) Necessary time in

N

Paid labour

Unpaid household Personal labour care

Financially egalitarian divorce (money-egalitarian) without kids men 238 16.48 13.13 women 238 21.45 13.13 total 476 18.97 13.13 with kids men 426 22.26 13.13 women 426 31.72 22.63 total 852 27.01 17.89 total men 664 19.95 13.13 women 664 27.61 18.83 total 1 328 23.80 15.99 Financially egalitarian divorce (time-egalitarian) without kids men 238 16.48 13.13 women 238 21.45 13.13 total 476 18.97 13.13 with kids men 426 22.57 13.13 women 426 30.07 22.63 total 852 26.34 17.89 total men 664 20.14 13.13 women 664 26.62 18.83 total 1 328 23.39 15.99 Strictly egalitarian divorce (money-egalitarian) without kids men 238 16.48 13.13 women 238 21.45 13.13 total 476 18.97 13.13

Discret’y time

56.47 56.47 56.47

81.93 76.96 79.44

56.47 56.47 56.47

76.14 57.18 66.63

56.47 56.47 56.47

78.46 65.09 71.75

56.47 56.47 56.47

81.93 76.96 79.44

56.47 56.47 56.47

75.83 58.84 67.30

56.47 56.47 56.47

78.27 66.08 72.16

56.47 56.47 56.47

81.93 76.96 79.44

Data

377

Table A2.12. (cont.) Necessary time in Paid labour

Unpaid household Personal labour care

Discret’y time

426 426 852

22.27 31.72 27.01

17.88 17.88 17.88

56.47 56.47 56.47

71.38 61.94 66.64

664 664 1 328

19.95 27.61 23.79

15.98 15.98 15.98

56.47 56.47 56.47

75.60 67.94 71.76

56.47 56.47 56.47

81.93 76.96 79.44

56.47 56.47 56.47

71.08 63.59 67.32

56.47 56.47 56.47

75.42 68.93 72.16

N with kids men women total total men women total

Strictly egalitarian divorce (time-egalitarian) without kids men 238 16.48 13.13 women 238 21.45 13.13 total 476 18.97 13.13 with kids men 426 22.57 17.88 women 426 30.07 17.88 total 852 26.34 17.88 total men 664 20.14 15.98 women 664 26.62 15.98 total 1 328 23.39 15.98

one adult

employed

total

employed

stay-at-home

single-earner

couples

employed

dual-earners

Without kids

total

353

316

669

women

total

66 132

men

women total

66

32

66

women

men

34

men

34

66

total

388 32

total men

women

194

194

men

women

N

Table A2.13. Discretionary time, France

17.04

18.34

15.90

11.87 11.43

10.99

22.86

24.14

21.61

0.00

0.00

10.29 0.00

9.41

11.16

6.19

11.79

11.79

11.79

16.80 13.47

10.14

9.38

12.70

6.17

17.56

20.77

8.32 14.25

10.45

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

care

Personal

Discret’y

Paid

77.57

76.27

78.71

77.73 81.50

85.27

74.16

69.56

78.62

88.84

85.63

87.79 92.15

86.54

89.04

time

21.12

22.46

19.94

14.08 13.56

13.05

27.13

28.64

25.66

0.00

0.00

12.19 0.00

11.13

13.26

11.79

11.79

11.79

16.80 13.47

10.14

9.38

12.70

6.17

17.56

20.77

8.32 14.25

10.45

6.19

labour labour

household

labour labour

Paid

Unpaid

household

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

care

Personal

Necessary time in

73.49

72.15

74.67

75.52 79.37

83.21

69.90

65.07

74.57

88.84

85.63

85.89 92.15

84.82

86.95

time

Discret’y

Post-taxes, post-transfers, pre-child-care-support

Unpaid

Necessary time in

Pre-taxes, pre-transfers, pre-child-care-support

household

21.12

22.46

19.94

14.08 13.56

13.05

27.13

28.64

25.66

0.00

0.00

12.19 0.00

11.13

13.26

11.79

11.79

11.79

16.80 13.47

10.14

9.38

12.70

6.17

17.56

20.77

8.32 14.25

10.45

6.19

labour labour

Paid

Unpaid

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

care

73.49

72.15

74.67

75.52 79.37

83.21

69.90

65.07

74.57

88.84

85.63

85.89 92.15

84.82

86.95

time

Personal Discret’y

Necessary time in

Post-taxes, post-transfers, post-child-care-support

one adult

employed

total

employed

stay-at-home

single-earner

couples

employed

total

employed

stay-at-home

dual-earners

With kids

total

66

total

576

139

146

women

total

7

433

men

total

total

217 216

102

216

women

men women

114

217

total

men

114

103

women

men

30.52

31.26

15.38

17.35

18.36 16.35

34.79

34.64

34.93

0.00

0.00

0.00

17.57

14.04 15.80

655

men

14.45

14.89

14.06

15.26

14.80 15.77

0.00

0.00

0.00

women 655 total 1 310

1 189

women

total

613

1 123

men

total

581 542

34

women

men women

32

men

18.88

18.85

19.42

20.16

13.76 26.59

14.56

20.77

9.01

25.73

31.79

19.02

16.87 12.70

8.54

10.94

11.93

10.03

10.57

9.81 11.40

17.56

20.77

14.25

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

92.15

57.00

56.29

71.60

68.89

74.29 63.46

57.05

50.99

62.47

80.67

74.61

87.38

75.50 77.89

80.29

81.00

79.59

82.31

80.57

81.79 79.23

88.84

85.63

0.00

27.06

27.72

13.54

16.19

17.14 15.25

32.47

32.31

32.61

0.00

0.00

0.00

15.65 17.64

19.62

17.69

18.05

17.35

18.67

18.26 19.12

0.00

0.00

14.25

18.88

18.85

19.42

20.16

13.76 26.59

14.56

20.77

9.01

25.73

31.79

19.02

16.87 12.70

8.54

10.94

11.93

10.03

10.57

9.81 11.40

17.56

20.77

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

92.15

60.46

59.83

73.44

70.05

75.51 64.56

59.38

53.32

64.79

80.67

74.61

87.38

73.88 76.06

78.24

77.77

76.42

79.02

77.16

78.33 75.87

88.84

85.63

0.00

25.81

26.41

13.44

16.19

17.14 15.25

32.47

32.31

32.61

0.00

0.00

0.00

14.84 16.75

18.66

17.69

18.05

17.35

18.67

18.26 19.12

0.00

0.00

14.25

18.88

18.85

19.42

16.96

11.63 22.31

12.30

17.68

7.49

21.59

26.46

16.21

16.87 12.70

8.54

10.94

11.93

10.03

10.57

9.81 11.40

17.56

20.77

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

92.15

61.71

61.14

73.54

73.25

77.64 68.84

61.63

56.41

66.30

84.81

79.94

90.19

74.69 76.95

79.20

77.77

76.42

79.02

77.16

78.33 75.87

88.84

85.63

stay-at-home

single-earner

couples

employed

total

employed

stay-at-home

dual-earners

Total

total

Table A2.13. (cont.)

148

283

total

135

1 698

women

men

total

849

849

men

0.00

0.00

0.00

14.44

12.90

15.99

17.13 17.41

women 1 010 total 1 889

19.71

19.32

20.16

0.00

0.00

0.00

17.75

women

hold

23.65

29.05

17.76

11.62

15.28

7.96

19.26 14.96

9.95

13.54

17.65

8.71

25.73

31.79

19.02

61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60

61.60

care

Personal

Discret’y

Paid

82.75

77.35

88.64

80.34

78.22

82.45

70.01 74.03

78.70

73.15

69.43

77.52

80.67

74.61

87.38

time

0.00

0.00

0.00

16.29

14.54

18.05

17.38 18.10

18.94

20.49

19.61

21.52

0.00

0.00

0.00

23.65

29.05

17.76

11.62

15.28

7.96

19.26 14.96

9.95

13.54

17.65

8.71

25.73

31.79

19.02

labour labour

hold

house-

houselabour labour

879

men

1 672

896

total

women

217

total

776

114

women

men

103

men

N

Paid

Unpaid

Unpaid

61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60

61.60

care

Personal

Necessary time in

82.75

77.35

88.64

78.49

76.58

80.39

69.76 73.34

77.51

72.37

69.14

76.17

80.67

74.61

87.38

time

Discret’y

pre-child-care-support

pre-child-care-support Necessary time in

Post-taxes, post-transfers,

Pre-taxes, pre-transfers,

hold

house-

0.00

0.00

0.00

15.63

13.92

17.33

16.67 17.39

18.23

19.68

18.81

20.71

0.00

0.00

0.00

20.56

25.04

15.69

11.62

15.28

7.96

18.34 14.22

9.41

13.24

17.30

8.49

21.59

26.46

16.21

labour labour

Paid

Unpaid

61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60

61.60

care

85.84

81.36

90.71

79.15

77.19

81.11

71.39 74.79

78.76

73.47

70.30

77.20

84.81

79.94

90.19

time

Personal Discret’y

Necessary time in

post-child-care-support

Post-taxes, post-transfers,

stay-at-home

total

total

employed

employed

one adult

total

employed

total

16.22

16.14

total

3 078

16.05

1 492

women 1 586

men

17.84

17.72

women 1 438

2 795

0.00 17.59

0.00

0.00

19.17

21.92

15.89

15.84

16.47 15.20

31.73

31.87

31.61

283 1 357

total men

148

women

total

135

455

815

women

men

360

565

men

total

282

total

283 282

134

women

men women

148

men

8.30

13.23

16.30

9.99

12.21

15.04

23.65 9.24

29.05

17.76

12.91

13.74

11.91

18.45

12.83 24.08

13.23

18.64

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

77.03

73.87

80.36

76.47

73.53

82.75 79.57

77.35

88.64

74.33

70.74

78.60

72.12

77.09 67.12

61.44

55.90

66.49

17.92

17.65

18.21

19.67

19.41

0.00 19.95

0.00

0.00

22.05

23.92

19.84

15.52

16.09 14.95

31.10

31.34

30.88

8.30

13.23

16.30

9.99

12.21

15.04

23.65 9.24

29.05

17.76

12.91

13.74

11.91

18.45

12.83 24.08

13.23

18.64

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

75.25

72.44

78.21

74.52

71.96

82.75 77.20

77.35

88.64

71.44

68.74

74.65

72.43

77.48 67.37

62.07

56.42

67.22

17.52

17.23

17.83

19.23

18.94

0.00 19.54

0.00

0.00

21.86

23.55

19.84

15.52

16.09 14.95

31.10

31.34

30.88

7.16

12.80

15.75

9.70

12.05

14.83

20.56 9.12

25.04

15.69

12.91

13.74

11.91

16.07

11.25 20.90

11.55

16.36

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

61.60

61.60

61.60

61.60 61.60

61.60

61.60

61.60

76.08

73.42

78.88

75.12

72.63

85.84 77.74

81.36

90.71

71.64

69.10

74.65

74.81

79.06 70.55

63.75

58.70

68.36

382

Appendix 2

Table A2.14. Spare time, France Actual time in

N Without kids dual-earners

employed

single-earner couples

stay-athome employed

total

one adult

employed

total

stay-athome employed

total

With kids dual-earners

employed

single-earner couples

stay-athome

Paid labour

Unpaid household labour

Personal care

Spare time

men women total men women total men women total men women total men women total men women total men women total men women total

390 390 780 19 94 113 94 19 113 113 113 226 351 316 667 19 94 113 835 725 1 560 854 819 1 673

46.48 40.33 43.40 0.00 0.00 0.00 45.62 27.19 42.14 37.00 5.14 21.07 42.28 42.26 42.27 0.00 0.00 0.00 44.42 40.83 42.80 43.41 36.41 40.09

16.60 25.55 21.07 29.98 42.09 39.80 14.45 27.22 16.86 17.38 39.28 28.33 18.10 22.27 19.94 29.98 42.09 39.80 17.09 24.12 20.27 17.39 26.06 21.50

74.42 77.36 75.89 79.03 85.36 84.16 76.52 72.96 75.85 76.99 83.02 80.01 71.16 72.44 71.73 79.03 85.36 84.16 73.10 75.02 73.97 73.24 76.13 74.61

30.50 24.77 27.63 58.99 40.55 44.03 31.41 40.63 33.15 36.62 40.56 38.59 36.45 31.03 34.06 58.99 40.55 44.03 33.39 28.04 30.97 33.97 29.39 31.80

men women total men women total

950 950 1 900 37 397 434

43.18 31.29 37.24 0.00 0.00 0.00

20.45 35.08 27.77 33.09 54.15 52.03

74.95 77.37 76.16 85.26 81.13 81.55

29.41 24.26 26.83 49.65 32.72 34.43

Data

383

Table A2.14. (cont.) Actual time in

N employed

total

one adult

employed

total

stay-athome employed

total

Total dual-earners

employed

single-earner couples

stay-athome employed

total

one adult

employed

Paid labour

Unpaid household labour

Personal care

Spare time

men women total men women total men women total men women total men women total men women total

397 37 434 434 434 868 18 116 134 37 397 434 1 365 1 103 2 468 1 402 1 500 2 902

42.50 22.80 40.52 38.23 2.29 20.26 44.91 33.26 34.86 0.00 0.00 0.00 43.01 31.17 37.69 41.63 22.88 31.92

17.20 34.75 18.96 18.80 52.19 35.50 25.78 30.59 29.93 33.09 54.15 52.03 19.57 34.56 26.30 20.00 39.77 30.24

76.22 81.94 76.79 77.13 81.21 79.17 74.49 76.36 76.11 85.26 81.13 81.55 75.32 77.44 76.27 75.64 78.42 77.08

32.08 28.51 31.72 33.85 32.30 33.07 22.82 27.79 27.11 49.65 32.72 34.43 30.10 24.83 27.73 30.73 26.93 28.76

men women total men women total men women total men women total men women total

1 340 1 340 2 680 56 491 547 491 56 547 547 547 1 094 369 432 801

44.24 34.18 39.21 0.00 0.00 0.00 43.14 24.32 40.88 37.96 2.92 20.44 42.39 40.05 41.21

19.22 32.03 25.63 32.01 51.69 49.33 16.64 32.13 18.50 18.48 49.34 33.91 18.41 24.31 21.37

74.78 77.36 76.07 83.09 81.99 82.13 76.28 78.82 76.58 77.10 81.61 79.36 71.29 73.41 72.35

29.76 24.43 27.09 52.90 34.32 36.55 31.94 32.72 32.04 34.46 34.13 34.29 35.91 30.23 33.06

384

Appendix 2

Table A2.14. (cont.) Actual time in

N total

stay-athome employed

total

men women total men women total men women total

56 491 547 2 200 1 828 4 028 2 256 2 319 4 575

Paid labour 0.00 0.00 0.00 43.62 35.35 39.89 42.39 28.11 35.25

Unpaid household labour

Personal care

Spare time

32.01 51.69 49.33 18.50 30.03 23.70 18.88 34.47 26.68

83.09 81.99 82.13 74.37 76.39 75.28 74.61 77.54 76.08

52.90 34.32 36.55 31.51 26.22 29.13 32.11 27.88 29.99

Data

385

Table A2.15. The state impact on discretionary time and the time-pressure illusion, France Impact of state taxes-andtransfers on discret’y time Without kids dual-earners employed

single-earner stay-atcouples home employed

total

one adult

employed

total

stay-athome employed

total

With kids dual-earners

employed

single-earner stay-atcouples home

Impact of state childcare support on discret’y time

Total state impact on discret’y time

Timepressure illusion

men women total men women total men women total men women total men women total men women total men women total men women total

2.10 1.72 1.91 0.00 0.00 0.00 4.05 4.50 4.27 2.06 2.21 2.13 4.04 4.12 4.08 0.00 0.00 0.00 3.46 3.36 3.41 3.29 3.17 3.23

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

2.10 1.72 1.91 0.00 0.00 0.00 4.05 4.50 4.27 2.06 2.21 2.13 4.04 4.12 4.08 0.00 0.00 0.00 3.46 3.36 3.41 3.29 3.17 3.23

56.45 60.05 58.25 33.16 45.08 44.80 43.16 24.43 36.74 46.59 34.96 40.77 38.22 41.12 39.44 33.16 45.08 44.80 44.94 47.83 46.19 45.05 47.03 45.98

men women total men women total

2.05 1.61 1.83 0.00 0.00 0.00

0.96 0.81 0.88 2.81 5.34 4.14

1.09 0.80 0.95 2.81 5.34 4.14

49.79 50.43 50.11 40.54 47.22 50.38

386

Appendix 2

Table A2.15. (cont.) Impact of state taxes-andtransfers on discret’y time employed

total

one adult

employed

total

stay-athome employed

total

Total dual-earners

employed

single-earner stay-atcouples home employed

total

Impact of state childcare support on discret’y time

Total state impact on discret’y time

Timepressure illusion

men women total men women total men women total men women total men women total men women total

2.32 2.33 2.33 1.22 1.10 1.16 1.84 3.54 3.46 0.00 0.00 0.00 1.35 0.29 0.78 1.19 0.25 0.69

1.51 3.09 2.26 2.13 4.28 3.20 0.10 1.31 1.25 2.81 5.34 4.14 1.03 1.16 1.10 1.24 1.63 1.45

3.83 5.43 4.58 3.35 5.38 4.36 1.94 4.85 4.71 2.81 5.34 4.14 0.32 0.87 0.32 0.05 1.38 0.77

34.22 27.90 29.91 43.79 36.54 40.17 50.72 33.35 34.61 40.54 47.22 50.38 47.10 45.47 45.74 48.03 44.47 46.03

men women total men women total men women total men women total

2.06 1.64 1.85 0.00 0.00 0.00 0.73 0.53 0.64 0.38 0.25 0.32

0.72 0.61 0.67 2.07 4.01 3.08 1.14 2.28 1.68 1.58 3.18 2.38

1.34 1.03 1.18 2.07 4.01 3.08 1.87 2.80 2.31 1.97 3.43 2.70

51.35 52.77 52.06 37.81 47.04 49.29 36.42 25.98 31.71 44.60 36.43 40.52

Data

387

Table A2.15. (cont.) Impact of state taxes-andtransfers on discret’y time one adult

employed

total

stay-athome employed

total

men women total men women total men women total men women total

3.95 2.00 2.89 0.00 0.00 0.00 2.36 1.57 1.96 2.16 1.43 1.78

Impact of state childcare support on discret’y time 0.00 0.36 0.20 2.07 4.01 3.08 0.54 0.67 0.61 0.67 0.97 0.83

Total state impact on discret’y time 3.95 1.64 2.69 2.07 4.01 3.08 1.83 0.90 1.35 1.49 0.45 0.96

Timepressure illusion 38.74 38.87 38.57 37.81 47.04 49.29 46.23 46.41 45.99 46.76 45.54 46.08

388

Appendix 2

Table A2.16. Discretionary time (post-taxes, post-transfers, post-childcare-support) under alternative household rules, France Necessary time in Unpaid household Personal labour care

Discret’y time

13.26 11.13 12.19

6.19 10.45 8.32

61.60 61.60 61.60

86.95 84.83 85.89

18.66 14.84 16.75

8.54 16.87 12.70

61.60 61.60 61.60

79.20 74.69 76.95

17.33 13.92 15.63

7.96 15.28 11.62

61.60 61.60 61.60

81.11 77.19 79.15

194 194 388

23.05 0.00 11.52

6.17 20.77 13.47

61.60 61.60 61.60

77.18 85.63 81.41

655 655 1 310

26.94 0.00 13.47

7.45 26.31 16.88

61.60 61.60 61.60

72.00 80.09 76.05

849 849 1 698

25.98 0.00 12.99

7.14 24.94 16.04

61.60 61.60 61.60

73.28 81.46 77.37

Most-efficient breadwinner without kids men 193 women 193 total 386

14.14 6.38 10.26

8.61 18.34 13.47

61.60 61.60 61.60

83.65 81.68 82.67

Paid labour

N Conventional dual-earner without kids men 194 women 194 total 388 with kids men 655 women 655 total 1 310 total men 849 women 849 total 1 698 Male breadwinner without kids men women total with kids men women total total men women total

Data

389

Table A2.16. (cont.) Necessary time in

N with kids men women total total men women total

Paid labour

Unpaid household Personal labour care

Discret’y time

655 655 1 310

14.47 8.49 11.48

12.40 21.37 16.88

61.60 61.60 61.60

79.53 76.55 78.04

848 848 1 696

14.39 7.97 11.18

11.47 20.62 16.04

61.60 61.60 61.60

80.54 77.81 79.18

11.97 15.56 13.76

8.34 8.34 8.34

61.60 61.60 61.60

86.09 82.51 84.30

17.27 20.72 18.99

12.72 12.72 12.72

61.60 61.60 61.60

76.42 72.96 74.69

15.96 19.45 17.71

11.64 11.64 11.64

61.60 61.60 61.60

78.80 75.31 77.06

12.22 12.22 12.22

8.34 8.34 8.34

61.60 61.60 61.60

85.84 85.84 85.84

16.82 16.82 16.82

12.71 12.71 12.71

61.60 61.60 61.60

76.87 76.87 76.87

15.68 15.68 15.68

11.63 11.63 11.63

61.60 61.60 61.60

79.08 79.08 79.08

Equal monetary contribution without kids men 193 women 193 total 386 with kids men 654 women 654 total 1 308 total men 847 women 847 total 1 694 Equal temporal contribution without kids men 194 women 194 total 388 with kids men 655 women 655 total 1 310 total men 849 women 849 total 1 698

390

Appendix 2

Table A2.16. (cont.) Necessary time in Paid labour

N Atomistic divorce without kids men women total with kids men women total total men women total Self-reliant divorce without kids men women total with kids men women total total men women total Gendered divorce without kids men women total with kids men women total

Unpaid household Personal labour care

Discret’y time

194 194 388

18.08 23.81 20.95

11.79 11.79 11.79

61.60 61.60 61.60

76.54 70.80 73.67

655 655 1 310

19.77 23.39 21.58

11.79 11.79 11.79

61.60 61.60 61.60

74.84 71.22 73.03

849 849 1 698

19.35 23.49 21.42

11.79 11.79 11.79

61.60 61.60 61.60

75.26 71.12 73.19

194 194 388

18.08 23.81 20.95

11.79 11.79 11.79

61.60 61.60 61.60

76.54 70.80 73.67

655 655 1 310

23.76 28.24 26.00

21.73 21.73 21.73

61.60 61.60 61.60

60.91 56.43 58.67

849 849 1 698

22.36 27.15 24.75

19.28 19.28 19.28

61.60 61.60 61.60

64.77 59.97 62.37

194 194 388

18.08 23.81 20.95

11.79 11.79 11.79

61.60 61.60 61.60

76.54 70.80 73.67

655 655 1 310

19.77 28.24 24.00

11.79 21.73 16.76

61.60 61.60 61.60

74.84 56.43 65.64

Data

391

Table A2.16. (cont.) Necessary time in Unpaid household Personal labour care

Discret’y time

19.35 27.15 23.25

11.79 19.28 15.53

61.60 61.60 61.60

75.26 59.97 67.62

18.08 23.81 20.95

11.79 11.79 11.79

61.60 61.60 61.60

76.54 70.80 73.67

23.61 23.39 23.50

11.79 21.73 16.76

61.60 61.60 61.60

71.01 61.28 66.14

22.24 23.49 22.87

11.79 19.28 15.53

61.60 61.60 61.60

72.37 63.63 68.00

11.79 11.79 11.79

61.60 61.60 61.60

76.55 70.78 73.67

15.58 17.93 16.76

61.60 61.60 61.60

68.79 63.63 66.21

14.65 16.42 15.54

61.60 61.60 61.60

70.69 65.39 68.04

Financially egalitarian divorce (money-egalitarian) without kids men 194 18.08 11.79 women 194 23.81 11.79 total 388 20.95 11.79

61.60 61.60 61.60

76.54 70.80 73.67

N total men women total

849 849 1 698

Male-breadwinner divorce without kids men 194 women 194 total 388 with kids men 655 women 655 total 1 310 total men 849 women 849 total 1 698

Paid labour

Most-efficient breadwinner divorce without kids men 193 18.06 women 193 23.83 total 386 20.95 with kids men 655 22.03 women 655 24.83 total 1 310 23.43 total men 848 21.05 women 848 24.59 total 1 696 22.82

392

Appendix 2

Table A2.16. (cont.) Necessary time in

N with kids men women total total men women total

Paid labour

Unpaid household Personal labour care

Discret’y time

655 655 1 310

21.77 25.79 23.78

11.79 21.73 16.76

61.60 61.60 61.60

72.84 58.88 65.86

849 849 1 698

20.86 25.31 23.08

11.79 19.28 15.53

61.60 61.60 61.60

73.75 61.82 67.78

61.60 61.60 61.60

76.54 70.80 73.67

61.60 61.60 61.60

73.00 59.49 66.25

61.60 61.60 61.60

73.88 62.28 68.08

61.60 61.60 61.60

76.54 70.80 73.67

61.60 61.60 61.60

67.88 63.63 65.75

61.60 61.60 61.60

70.01 65.40 67.71

Financially egalitarian divorce (time-egalitarian) without kids men 194 18.08 11.79 women 194 23.81 11.79 total 388 20.95 11.79 with kids men 655 21.61 11.79 women 655 25.18 21.73 total 1 310 23.40 16.76 total men 849 20.74 11.79 women 849 24.85 19.28 total 1 698 22.79 15.53 Strictly egalitarian divorce (money-egalitarian) without kids men 194 18.08 11.79 women 194 23.81 11.79 total 388 20.95 11.79 with kids men 655 21.76 16.76 women 655 26.01 16.76 total 1 310 23.89 16.76 total men 849 20.85 15.53 women 849 25.47 15.53 total 1 698 23.16 15.53

Data

393

Table A2.16. (cont.) Necessary time in

N

Paid labour

Unpaid household Personal labour care

Strictly egalitarian divorce (time-egalitarian) without kids men 194 18.08 11.79 women 194 23.81 11.79 total 388 20.95 11.79 with kids men 655 21.61 16.76 women 655 25.18 16.76 total 1 310 23.40 16.76 total men 849 20.74 15.53 women 849 24.85 15.53 total 1 698 22.79 15.53

Discret’y time

61.60 61.60 61.60

76.54 70.80 73.67

61.60 61.60 61.60

68.03 64.46 66.25

61.60 61.60 61.60

70.13 66.02 68.08

one adult

couples

earner

single-

earners

dual-

kids

Without

employed

total

employed

home

stay-at-

employed

80

total

268

868

women

total

161

600

men

total

80 81

21

women

men women

59

81

60

21

628 1 256

628

men

total

women

men

women total

men

N

19.68

18.87

20.10

17.86

27.57 8.27

35.96

32.83

37.04

0.00

0.00

0.00

8.77 9.82

10.86

Pre-taxes, pre-transfers, pre-child-care-support

Paid

13.42

13.42

13.42

14.02

8.77 19.20

8.04

11.37

6.90

19.92

21.84

14.23

10.31 8.40

6.48

labour

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

care

81.24

82.05

80.82

82.46

78.00 86.87

70.34

70.14

70.41

94.42

92.50

100.12

95.26 96.13

96.99

time

22.07

21.10

22.58

19.17

29.63 8.85

38.61

35.12

39.80

0.00

0.00

0.00

9.65 10.79

11.94

labour

13.42

13.42

13.42

14.02

8.77 19.20

8.04

11.37

6.90

19.92

21.84

14.23

10.31 8.40

6.48

labour

household

Personal Discret’y

Unpaid

household

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

care

78.85

79.82

78.34

81.15

75.94 86.29

67.69

67.85

67.64

94.42

92.50

100.12

94.38 95.15

95.92

time

Post-taxes, post-transfers, post-child-care-support

22.07

21.10

22.58

19.17

29.63 8.85

38.61

35.12

39.80

0.00

0.00

0.00

9.65 10.79

11.94

13.42

13.42

13.42

14.02

8.77 19.20

8.04

11.37

6.90

19.92

21.84

14.23

10.31 8.40

6.48

labour

household

Unpaid

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

care

78.85

79.82

78.34

81.15

75.94 86.29

67.69

67.85

67.64

94.42

92.50

100.12

94.38 95.15

95.92

time

Personal Discret’y

Necessary time in

labour

Personal Discret’y Paid

Necessary time in

Post-taxes, post-transfers, pre-child-care-support

Unpaid

Necessary time in

labour

Paid

Table A2.17. Discretionary time, Sweden

stay-at-

total

home

employed

total

employed

home

stay-at-

employed

total

employed

home

stay-at-

one adult

earner couples

single-

earners

dual-

With kids

total

2 204

116

141

women

total

25

204

men

total

281

total

55 149

141

women

men women

140

140

25

115

116 141

men

total

women

men

women total

25

3 400

total

men

1 700

1 700

2 285

977

women

men

total

women

1 308

total

men

1 287 917

81

total

men women

60

21

women

men

0.00

0.00

0.00

29.48

22.99 31.64

18.44

7.25

29.70

37.01

36.38

37.17

0.00 0.00

0.00

16.35

13.91

18.80

15.79

13.47

17.42

16.23

17.63 14.17

0.00

0.00

0.00

29.89

32.16

20.75

20.98

20.61 21.10

20.77

29.21

12.26

11.59

17.38

10.13

32.16 29.89

20.75

12.58

16.32

8.84

11.52

12.35

10.94

11.29

10.90 11.86

19.92

21.84

14.23

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

84.46

82.18

93.59

63.88

70.74 61.60

75.13

77.88

72.37

65.74

60.58

67.04

82.18 84.46

93.59

85.41

84.11

86.71

87.02

88.52

85.98

86.82

85.81 88.31

94.42

92.50

100.12

0.00

0.00

0.00

0.00

25.88

19.68 27.93

15.03

5.77

24.35

30.16

28.96

30.46

0.00 0.00

0.00

15.18

12.91

17.44

17.58

14.96

19.42

18.07

19.65 15.74

0.00

0.00

14.23

29.89

32.16

20.75

20.98

20.61 21.10

20.77

29.21

12.26

11.59

17.38

10.13

32.16 29.89

20.75

12.58

16.32

8.84

11.52

12.35

10.94

11.29

10.90 11.86

19.92

21.84

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

84.46

82.18

93.59

67.49

74.04 65.31

78.54

79.35

77.73

72.59

68.00

73.74

82.18 84.46

93.59

86.59

85.11

88.06

85.23

87.03

83.98

84.98

83.79 86.74

94.42

92.50

100.12

0.00

0.00

0.00

23.44

18.54 25.07

15.03

5.77

24.35

30.16

28.96

30.46

0.00 0.00

0.00

14.02

11.88

16.15

17.58

14.96

19.42

18.07

19.65 15.74

0.00

0.00

0.00

29.89

32.16

20.75

20.98

20.61 21.10

20.77

29.21

12.26

11.59

17.38

10.13

32.16 29.89

20.75

12.58

16.32

8.84

11.52

12.35

10.94

11.29

10.90 11.86

19.92

21.84

14.23

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

84.46

82.18

93.59

69.92

75.19 68.18

78.54

79.35

77.73

72.59

68.00

73.74

82.18 84.46

93.59

87.74

86.14

89.35

85.23

87.03

83.98

84.98

83.79 86.74

94.42

92.50

100.12

couples

singleearner

earners

dual-

Total

employed

stay-athome

employed

total

employed

Table A2.17. (cont.)

3 885

total

total

total

women

men

220

46

174

222

46 176

4 656

total

men women

2 328

women

2 328

1 990

men

1 895

3 744

total

women

1 874

women

men

1 870

men

N

36.63

34.91

37.12

0.00

0.00 0.00

14.65

12.57

16.72

17.88

16.10

19.86

18.56

17.05

20.17

labour

Paid

Paid

10.32

14.88

9.03

26.30

18.04 28.60

11.49

14.75

8.22

14.06

17.93

9.75

13.46

17.09

9.58

labour

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

care

67.39

64.55

68.19

88.04

96.30 85.74

88.21

87.02

89.40

82.40

80.31

84.73

82.33

80.20

84.59

time

33.19

31.52

33.66

0.00

0.00 0.00

14.03

12.06

16.00

16.29

14.66

18.09

16.90

15.52

18.38

labour

10.32

14.88

9.03

26.30

18.04 28.60

11.49

14.75

8.22

14.06

17.93

9.75

13.46

17.09

9.58

labour

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

care

70.83

67.94

71.65

88.04

96.30 85.74

88.82

87.53

90.11

84.00

81.75

86.50

83.98

81.73

86.39

time

33.19

31.52

33.66

0.00

0.00 0.00

13.18

11.30

15.05

15.08

13.43

16.91

15.65

14.22

17.18

labour

10.32

14.88

9.03

26.30

18.04 28.60

11.49

14.75

8.22

14.06

17.93

9.75

13.46

17.09

9.58

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66

care

labour

70.83

67.94

71.65

88.04

96.30 85.74

89.68

88.29

91.07

85.21

82.98

87.68

85.23

83.03

87.59

time

Personal Discret’y

hold

house-

hold

hold

Personal Discret’y Paid

housePersonal Discret’y

Unpaid

Unpaid

house-

Necessary time in

post-child-care-support

Post-taxes, post-transfers,

Unpaid

Necessary time in

pre-child-care-support

pre-child-care-support Necessary time in

Post-taxes, post-transfers,

Pre-taxes, pre-transfers,

stay-at-

total

total

employed

home

employed

one adult

total

2 967 6 170

women total

5 948

total

3 203

2 791

men

3 157

women

total

men

176

222

women

1 072 46

417

total men

655

442

total

women

222

women

men

220

men

15.12 16.98

18.68

17.55

15.97

18.94

0.00

0.00

21.62 0.00

23.34

20.35

18.23

7.62

28.94

15.85 12.97

10.33

12.52

15.14

10.22

26.30

28.60

14.91 18.04

16.11

14.04

18.35

25.61

11.01

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

83.37 84.39

85.33

84.27

83.23

85.18

88.04

85.74

77.81 96.30

74.90

79.96

77.77

81.11

74.39

14.77 16.85

18.74

17.41

15.60

19.00

0.00

0.00

22.82 0.00

23.49

22.33

16.51

6.88

26.24

15.85 12.97

10.33

12.52

15.14

10.22

26.30

28.60

14.91 18.04

16.11

14.04

18.35

25.61

11.01

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

83.72 84.53

85.27

84.41

83.60

85.12

88.04

85.74

76.61 96.30

74.74

77.98

79.48

81.85

77.09

14.00 16.16

18.13

16.70

14.79

18.38

0.00

0.00

22.34 0.00

22.49

22.23

16.51

6.88

26.24

11.01

15.85 12.97

10.33

12.52

15.14

10.22

26.30

28.60

14.91 18.04

16.11

14.04

18.35

25.61

53.66 53.66

53.66

53.66

53.66

53.66

53.66

53.66

53.66 53.66

53.66

53.66

53.66

53.66

53.66

84.49 85.22

85.88

85.12

84.42

85.74

88.04

85.74

77.09 96.30

75.75

78.07

79.48

81.85

77.09

398

Appendix 2

Table A2.18. Spare time, Sweden Actual time in

N

Paid labour

Unpaid household labour

Personal care

Spare time

565 581 1 146 17 34 51 55 42 97 72 76 148 425 293 718 17 34 51 1 045 916 1 961 1 062 950 2 012

55.17 44.33 49.72 0.00 0.00 0.00 50.07 25.65 39.55 38.58 13.67 25.56 51.65 46.22 49.43 0.00 0.00 0.00 53.40 44.15 49.12 52.61 42.49 47.87

17.04 24.99 21.03 31.92 49.92 44.35 17.79 24.67 20.75 21.03 36.46 29.10 15.11 19.32 16.83 31.92 49.92 44.35 16.25 23.03 19.39 16.48 24.04 20.02

64.37 69.56 66.98 77.64 72.29 73.95 68.64 79.85 73.47 70.70 76.32 73.64 64.41 67.36 65.61 77.64 72.29 73.95 64.60 69.27 66.76 64.80 69.38 66.94

31.43 29.12 30.27 58.44 45.79 49.71 31.50 37.84 34.23 37.68 41.55 39.71 36.83 35.10 36.12 58.44 45.79 49.71 33.74 31.56 32.73 34.11 32.09 33.17

men 997 women 1 041 total 2 038 single-earner stay-at-home men 21 couples women 225 total 246

55.65 35.54 45.74 0.00 0.00 0.00

20.24 36.45 28.23 40.23 65.17 62.94

64.47 68.90 66.66 75.85 71.74 72.11

27.64 27.11 27.37 51.92 31.09 32.96

Without kids dual-earners employed

men women total single-earner stay-at-home men couples women total employed men women total total men women total one adult employed men women total total stay-at-home men women total employed men women total total men women total With kids dual-earners

employed

399

Table A2.18. (cont.) Actual time in

N

Paid labour

Unpaid household labour

Personal care

Spare time

174 26 200 195 251 446 40 158 198 21 225 246 1 211 1 225 2 436 1 232 1 450 2 682

48.97 42.07 48.12 43.73 4.32 22.12 44.16 45.25 45.02 0.00 0.00 0.00 54.29 37.06 45.87 53.37 31.32 41.77

22.18 36.74 23.96 24.11 62.25 45.02 34.16 29.71 30.62 40.23 65.17 62.94 21.01 35.50 28.09 21.33 40.10 31.21

69.16 69.30 69.18 69.88 71.49 70.76 64.94 65.15 65.11 75.85 71.74 72.11 65.16 68.38 66.73 65.34 68.90 67.21

27.69 19.89 26.73 30.28 29.94 30.10 24.74 27.89 27.25 51.92 31.09 32.96 27.54 27.07 27.31 27.96 27.69 27.82

men 1 562 women 1 622 total 3 184 single-earner stay-at-home men 38 couples women 259 total 297 employed men 229 women 68 total 297 total men 267 women 327 total 594 one adult employed men 465 women 451 total 916

55.47 38.81 47.20 0.00 0.00 0.00 49.23 31.70 45.31 42.35 6.64 23.01 51.02 45.89 48.51

19.08 32.19 25.59 36.57 62.93 59.50 21.12 29.11 22.91 23.29 55.84 1.91 16.71 22.85 19.71

64.43 69.15 66.77 76.64 71.82 72.45 69.04 75.96 70.58 70.10 72.69 71.50 64.45 66.61 65.51

29.02 27.85 28.44 54.79 33.25 36.06 28.61 31.23 29.19 32.27 32.83 32.57 35.82 32.65 34.27

employed

one adult

total

Total dual-earners

men women total total men women total employed men women total stay-at-home men women total employed men women total total men women total employed

400

Appendix 2

Table A2.18. (cont.) Actual time in

total

stay-at-home men women total employed men women total total men women total

N

Paid labour

Unpaid household labour

Personal care

Spare time

38 259 297 2 256 2 141 4 397 2 294 2 400 4 694

0.00 0.00 0.00 53.87 40.23 47.37 53.01 35.95 44.48

36.57 62.93 59.50 18.76 29.93 24.08 19.04 33.43 26.24

76.64 71.82 72.45 64.90 68.77 66.74 65.08 69.10 67.09

54.79 33.25 36.06 30.48 29.07 29.81 30.86 29.52 30.19

Data

401

Table A2.19. The state impact on discretionary time and the time-pressure illusion, Sweden Impact of state taxesand-transfers on discret’y time Without kids dualemployed earners singleearner couples

stay-athome employed

total

one adult

employed

total

stay-athome employed

total

With kids dual-earners

employed

single-earner stay-atcouples home

Impact of state childcare support on discret’y time

Total state impact on Timediscret’y pressure time illusion

men women total men women total men women total men women total men women total men women total men women total men women total

1.08 0.88 0.98 0.00 0.00 0.00 2.77 2.29 2.64 2.06 0.58 1.31 2.48 2.23 2.39 0.00 0.00 0.00 2.02 1.57 1.84 2.00 1.49 1.79

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

1.08 0.88 0.98 0.00 0.00 0.00 2.77 2.29 2.64 2.06 0.58 1.31 2.48 2.23 2.39 0.00 0.00 0.00 2.02 1.57 1.84 2.00 1.49 1.79

64.48 65.26 64.88 41.67 46.71 44.71 36.14 30.01 33.46 38.26 44.74 41.44 41.51 44.72 42.73 41.67 46.71 44.71 50.04 55.19 52.24 49.86 54.94 52.07

men women total men women total

1.35 1.00 1.18 0.00 0.00 0.00

1.29 1.03 1.16 0.00 0.00 0.00

2.64 2.03 2.34 0.00 0.00 0.00

61.71 59.03 60.37 41.67 51.09 51.50

402

Appendix 2

Table A2.19. (cont.) Impact of state taxesand-transfers on discret’y time employed

total

one adult

employed

total

stay-athome employed

total

Total dual-earners

employed

single-earner stay-atcouples home employed

total

Impact of state childcare support on discret’y time

Total state impact on Timediscret’y pressure time illusion

men women total men women total men women total men women total men women total men women total

6.70 7.42 6.85 5.36 1.48 3.41 3.30 3.70 3.60 0.00 0.00 0.00 1.79 1.53 1.66 1.77 1.44 1.60

0.00 0.00 0.00 0.00 0.00 0.00 1.14 2.87 2.44 0.00 0.00 0.00 1.20 1.30 1.25 1.18 1.23 1.21

6.70 7.42 6.85 5.36 1.48 3.41 4.45 6.57 6.04 0.00 0.00 0.00 3.00 2.83 2.91 2.95 2.67 2.80

46.05 48.11 45.86 47.44 49.41 48.45 50.45 40.28 42.68 41.67 51.09 51.50 60.04 55.96 57.93 59.72 55.29 57.39

men women total men women total men women total men women total

0.72 0.51 0.62 0.00 0.00 0.00 3.46 3.38 3.44 2.70 0.74 1.71

0.95 0.76 0.86 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

1.67 1.27 1.47 0.00 0.00 0.00 3.46 3.38 3.44 2.70 0.74 1.71

62.05 60.44 61.24 41.51 52.49 51.99 43.05 36.71 41.64 44.82 49.02 46.91

Data

403

Table A2.19. (cont.) Impact of state taxesand-transfers on discret’y time one adult

employed

total

stay-athome employed

total

men women total men women total men women total men women total

1.98 0.15 1.21 0.00 0.00 0.00 0.06 0.37 0.14 0.06 0.35 0.14

Impact of state childcare support on discret’y time 0.10 1.00 0.48 0.00 0.00 0.00 0.62 0.82 0.71 0.61 0.77 0.69

Total state impact on Timediscret’y pressure time illusion 1.88 0.85 0.73 0.00 0.00 0.00 0.56 1.18 0.85 0.55 1.12 0.82

42.26 43.09 42.82 41.51 52.49 51.99 55.26 55.35 55.32 55.02 54.97 55.03

404

Appendix 2

Table A2.20. Discretionary time (post-taxes, post-transfers, post-child-caresupport) under alternative household rules, Sweden Necessary time in

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

628 628 1 256

11.94 9.65 10.79

6.48 10.31 8.40

53.66 53.66 53.66

95.92 94.38 95.15

1 700 1 700 3 400

16.15 11.88 14.02

8.84 16.32 12.58

53.66 53.66 53.66

89.35 86.14 87.74

2 328 2 328 4 656

15.05 11.30 13.18

8.22 14.75 11.49

53.66 53.66 53.66

91.07 88.29 89.68

628 628 1 256

21.81 0.00 10.91

6.90 21.84 14.37

53.66 53.66 53.66

85.63 92.50 89.07

1 700 1 700 3 400

24.00 0.00 12.00

9.69 30.77 20.23

53.66 53.66 53.66

80.65 83.57 82.11

2 328 2 328 4 656

23.43 0.00 11.71

8.96 28.44 18.70

53.66 53.66 53.66

81.95 85.90 83.93

Most-efficient breadwinner without kids men 628 women 628 total 1 256

12.89 4.98 8.93

6.93 23.11 15.02

53.66 53.66 53.66

94.52 86.26 90.39

Conventional dual-earner without kids men women total with kids men women total total men women total Male breadwinner without kids men women total with kids men women total total men women total

Data

405

Table A2.20. (cont.) Necessary time in

with kids men women total total men women total

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

1 699 1 699 3 398

13.66 5.64 9.65

10.82 31.14 20.98

53.66 53.66 53.66

89.87 77.56 83.72

2 327 2 327 4 654

13.46 5.46 9.46

9.80 29.04 19.42

53.66 53.66 53.66

91.08 79.84 85.46

11.75 13.69 12.72

9.49 9.49 9.49

53.66 53.66 53.66

93.10 91.16 92.13

14.75 22.71 18.73

13.36 13.36 13.36

53.66 53.66 53.66

86.22 78.27 82.25

13.97 20.35 17.16

12.35 12.35 12.35

53.66 53.66 53.66

88.02 81.64 84.83

10.79 10.79 10.79

9.49 9.49 9.49

53.66 53.66 53.66

94.06 94.06 94.06

14.14 14.14 14.14

13.36 13.36 13.36

53.66 53.66 53.66

86.84 86.84 86.84

13.26 13.26 13.26

12.35 12.35 12.35

53.66 53.66 53.66

88.73 88.73 88.73

Equal monetary contribution without kids men 628 women 628 total 1 256 with kids men 1 700 women 1 700 total 3 400 total men 2 328 women 2 328 total 4 656 Equal temporal contribution without kids men 628 women 628 total 1 256 with kids men 1 700 women 1 700 total 3 400 total men 2 328 women 2 328 total 4 656

406

Appendix 2

Table A2.20. (cont.) Necessary time in

Atomistic divorce without kids men women total with kids men women total total men women total Self-reliant divorce without kids men women total with kids men women total total men women total Gendered divorce without kids men women total

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

628 628 1 256

17.10 20.14 18.62

13.42 13.42 13.42

53.66 53.66 53.66

83.82 80.78 82.30

1 700 1 700 3 400

18.38 27.75 23.07

13.42 13.42 13.42

53.66 53.66 53.66

82.54 73.17 77.85

2 328 2 328 4 656

18.05 25.76 21.91

13.42 13.42 13.42

53.66 53.66 53.66

82.87 75.16 79.01

628 628 1 256

17.10 20.14 18.62

13.42 13.42 13.42

53.66 53.66 53.66

83.82 80.78 82.30

1 699 1 699 3 398

23.66 35.15 29.41

22.93 22.93 22.93

53.66 53.66 53.66

67.75 56.26 62.01

2 327 2 327 4 654

21.94 31.23 26.59

20.44 20.44 20.44

53.66 53.66 53.66

71.95 62.67 67.31

628 628 1 256

17.10 20.14 18.62

13.42 13.42 13.42

53.66 53.66 53.66

83.82 80.78 82.30

Data

407

Table A2.20. (cont.) Necessary time in

with kids men women total total men women total Male-breadwinner divorce without kids men women total with kids men women total total men women total

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

1 700 1 699 3 399

18.38 35.15 26.77

13.42 22.93 18.17

53.66 53.66 53.66

82.54 56.26 69.40

2 328 2 327 4 655

18.05 31.23 24.64

13.42 20.44 16.93

53.66 53.66 53.66

82.87 62.67 72.77

628 628 1 256

17.10 20.14 18.62

13.42 13.42 13.42

53.66 53.66 53.66

83.82 80.78 82.30

1 699 1 700 3 399

23.66 27.75 25.71

13.42 22.93 18.18

53.66 53.66 53.66

77.26 63.66 70.46

2 327 2 328 4 655

21.94 25.76 23.85

13.42 20.44 16.93

53.66 53.66 53.66

78.98 68.13 73.55

17.10 20.14 18.62

13.42 13.42 13.42

53.66 53.66 53.66

83.82 80.78 82.30

21.67 29.07 25.37

15.86 20.49 18.17

53.66 53.66 53.66

76.82 64.78 70.80

20.48 26.73 23.60

15.22 18.65 16.93

53.66 53.66 53.66

78.65 68.96 73.80

Most-efficient breadwinner divorce without kids men 628 women 628 total 1 256 with kids men 1 699 women 1 699 total 3 398 total men 2 327 women 2 327 total 4 654

408

Appendix 2

Table A2.20. (cont.) Necessary time in

N

Paid labour

Unpaid household labour

Financially egalitarian divorce (money-egalitarian) without kids men 628 17.10 13.42 women 628 20.14 13.42 total 1 256 18.62 13.42 with kids men 1 699 21.04 13.42 women 1 699 31.57 22.93 total 3 398 26.30 18.17 total men 2 327 20.01 13.42 women 2 327 28.58 20.44 total 4 654 24.29 16.93 Financially egalitarian divorce (time-egalitarian) without kids men 628 17.10 13.42 women 628 20.14 13.42 total 1 256 18.62 13.42 with kids men 1 699 21.24 13.42 women 1 699 30.53 22.93 total 3 398 25.89 18.17 total men 2 327 20.16 13.42 women 2 327 27.81 20.44 total 4 654 23.99 16.93 Strictly egalitarian divorce (money-egalitarian) without kids men 628 17.10 13.42 women 628 20.14 13.42 total 1 256 18.62 13.42

Personal care

Discret’y time

53.66 53.66 53.66

83.82 80.78 82.30

53.66 53.66 53.66

79.88 59.85 69.86

53.66 53.66 53.66

80.91 65.32 73.11

53.66 53.66 53.66

83.82 80.78 82.30

53.66 53.66 53.66

79.68 60.88 70.28

53.66 53.66 53.66

80.76 66.08 73.42

53.66 53.66 53.66

83.82 80.78 82.30

Data

409

Table A2.20. (cont.) Necessary time in

with kids men women total total men women total

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

1 699 1 699 3 398

21.04 31.57 26.30

18.17 18.17 18.17

53.66 53.66 53.66

75.13 64.60 69.86

2 327 2 327 4 654

20.01 28.58 24.29

16.93 16.93 16.93

53.66 53.66 53.66

77.40 68.83 73.11

13.42 13.42 13.42

53.66 53.66 53.66

83.82 80.78 82.30

18.17 18.17 18.17

53.66 53.66 53.66

74.92 65.64 70.28

16.93 16.93 16.93

53.66 53.66 53.66

77.25 69.60 73.42

Strictly egalitarian divorce (time-egalitarian) without kids men 628 17.10 women 628 20.14 total 1 256 18.62 with kids men 1 699 21.24 women 1 699 30.53 total 3 398 25.89 total men 2 327 20.16 women 2 327 27.81 total 4 654 23.99

stay-at-

total

home

employed

total

employed

stay-athome

employed

one adult

couples

singleearner

earners

dual-

Without kids

0.00

0.00 0.00

37 14.84

73 16.45

women

total

10

16

women

total

6

32 11.88 36 18.07

total men

men

16

women

0.00

0.00

0.00

8.17

16 15.59

16 23.76

6 23.52

10 23.89

16

6 10

128 10.09

64 10.04

64 10.15

men

total

women

men

total

men women

total

women

men

N

Paid labour

7.20

13.43

14.81

10.84

12.43

12.43

11.35 12.43

13.79

8.91

9.27

11.87

7.89

13.43

10.84 14.81

8.79

10.38

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

99.35

97.96

101.94

83.90

85.51

89.55 82.28

90.82

88.27

79.75

77.39

81.00

99.35

101.94 97.96

93.90

92.36

95.43

0.00

0.00

0.00

22.67

20.89

14.48 24.48

9.91

19.05

28.97

28.53

29.20

0.00

0.00 0.00

12.37

12.30

12.45

13.43

14.81

10.84

12.43

12.43

11.35 12.43

13.79

8.91

9.27

11.87

7.89

13.43

10.84 14.81

8.79

10.38

7.20

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

household Personal labour care

Unpaid

99.35

97.96

101.94

77.68

79.46

86.94 75.87

89.07

84.81

74.54

72.38

75.69

99.35

101.94 97.96

91.62

90.10

93.13

0.00

0.00

0.00

22.67

20.89

14.48 24.48

9.91

19.05

28.97

28.53

29.20

0.00

0.00 0.00

12.37

12.30

12.45

Discret’y Paid time labour

Post-taxes, post-transfers, pre-child-care-support Necessary time in

household Personal Discret’y Paid labour care time labour

Unpaid

Necessary time in

Pre-taxes, pre-transfers, pre-child-care-support

Table A2.21. Discretionary time, Finland

13.43

14.81

10.84

12.43

12.43

11.35 12.43

13.79

8.91

9.27

11.87

7.89

13.43

10.84 14.81

8.79

10.38

7.20

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

household Personal labour care

Unpaid

Necessary time in

99.35

97.96

101.94

77.68

79.46

86.94 75.87

89.07

84.81

74.54

72.38

75.69

99.35

101.94 97.96

91.62

90.10

93.13

Discret’y time

Post-taxes, post-transfers, post-child-care-support

stay-at-

total

home

employed

total

employed

home

stay-at-

employed

one adult

couples

earner

single-

earners

dual-

With kids

total

employed

233 12.42

total

0.00

0.00

0.00

35

44

women

total

9

0.00

0.00

0.00

17 36.03

total

men

13 38.29

women

4 25.81

88 14.96

total

men

44 22.03 44 7.89

44 29.92

9 37.94

35 27.81

44

35

9

72 14.32 144 14.42

men women

total

women

men

total

women

men

women total

72 14.52

117 11.36

men

116 13.49

217 13.35

total

women

107 12.48

women

men

110 14.17

men

9.06

26.59

28.30

20.08

20.86

20.86

20.85

19.23

12.51 25.94

11.86

16.94

10.53

26.59

28.30

20.08

16.84 13.64

10.43

10.33

11.52

9.14

10.10

11.20

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22

86.18

84.47

92.70

55.88

53.62

66.12

78.59

78.23 78.95

71.00

57.90

74.44

86.18

84.47

92.70

81.61 84.72

87.82

90.02

89.90

90.14

89.33

89.10

89.54

0.00

0.00

0.00

37.69

40.02

27.16

12.89

19.00 6.79

25.79

32.66

23.98

0.00

0.00

0.00

15.82 15.93

16.03

16.04

14.79

17.29

17.23

16.26

18.17

9.06

26.59

28.30

20.08

20.86

20.86

20.85

19.23

12.51 25.94

11.86

16.94

10.53

26.59

28.30

20.08

16.84 13.64

10.43

10.33

11.52

9.14

10.10

11.20

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22

86.18

84.47

92.70

54.22

51.89

64.76

80.66

81.27 80.05

75.13

63.17

78.27

86.18

84.47

92.70

80.11 83.21

86.31

86.40

86.46

86.34

85.44

85.32

85.55

0.00

0.00

0.00

32.22

33.56

26.16

12.89

19.00 6.79

25.79

32.66

23.98

0.00

0.00

0.00

14.48 14.64

14.79

16.04

14.79

17.29

17.23

16.26

18.17

9.06

26.59

28.30

20.08

20.86

20.86

20.85

19.23

12.51 25.94

11.86

16.94

10.53

26.59

28.30

20.08

16.84 13.64

10.43

10.33

11.52

9.14

10.10

11.20

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22

86.18

84.47

92.70

59.69

58.35

65.77

80.66

81.27 80.05

75.13

63.17

78.27

86.18

84.47

92.70

81.46 84.50

87.55

86.40

86.46

86.34

85.44

85.32

85.55

earner couples

single-

earners

dual-

Total

total

employed

home

stay-at-

employed

total

employed

Table A2.21. (cont.)

249 16.19

total

120 14.13

7.96

60

total

women

60 28.26

total

60 20.29

15 32.44

men

45 26.90

0.00 0.00

women

45 60

0.00

men

women total

15

272 12.38

total

men

136 12.30

women

136 12.46

129 14.87

women

men

120 17.64

205 19.71

total

men

111 19.13 94 20.37

labour

men women

N

Paid

17.11

22.67

11.54

11.16

15.00

9.91

25.16 23.05

16.55

11.35

13.79

8.91

16.16

20.40

11.50

13.89

10.78 17.48

labour

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

care

81.54

82.15

80.94

73.36

65.33

75.97

87.62 89.73

96.22

89.04

86.68

91.41

80.43

77.51

83.63

79.18

82.87 74.93

time

13.32

7.63

19.01

26.64

31.09

25.20

0.00 0.00

0.00

14.25

14.16

14.34

16.43

15.51

17.45

20.00

18.92 21.25

labour

household Personal Discret’y Paid

Unpaid

17.11

22.67

11.54

11.16

15.00

9.91

25.16 23.05

16.55

11.35

13.79

8.91

16.16

20.40

11.50

13.89

10.78 17.48

labour

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

care

household Personal

Unpaid

Necessary time in

82.35

82.48

82.22

74.97

66.68

77.67

87.62 89.73

96.22

87.17

84.82

89.53

80.18

76.86

83.83

78.88

83.08 74.05

time

13.32

7.63

19.01

26.64

31.09

25.20

0.00 0.00

0.00

13.57

13.45

13.69

15.30

14.04

16.68

18.62

18.09 19.24

labour

Discret’y Paid

pre-child-care-support

pre-child-care-support

Necessary time in

Post-taxes, post-transfers,

Pre-taxes, pre-transfers,

17.11

22.67

11.54

11.16

15.00

9.91

25.16 23.05

16.55

11.35

13.79

8.91

16.16

20.40

11.50

13.89

10.78 17.48

labour

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

care

household Personal

Unpaid

Necessary time in

82.35

82.48

82.22

74.97

66.68

77.67

87.62 89.73

96.22

87.85

85.53

90.18

81.32

78.33

84.60

80.26

83.91 76.06

time

Discret’y

post-child-care-support

Post-taxes, post-transfers,

stay-athome

total

total

employed

employed

one adult

236 15.57

246 13.20

482 14.35

women

total

422 16.41

total

men

201 16.18

0.00

women

60

221 16.62

men

total

0.00 0.00

90 20.17

total

15 45

50 21.32

women

men women

40 18.68

men

13.32

16.17

10.32

11.93

14.14

9.91

23.05

16.55 25.16

14.03

14.76

13.09

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

85.10

83.41

86.88

84.44

82.46

86.25

89.73

96.22 87.62

78.58

76.70

81.01

16.24

15.17

17.37

18.56

18.59

18.54

0.00

0.00 0.00

25.52

26.18

24.69

13.32

16.17

10.32

11.93

14.14

9.91

23.05

16.55 25.16

14.03

14.76

13.09

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

55.22

75.00

83.22

81.44

85.08

82.28

80.04

84.33

89.73

96.22 87.62

73.23

71.84

24.61

15.66

14.40

16.99

17.90

17.65

18.13

0.00

0.00 0.00

24.49

24.39

13.09

13.32

16.17

10.32

11.93

14.14

9.91

23.05

16.55 25.16

14.03

14.76

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22

55.22 55.22

55.22

55.22

75.08

83.80

82.21

85.47

82.95

80.99

84.74

89.73

96.22 87.62

74.26

73.63

414

Appendix 2

Table A2.22. Spare time, Finland Actual time in

Without kids dual-earners employed

singleearner couples

stay-athome employed

total

one adult

employed

total

stay-athome employed

total

With kids dual-earners employed

singleearner couples

stay-athome

N

Paid labour

Unpaid household labour

127 128 255 12 20 32 20 12 32 32 32 64 72 74 146 12 20 32 219 214 433 231 234 465

41.84 38.24 40.03 0.00 0.00 0.00 51.07 36.62 46.05 33.33 12.72 23.02 47.45 32.65 40.02 0.00 0.00 0.00 44.65 36.13 40.47 42.50 32.88 37.67

17.08 21.22 19.16 24.11 31.81 29.13 15.22 22.15 17.63 18.30 28.45 23.38 14.03 18.47 16.26 24.11 31.81 29.13 15.85 20.28 18.02 16.25 21.32 18.79

66.69 68.61 67.66 83.31 83.03 83.13 67.01 71.19 68.46 72.67 78.92 75.79 64.05 67.81 65.94 83.31 83.03 83.13 65.81 68.45 67.11 66.66 69.77 68.22

42.39 39.93 41.16 60.59 53.16 55.74 34.70 38.04 35.86 43.70 47.91 45.80 42.47 49.07 45.78 60.59 53.16 55.74 41.68 43.13 42.40 42.60 44.04 43.32

men 142 women 144 total 286 men 17 women 70 total 87

40.71 28.73 34.67 0.00 0.00 0.00

22.99 37.29 30.20 47.93 53.57 52.46

67.14 68.49 67.82 72.49 74.27 73.92

37.16 33.49 35.31 47.58 40.16 41.63

men women total men women total men women total men women total men women total men women total men women total men women total

Personal care

Spare time

Data

415

Table A2.22. (cont.) Actual time in

employed

total

one adult

employed

total

stay-athome employed

total

Total dual-earners employed

singleearner couples

stay-athome employed

total

one adult

employed

N

Paid labour

Unpaid household labour

men women total men women total men women total men women total men women total men women total

70 18 88 87 88 175 6 25 31 17 70 87 218 187 405 235 257 492

43.61 20.98 38.90 34.97 4.36 19.57 29.18 25.98 26.47 0.00 0.00 0.00 41.37 27.56 34.91 38.29 20.09 28.69

21.43 41.61 25.62 26.67 51.09 38.96 20.05 33.49 31.41 47.93 53.57 52.46 22.41 37.13 29.29 24.31 41.59 33.42

68.04 69.12 68.27 68.92 73.20 71.07 64.26 72.41 71.15 72.49 74.27 73.92 67.36 69.15 68.20 67.75 70.53 69.22

34.93 36.29 35.21 37.43 39.36 38.40 54.51 36.13 38.96 47.58 40.16 41.63 36.85 34.16 35.60 37.65 35.79 36.67

men women total men women total men women total men women total men women total

269 272 541 29 90 119 90 30 120 119 120 239 78 99 177

41.24 33.21 37.20 0.00 0.00 0.00 45.34 26.94 40.83 34.52 6.61 20.51 46.29 30.86 37.59

20.20 29.72 24.99 38.50 48.50 46.11 19.98 34.19 23.47 24.40 44.99 34.74 14.41 22.51 18.98

66.93 68.55 67.74 76.77 76.31 76.42 67.80 69.91 68.32 69.94 74.74 72.35 64.07 69.04 66.87

39.63 36.53 38.07 52.73 43.19 45.47 34.87 36.96 35.39 39.14 41.66 40.40 43.23 45.59 44.56

Personal care

Spare time

416

Appendix 2

Table A2.22. (cont.) Actual time in

total

stay-athome employed

total

men women total men women total men women total

N

Paid labour

Unpaid household labour

29 90 119 437 401 838 466 491 957

0.00 0.00 0.00 43.05 32.13 37.81 40.40 26.20 33.09

38.50 48.50 46.11 19.06 28.15 23.42 20.26 31.91 26.26

Personal care

Spare time

76.77 76.31 76.42 66.57 68.78 67.63 67.20 70.17 68.73

52.73 43.19 45.47 39.32 38.95 39.14 40.14 39.73 39.93

Data

417

Table A2.23. The state impact on discretionary time and the time-pressure illusion, Finland Impact of state taxesand-transfers on discret’y time

Impact of state childcare support on discret’y time

Total state impact on Timediscret’y pressure time illusion

Without kids dualemployed men earners women total singlestay-atmen home women earner couples total employed men women total total men women total one adult employed men women total total stay-atmen home women total employed men women total total men women total

2.30 2.27 2.28 0.00 0.00 0.00 5.31 5.01 5.21 3.47 1.74 2.60 6.40 6.05 6.23 0.00 0.00 0.00 3.99 3.78 3.89 3.80 3.44 3.62

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

2.30 2.27 2.28 0.00 0.00 0.00 5.31 5.01 5.21 3.47 1.74 2.60 6.40 6.05 6.23 0.00 0.00 0.00 3.99 3.78 3.89 3.80 3.44 3.62

50.74 50.16 50.46 41.35 44.80 43.60 40.98 34.34 38.67 41.11 41.17 41.14 33.41 30.40 31.89 41.35 44.80 43.60 43.87 42.19 43.04 43.74 42.43 43.08

With kids dualemployed men earners women total singlestay-atmen home women earner couples total

1.51 1.50 1.50 0.00 0.00 0.00

1.24 1.34 1.29 0.00 0.00 0.00

0.27 0.16 0.21 0.00 0.00 0.00

50.39 47.97 49.19 45.12 44.31 44.55

418

Appendix 2

Table A2.23. (cont.) Impact of state taxesand-transfers on discret’y time

Impact of state childcare support on discret’y time

Total state impact on Timediscret’y pressure time illusion

employed men women total total men women total one adult employed men women total total stay-atmen home women total employed men women total total men women total

3.83 5.27 4.13 3.03 1.10 2.07 1.35 1.73 1.66 0.00 0.00 0.00 0.21 0.88 0.29 0.20 0.64 0.24

0.00 0.00 0.00 0.00 0.00 0.00 1.00 6.46 5.47 0.00 0.00 0.00 0.83 2.01 1.38 0.77 1.47 1.14

3.83 5.27 4.13 3.03 1.10 2.07 0.35 4.73 3.81 0.00 0.00 0.00 1.05 1.13 1.09 0.96 0.83 0.89

43.34 26.88 39.92 43.83 40.69 42.26 11.25 22.22 20.73 45.12 44.31 44.55 47.06 41.90 44.67 46.94 42.55 44.65

Total dualemployed men earners women total singlestay-atmen earner home women couples total employed men women total total men women total one adult employed men women total

1.88 1.86 1.87 0.00 0.00 0.00 1.70 1.35 1.61 1.28 0.33 0.81 6.01 4.85 5.36

0.66 0.71 0.68 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.08 1.78 1.04

1.22 1.15 1.19 0.00 0.00 0.00 1.70 1.35 1.61 1.28 0.33 0.81 5.93 3.07 4.32

50.55 49.00 49.79 43.49 44.43 44.26 42.79 29.72 39.58 43.08 40.82 41.95 31.85 28.04 29.70

Data

419

Table A2.23. (cont.) Impact of state taxesand-transfers on discret’y time total

stay-athome

men women total employed men women total total men women total

0.00 0.00 0.00 1.92 2.42 2.16 1.80 1.97 1.89

Impact of state childcare support on discret’y time 0.00 0.00 0.00 0.41 0.94 0.66 0.39 0.77 0.58

Total state impact on Timediscret’y pressure time illusion 0.00 0.00 0.00 1.51 1.48 1.49 1.41 1.20 1.31

43.49 44.43 44.26 45.43 42.04 43.81 45.33 42.48 43.87

420

Appendix 2

Table A2.24. Discretionary time (post-taxes, post-transfers, post-child-caresupport) under alternative household rules, Finland Necessary time in

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

64 64 128

12.45 12.30 12.37

7.20 10.38 8.79

55.22 55.22 55.22

93.13 90.10 91.62

72 72 144

14.79 14.48 14.64

10.43 16.84 13.64

55.22 55.22 55.22

87.55 81.46 84.50

136 136 272

13.69 13.45 13.57

8.91 13.79 11.35

55.22 55.22 55.22

90.18 85.53 87.85

64 64 128

23.12 0.00 11.56

7.89 14.81 11.35

55.22 55.22 55.22

81.77 97.96 89.86

72 72 144

20.17 0.00 10.09

9.39 24.83 17.11

55.22 55.22 55.22

83.22 87.95 85.58

136 136 272

21.56 0.00 10.78

8.68 20.11 14.40

55.22 55.22 55.22

82.53 92.67 87.60

Most-efficient breadwinner without kids men 64 women 64 total 128

14.34 5.87 10.11

6.05 16.65 11.35

55.22 55.22 55.22

92.38 90.25 91.32

Conventional dual-earner without kids men women total with kids men women total total men women total Male breadwinner without kids men women total with kids men women total total men women total

Data

421

Table A2.24. (cont.) Necessary time in

with kids men women total total men women total

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

72 72 144

15.22 3.55 9.39

5.90 28.31 17.11

55.22 55.22 55.22

91.65 80.91 86.28

136 136 272

14.81 4.64 9.73

5.97 22.82 14.40

55.22 55.22 55.22

91.99 85.32 88.65

11.75 14.95 13.35

8.79 8.79 8.79

55.22 55.22 55.22

92.24 89.04 90.64

13.08 18.83 15.96

13.64 13.64 13.64

55.22 55.22 55.22

86.06 80.31 83.18

12.46 17.00 14.73

11.35 11.35 11.35

55.22 55.22 55.22

88.97 84.43 86.70

12.16 12.16 12.16

8.79 8.79 8.79

55.22 55.22 55.22

91.83 91.83 91.83

14.54 14.54 14.54

13.64 13.64 13.64

55.22 55.22 55.22

84.60 84.60 84.60

13.41 13.41 13.41

11.35 11.35 11.35

55.22 55.22 55.22

88.01 88.01 88.01

Equal monetary contribution without kids men 64 women 64 total 128 with kids men 72 women 72 total 144 total men 136 women 136 total 272 Equal temporal contribution without kids men 64 women 64 total 128 with kids men 72 women 72 total 144 total men 136 women 136 total 272

422

Appendix 2

Table A2.24. (cont.) Necessary time in

Atomistic divorce without kids men women total with kids men women total total men women total Self-reliant divorce without kids men women total with kids men women total total men women total Gendered divorce without kids men women total

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

64 64 128

18.00 22.93 20.47

12.43 12.43 12.43

55.22 55.22 55.22

82.35 77.42 79.89

72 72 144

16.20 23.05 19.63

12.43 12.43 12.43

55.22 55.22 55.22

84.15 77.30 80.72

136 136 272

17.05 23.00 20.02

12.43 12.43 12.43

55.22 55.22 55.22

83.30 77.35 80.33

64 64 128

18.00 22.93 20.46

12.43 12.43 12.43

55.22 55.22 55.22

82.35 77.42 79.89

72 72 144

21.55 30.90 26.22

23.39 23.39 23.39

55.22 55.22 55.22

67.84 58.49 63.17

136 136 272

19.88 27.14 23.51

18.22 18.22 18.22

55.22 55.22 55.22

74.68 67.41 71.04

64 64 128

18.00 22.93 20.46

12.43 12.43 12.43

55.22 55.22 55.22

82.35 77.42 79.89

Data

423

Table A2.24. (cont.) Necessary time in

with kids men women total total men women total Male-breadwinner divorce without kids men women total with kids men women total total men women total

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

72 72 144

16.20 30.90 23.55

12.43 23.39 17.91

55.22 55.22 55.22

84.15 58.49 71.32

136 136 272

17.05 27.14 22.09

12.43 18.22 15.32

55.22 55.22 55.22

83.30 67.41 75.36

64 64 128

18.00 22.93 20.46

12.43 12.43 12.43

55.22 55.22 55.22

82.35 77.42 79.89

72 72 144

21.55 23.05 22.30

12.43 23.39 17.91

55.22 55.22 55.22

78.80 66.33 72.57

136 136 272

19.88 23.00 21.44

12.43 18.22 15.32

55.22 55.22 55.22

80.48 71.56 76.02

18.00 22.93 20.46

12.43 12.43 12.43

55.22 55.22 55.22

82.35 77.42 79.89

20.05 24.12 22.09

14.24 21.57 17.91

55.22 55.22 55.22

78.48 67.08 72.78

19.08 23.56 21.32

13.39 17.26 15.32

55.22 55.22 55.22

80.31 71.95 76.13

Most-efficient breadwinner divorce without kids men 64 women 64 total 128 with kids men 72 women 72 total 144 total men 136 women 136 total 272

424

Appendix 2

Table A2.24. (cont.) Necessary time in

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

55.22 55.22 55.22

82.35 77.42 79.89

55.22 55.22 55.22

81.47 62.39 71.93

55.22 55.22 55.22

81.89 69.47 75.68

12.43 12.43 12.43

55.22 55.22 55.22

82.35 77.42 79.89

12.43 23.39 17.91

55.22 55.22 55.22

81.22 63.37 72.29

12.43 18.22 15.32

55.22 55.22 55.22

81.75 69.99 75.87

12.43 12.43 12.43

55.22 55.22 55.22

82.35 77.42 79.89

Financially egalitarian divorce (money-egalitarian) without kids men 64 18.00 12.43 women 64 22.93 12.43 total 128 20.46 12.43 with kids men 72 18.88 12.43 women 72 27.00 23.39 total 144 22.94 17.91 total men 136 18.46 12.43 women 136 25.08 18.22 total 272 21.77 15.32 Financially egalitarian divorce (time-egalitarian) without kids men 64 18.00 women 64 22.93 total 128 20.46 with kids men 72 19.13 women 72 26.02 total 144 22.57 total men 136 18.60 women 136 24.56 total 272 21.58 Strictly egalitarian divorce (money-egalitarian) without kids men 64 18.00 women 64 22.93 total 128 20.46

Data

425

Table A2.24. (cont.) Necessary time in

with kids men women total total men women total

N

Paid labour

Unpaid household labour

Personal care

Discret’y time

72 72 144

18.88 27.00 22.94

17.91 17.91 17.91

55.22 55.22 55.22

75.99 67.87 71.93

136 136 272

18.46 25.08 21.77

15.32 15.32 15.32

55.22 55.22 55.22

78.99 72.37 75.68

12.43 12.43 12.43

55.22 55.22 55.22

82.35 77.42 79.89

17.91 17.91 17.91

55.22 55.22 55.22

75.74 68.85 72.29

15.32 15.32 15.32

55.22 55.22 55.22

78.86 72.89 75.87

Strictly egalitarian divorce (time-egalitarian) without kids men 64 18.00 women 64 22.93 total 128 20.46 with kids men 72 19.13 women 72 26.02 total 144 22.57 total men 136 18.60 women 136 24.56 total 272 21.58

Bibliography

Abraham, Katharine G. and Christopher Mackie, eds. 2005. Beyond the Market: Designing Nonmarket Accounts for the United States. Report of the Panel to Study the Design of Nonmarket Accounts, Committee on National Statistics, Division of Behavioral and Social Sciences & Education, National Research Council. Washington, DC: National Academies Press. Abrahamson, Peter. 1999. The Scandinavian model of welfare. Vol. IV, pp. 15–30 in Comparing Social Welfare Systems in Nordic Europe and France, ed. Denis Bouget and Bruno Palier. Paris: MIRE-DREES. Abramson, L. Y., G. I. Metalsky and L. B. Alloy. 1989. Hopelessness depression: a theory-based subtype of depression. Psychological Review 96: 358–72. Adam, Barbara. 2004. Time. Cambridge: Polity. Adam Smith Institute. 2005. About Tax Freedom Day. Available at www. adamsmith.org/tax/Tax%20Freedom%20Day/About%20TFD.html. Adams, Charles. 2000. Slavery reexamined. Available at: www.fff.org/freedom/ 0100g.asp (Part I) and www.fff.org/freedom/0200e.asp (Part II). Alessina, Alberto, Edward Glaeser and Bruce Sacerdote. 2005. Work and leisure in the US and Europe: why so different? NBER Working Paper No. 11278. Boston: NBER. Available at: www.nber.org/papers/w11278. Ambler, John S., ed. 1991. The French Welfare State. New York University Press. Andorka, Rudolf. 1987. Time budgets and their uses. Annual Review of Sociology 13: 149–64. Anonymous. n.d. France 1994: survey information. Available at: www. lisproject.org/techdoc/fr/fr94survey.pdf. Arneson, Richard J. 1989. Equality and equality of opportunity for welfare. Philosophical Studies 56: 77–93. 1997. Egalitarianism and the undeserving poor. Journal of Political Philosophy 5: 327–50. A˚s, Dagfinn. 1978. Studies of time use: problems and prospects. Acta Sociologica 15: 125–41. 1982. Designs for large scale, time use studies of the 24-hour day. Pp. 17–53 in It’s About Time: Proceedings of the International Research Group on Time Budgets and Social Activities, ed. Z. Staikov.

426

Bibliography

427

Sofia: Institute of Sociology at the Bulgarian Academy of Sciences and Bulgarian Sociological Association. Ashford, Douglas E. 1991. Advantages of complexity: social insurance in France. In Ambler 1991, pp. 32–57. Atkinson, A. B. 1987. On the measurement of poverty. Econometrica 55: 749–64. 1995. Incomes and the Welfare State. Cambridge University Press. 1998. Poverty in Europe. Oxford: Blackwell. Atkinson, A. B., L. Rainwater and T. M. Smeeding. 1995. Income Distribution in OECD Countries: The Evidence from the Luxembourg Income Study (LIS), Social Policy Studies 18. Paris: OECD. Australian Bureau of Statistics (ABS). 1993. Time Use Survey, Australia, 1992: User’s Guide. ABS catalogue No. 4150.0. Canberra: ABS. 1994. How Australians Use Their Time. Revised edn. Catalogue No. 4153.0. Canberra: ABS. 1996. In Tune: Time Use News & Events 1 (June). Available at: www. stmarys.ca/partners/iatur/choi7103.htm. 1998a. How Australians Use Their Time, 1997. Rev. edn. Catalogue No. 4153. Canberra: ABS. 1998b. Time Use Survey, Australia: User’s Guide, 1997. Catalogue No. 4150. Canberra: ABS. Baldwin, Peter. 1990. The Politics of Social Solidarity: Class Bases of the European Welfare State 1875–1975. Cambridge University Press. Bane, Mary Jo. 1988. Politics and policies of the feminization of poverty. In Weir et al. 1988, pp. 381–96. Barr, Nicholas. 2001. The Welfare State as Piggy Bank. Oxford University Press. Barrett, Garry F. 2001. The dynamics of participating in Parenting Payment (Single) and the Sole Parent Pension. Policy Research Paper Number 14. Canberra: Department of Family & Community Services, Commonwealth of Australia. Barry, Brian. 1989. Chance, choice and justice. Pp. 142–58 in Liberty and Justice. Oxford University Press. Baxter, Janeen and Diane Gibson. 1990. Double Take: The Links Between Paid and Unpaid Work. Canberra: Australian Government Publishing Service. Becker, Gary S. 1965. A theory of the allocation of time. Economic Journal 75: 493–517. 1976. The Economic Approach to Human Behavior. University of Chicago Press. 1981. A Treatise on the Family. Cambridge, MA: Harvard University Press.

428

Bibliography

Beckerman, Wilfred. 1978. Measures of Leisure, Equality and Welfare. Paris: Organisation for Economic Co-operation and Development. Bell, Linda and Richard Freeman. 2001. The incentive for working hard: explaining hours worked differences in the US and Germany. Labour Economics 8: 181–202. Bergman, Solveig. 2004. Collective organizing and claim making on child care in Norden: blurring the boundaries between inside and outside. Social Politics 11: 217–46. Berlin, Isaiah. 1958. Two concepts of liberty. Pp. 118–72 in Four Essays on Liberty. Oxford: Clarendon Press. Besharov, Douglas J., Jeffrey S. Morrow and Anne Fergyan Shi. 2006. Childcare data in the Survey of Income and Program Participation (SIPP): inaccuracies and corrections. Mimeo., Welfare Reform Academy, University of Maryland. Available at: www.welfareacademy.org/pubs/ sipp_childcare_data.pdf. Beveridge, William H. 1942. Social Insurance and Allied Services. Cmnd. 6404. London: HMSO. Bianchi, Suzanne M., John P. Robinson and Melissa A. Milkie. 2006. Changing Rhythms of American Family Life. New York: Russell Sage Foundation. Bittman, Michael. 1992. Juggling Time: How Australian Families Use Time. 2nd edn. Canberra: Office of the Status of Women. 1999a. Parenthood without penalty: time use and public policy in Australia and Finland. Feminist Economics 5 (3: Nov.): 27–42. 1999b. Social participation and family welfare: the money and time cost of leisure. SPRC Discussion Papers No. 95 (Feb. 1999). Kensington, NSW: Social Policy Research Centre, University of New South Wales. Bittman, Michael, Paula England, Liana Sayer, Nancy Folbre and George Matheson. 2002. When does gender trump money? Bargaining and time in household work. American Journal of Sociology 109 (1: July): 186–214. Bittman, Michael and Jocelyn Pixley. 1997. The Double Life of the Family. Sydney: Allen & Unwin. Bittman, Michael and James Mahmud Rice. 2002. The spectre of overwork: an analysis of trends between 1974 and 1997 using Australian time-use diaries. Labour and Industry 12: 5–25. Bittman, Michael and Judy Wajcman. 2000. The rush hour: the character of leisure time and gender equity. Social Forces 79 (1: Sept.): 165–89. Bjørnskov, Christian. 2003. The happy few: cross-country evidence on social capital and life satisfaction. Kyklos 56: 3–16. Blake, William. 1804. Jerusalem. Pp. 635–847 in The Complete Poems of William Blake, ed. Alicia Ostriker. Harmondsworth: Penguin, 1977.

Bibliography

429

Blanchflower, David and Andrew Oswald. 2006. Happiness and the Human Development Index: the paradox of Australia. National Bureau of Economics Working Paper No. 11416. Available at: www.nber.org/ papers/w11416. Blaug, Marc. 1963. The myth of the old poor law and the making of the new. Journal of Economic History 23: 151–84. Blomfield, L. et al. 1834. Report from His Majesty’s Commission for Enquiring into the Administration and Practical Operation of the Poor Law. London: HMSO. Reprinted in part as The Poor Law Report of 1834, ed. S. G. and E. O. A. Checkland (Harmondsworth: Penguin Books, 1974). Bo¨heim, Rene´ and Mark P. Taylor. 2004. Actual and preferred working hours. British Journal of Industrial Relations 42: 149–66. Boulin, Jean-Yves and Ulrich Mu¨ckenberger. 1999. Times in the city and quality of life. Best: European studies on time. European Foundation. Available at: www.fr.eurofound.eu.int/publications/files/EF9940EN.pdf. Bowles, Samuel and Yongjin Park. 2005. Emulation, inequality and work hours: was Thorsten Veblen right? Economic Journal 115 (Nov.): F397–412. Bradshaw, Jonathan, John Ditch, Hilary Holmes and Peter Whiteford. 1993. Support for Children: A Comparison of Arrangements in Fifteen Countries. Department of Social Security Research Report No. 21. London: HMSO. Braybrooke, David. 1987. Meeting Needs. Princeton University Press. Brock, Daniel W. 1995. Justice and the ADA: does prioritizing and rationing health care discriminate against the disabled? Social Philosophy & Policy 12: 159–85. Bryson, Lois, Michael Bittman and Sue Donath. 1994. Men’s welfare state, women’s welfare state: tendencies to convergence in practice and theory? In Sainsbury 1994, pp. 118–31. Burger, Jerry 1989. Negative reactions to increases in perceived personal control. Journal of Personality and Social Psychology 56 (2): 246–56. Butler, Eamonn. 2005. Tax Freedom Day points up Europe’s problems. Available at: www.adamsmith.org/cissues/tax-and-economy/wsj-tax.htm. Calmfors, Lars and Per Skedinger. 1995. Does active labour-market policy increase employment? Theoretical considerations and empirical evidence from Sweden. Oxford Review of Economic Policy 11 (1: Spring): 91–109. Campbell, Angus, Philip E. Converse and Willard L. Rogers. 1976. The Quality of American Life. New York: Russell Sage. Carney, Terry and Peter Hanks. 1994. Social Security in Australia. Oxford University Press.

430

Bibliography

Castles, Francis G. 1985. The Working Class and the Welfare State: Reflections on the Political Development of the Welfare State in Australia and New Zealand, 1890–1980. Sydney: Allen & Unwin. ed. 1993. Families of Nations. Aldershot: Dartmouth. 1996. Needs-based strategies of social protection in Australia and New Zealand. Pp. 88–115 in Welfare States in Transition, ed. Gøsta EspingAndersen. London: Sage for the United Nations Research Institute for Social Development. ed. 1998. Comparative Public Policy: Patterns of Post-war Transformation. Cheltenham: Elgar. 2001. A farewell to Australia’s welfare state. International Journal of Health Services 31: 537–44. Castles, Francis G. and Michael Flood. 1993. Why divorce rates differ: law, religious belief and modernity. In Castles 1993, pp. 293–326. Castles, Francis G. and Deborah Mitchell. 1992. Identifying welfare state regimes: the links between politics, instruments and outcomes. Governance 5: 1–26. Chambaz, Christine and Claude Martin. 2001. Lone parents, employment and social policy in France: lessons from a family-friendly policy. Pp. 129–50 in Lone Parent, Employment and Social Policy: Crossnational Comparisons, ed. Jane Millar and Karen Rowlingson. Bristol: Policy Press. Citro, Constance F. and Robert T. Michael. 1995. Measuring Poverty: A New Approach. Washington, DC: National Academic Press. Clark, Colin. 1958. The economics of housework. Bulletin of the Oxford Institute of Statistics 20: 205–11. Clarkberg, Marin and Phyllis Moen. 2001. The time-squeeze: is the increase in working time due to employer demands or employee preferences? American Behavioral Scientist 44: 1115–35. Cohen, G. A. 1983. The structure of proletarian unfreedom. Philosophy & Public Affairs 12: 3–33. 1989. On the currency of egalitarian justice. Ethics 99: 906–44. Coleman, M. T. and J. Pencavel. 1993a. Changes in work hours of male employees 1945–1988. Industrial and Labor Relations Review 46: 262–81. 1993b. Trends in market work behavior of women since 1940. Industrial and Labor Relations Review 46: 653–76. Dapre`, Bob. 2006. A Compendium of Legislative Changes in Social Security 1983–2000: Part 1, 1983–1993. Occasional Paper No. 13. Canberra: Department of Families, Community Services and Indigenous Affairs. Darwall, Stephen. 2006. The value of autonomy and autonomy of the will. Ethics 116: 263–84.

Bibliography

431

Deaton, Angus. 1997. The Analysis of Household Surveys: A Microeconomic Approach to Development Policy. Baltimore, MD: Johns Hopkins University Press for the World Bank. Deeming, Christopher. 2005. Minimum income standards: how might budget standards be set for the UK? Journal of Social Policy 34: 619–36. DeLay, Tom. 1998. The Reagan legacy lives on. Special Order Speech before the US House of Representatives, Washington, DC, on 3 February, 1998. Available at: nlc.natca.org/archive/delay.htm. De Swaan, Abram. 1988. In Care of the State. Oxford: Polity. Dickens, Charles. 1854. Hard Times. Harmondsworth: Penguin, 1969. Diener, Ed. 1984. Subjective well-being. Psychological Bulletin 95 (3): 542–75. Diener, Ed, Carol L. Gohm, Eunkook M. Shu and Shigehiro Oishi. 2000. Similiarity of the relations between marital status and subjective wellbeing across cultures. Journal of Cross Cultural Psychology 31: 419–36. Diener, Ed, Eunkook M. Shu, Richard E. Lucas and Heidi L. Smith. 1999. Subjective well-being: three decades of progress. Psychological Bulletin 125: 276–302. Donovan, Nick and David Halpern. 2002. Life Satisfaction. London: Cabinet Office Strategy Unit. Available at: www.wcfia.harvard.edu/ conferences/socialcapital/papers.asp. Douglas, Mary and Brian Isherwood. 1979. The World of Goods. London: Allen Lane. Dow, Greg K. and F. Thomas Juster. 1985. Goods, time and well-being: the joint dependence problem. In Juster and Stafford 1985b, pp. 397–414. Doyal, Len and Ian Gough. 1991. A Theory of Human Need. Basingstoke: Macmillan. Dubay, Curtis S., Sumeet Sagoo and Scott A. Hodge. 2005. America celebrates Tax Freedom Day1. Special Report No. 134 (April). Available at: www.taxfoundation.org/taxfreedomday. Duncan-Myers, Audrey and Ruth Huebner. 2000. Relationship between choice and quality of life among residents in long-term-care facilities. American Journal of Occupational Therapy 54 (5): 504–8. Dusenberry, James. 1949. Income, Savings and the Theory of Consumer Behavior. Cambridge, MA: Harvard University Press. Dworkin, Gerald. 1982. Is more choice better than less? Midwest Studies in Philosophy 7: 47–61. Reprinted pp. 62–84 in Dworkin 1988. 1988. The Theory and Practice of Autonomy. Cambridge University Press. Dworkin, Ronald M. 1978. Liberalism. Pp. 113–43 in Public and Private Morality, ed. Stuart Hampshire. Cambridge University Press. 1981. What is equality? Philosophy & Public Affairs 10: 185–246, 283–345.

432

Bibliography

2000. The Sovereign Virtue. Cambridge, MA: Harvard University Press. Easterlin, Richard. 2001. Income and happiness: towards a unified theory. Economic Journal 111: 465–84. Edwards, Meredith. 2001. From child maintenance to child support: an unlikely policy. Pp. 57–96 in Social Policy, Public Policy: From Problem to Practice. Crows Nest, NSW: Allen & Unwin. Eekelaar, John. 1991. Regulating Divorce. Oxford: Clarendon Press. Ehrenreich, Barbara. 2002. Nickel and Dimed: On (Not) Getting By in America. New York: Owl Books, Henry Holt. Eisenstein, Hester. 1984. Contemporary Feminist Thought. London: Allen & Unwin. Eisner, Robert. 1988. Extended accounts for national income and product. Journal of Economic Literature 26: 1611–84. Ekirch, A. Roger. 2001. Sleep we have lost: pre-industrial slumber in the British Isles. American Historical Review 106 (2: April): 343–86. 2005. At Day’s Close: Night in Times Past. New York: Norton. Engels, Friedrich. 1844. The Condition of the Working Class in England, ed. and trans. W. O. Henderson and W. H. Chaloner. Oxford: Blackwell, 1958. England, Paula and Nancy Folbre. 1999. The cost of caring. Annals of the American Academy of Political and Social Science 561 (1): 39–51. Epstein, Cynthia Fuchs and Arne L. Kallenberg, eds. 2004. Fighting for Time: Shifting Boundaries of Work and Social Life. New York: Russell Sage Foundation. Erikson, Robert. 1993. Descriptions of inequality: the Swedish approach to welfare research. Pp. 67–83 in The Quality of Life, ed. Martha Nussbaum and Amartya Sen. Oxford: Clarendon Press. ˚ berg. 1987. The nature and distribution of Erikson, Robert and Rune A ˚ berg, Welfare in Transition: A welfare. Pp. 1–13 in Erikson and A Survey of Living Conditions in Sweden 1968–1981. Oxford: Clarendon Press. Eriksson, Lina and Robert E. Goodin. 2007. The measuring rod of time: the example of Swedish day-fines. Journal of Applied Philosophy 24: 125–36. Eriksson, Lina, James Mahmud Rice and Robert E. Goodin. 2007. Temporal aspects of life satisfaction. Social Indicators Research 80: 511–33. Esping-Andersen, Gøsta. 1990. The Three Worlds of Welfare Capitalism. Oxford: Polity. 1999. Social Foundations of Post Industrial Economies. Oxford University Press. 2002. Why We Need a New Welfare State. Oxford University Press.

Bibliography

433

Esping-Andersen, Gøsta and Walter Korpi. 1984. Social policy as class politics in post-war capitalism: Scandinavia, Austria and Germany. In Goldthorpe 1984, pp. 179–208. Etzioni, Amitai. 1993. The Spirit of Community. New York: HarperCollins. European Commission (EC), Social Protection Committee. 2001. Report on Indicators in the Field of Poverty and Social Exclusion. Available at: http://ec.europa.eu/employment_social/social_inclusion/ indicators_en.htm. Eurostat. 2000a. Guidelines on Harmonized European Time Use Surveys. Luxembourg: Eurostat. Available at: www.scp.nl/onderzoek/tbo/ achtergronden/guidelines.pdf. 2000b. Survey on Time Use: Activity Coding List. Final draft, replacing Guidelines on Harmonized European Time Use Surveys. Doc E2/TUS/ 5/00. Luxembourg: Eurostat. Available at: unstats.un.org/unsd/methods/ timeuse/tusresource_manuals/ eurostat_man.pdf. Feinberg, Joel. 1980. The idea of a free man. Pp. 3–29 in Feinberg, Rights, Justice and the Bounds of Liberty. Princeton University Press. Figart, D. and L. Golden. 1998. The social economics of work time: an introduction. Review of Social Economy 56: 411–24. Fineman, Martha Albertson. 1991. The Illusion of Equality: The Rhetoric and Reality of Divorce Reform. University of Chicago Press. Fisher, Irving. 1928. The Money Illusion. New York: Adelphi. Folbre, Nancy. 1982. Exploitation comes home: a critique of the Marxian theory of family labor. Cambridge Journal of Economics 6: 317–29. 1991. The unproductive housewife: her evolution in nineteenth century economic thought. Signs 16: 463–84. 1994. Who Pays for the Kids? Gender and the Structures of Constraint. London: Routledge. 2001. The Invisible Heart: Economics and Family Values. New York: New Press. 2006. Greed and Lust: A History of Economic Ideas. Mimeo, Department of Economics, University of Massachusetts, Amherst. Forthcoming from Oxford University Press. 2008. Valuing Children: Rethinking the Economics of the Family. Cambridge, MA: Harvard University Press. Forsse´n, Katja. 1997. Gender equality and welfare states models: outcomes perspective. Series B:14/1997. Turku, Finland: Department of Social Policy, University of Turku. 1998. Children, Families and the Welfare State: Studies on the Outcomes of Finnish Family Policy. STAKES Research Report 92. Helsinki: STAKES National Research and Development Centre for Welfare and Health.

434

Bibliography

Forsse´n, Katja and Mia Hakovirta. 1998. Work incentives in single parent families. Series B:15/1998. Turku, Finland: Department of Social Policy, University of Turku. Foucault, Michel. 1977. Discipline and Punish: The Birth of the Prison, trans. Alan Sheridan. Harmondsworth: Penguin; originally published 1975. Frank, Robert H. 1985. Choosing the Right Pond. Oxford University Press. Fraser, Nancy. 1994. After the family wage: gender equity and the welfare state. Political Theory 22: 591–618. Fraser, Nancy and Linda Gordon. 1992. Contract versus charity: why is there no social citizenship in the United States? Socialist Review 22 (3): 45–68. 1994. A genealogy of dependency: tracing a keyword of the US welfare state. Signs 19: 309–36. Frey, Bruno S. and Alois Stutzer. 2002a. Happiness & Economics. Princeton University Press. 2002b. What can economists learn from happiness research? Journal of Economic Literature 40 (2: June): 402–35. Friedman, Milton. 1962. Capitalism and Freedom. University of Chicago Press. Friedman, Milton and Rose Friedman. 1980. Free to Choose. Harmondsworth: Penguin. Fromm, Erich. 1941. Escape from Freedom. New York: Holt, Rinehart and Winston. Galston, William A. 1991. Liberal Purposes. Cambridge University Press. Garfinkel, Irving and Robert H. Haveman. 1977. Earnings capacity, economic status and poverty. Journal of Human Resources 12: 49–70. 1978. Earnings Capacity, Poverty and Inequality. New York: Academic Press. Garfinkel, Irwin, Jennifer L. Hochschild and Sara S. McLanahan, eds. 1996. Social Policies for Children. Washington, DC: Brookings Institution. Gauthier, Anne. 1996. The State and the Family: A Comparative Analysis of Family Policies in Industrialized Countries. Oxford: Clarendon Press. Germany. 1949. Basic Law for the Federal Republic of Germany of 23 May 1949. Pp. 127–207 in Comparing Constitutions, ed. S. E. Finer, Vernon Bogdanor and Bernard Rudden. Oxford University Press. Gershuny, Jonathan. 1992. Are we running out of time? Futures 24 (1: Jan./ Feb.): 3–22. 2000. Changing Times: Work and Leisure in Post-Industrial Society. Oxford University Press. 2005a. Business as the badge of honor for the new superordinate working class. Social Research 72: 287–314.

Bibliography

435

2005b. Time Allocation and the Comprehensive Accounting of Economic Activity. ISER Working Paper No. 2005–8. Colchester: University of Essex. Gershuny, Jonathan and John P. Robinson. 1988. Historical changes in the household division of labor. Demography 25: 537–52. Gershuny, Jonathan and Oriel Sullivan. 1998. The sociological uses of timeuse diary analysis. European Sociological Review 14: 69–85. 2003. Time use, gender, and public policy regimes. Social Politics 10: 205–28. Gilligan, Carol. 1982. In a Different Voice: Psychological Theory and Women’s Development. Cambridge, MA: Harvard University Press. Glass, D. V. 1940. Population Policies and Movements in Europe. Oxford: Clarendon Press. Goldschmidt-Clermont, Luisella. 1993. Monetary valuation of non-market productive time – methodological considerations. Review of Income and Wealth 39: 419–33. Available at: www.roiw.org/1993/419.pdf. Goldschmidt-Clermont, Luisella and Elisabetta Pagnossin-Aligisakis. 1995. Measures of Unrecorded Economic Activities in Fourteen Countries. Human Development Report Office Occasional Paper No. 20. New York: United Nations Development Programme. Available at: hdr.undp. org/docs/publications/ ocational_papers/oc20a.htm. 1999. Households’ non-SNA production: labour time, value of labour and of product, and contribution to extended private consumption. Review of Income and Wealth 45: 519–29. Available at: www.roiw.org/1999/ 519.pdf. Goldschneider, Frances K. and Linda J. Waite. 1991. New Families, No Families? Berkeley: University of California Press. Goldthorpe, John H., ed. 1984. Order and Conflict in Contemporary Capitalism. Oxford: Clarendon Press. Goodin, Robert E. 1982. Political Theory & Public Policy. University of Chicago Press. 1985. Protecting the Vulnerable. University of Chicago Press. 1988. Reasons for Welfare. Princeton University Press. 1990a. Relative needs. Pp. 12–33 in Needs and Welfare, ed. Alan Ware and Robert E. Goodin. London: Sage. 1990b. Stabilizing expectations: the role of earnings-related benefits in social welfare policy. Ethics 100: 530–53. 1995. Utilitarianism as a Public Philosophy. Cambridge University Press. 1998. Social welfare as a collective social responsibility. Pp. 97–194 in Social Welfare and Individual Responsibility, by David Schmidtz and Robert E. Goodin. Cambridge University Press.

436

Bibliography

2001. Toward a post-productivist welfare state. British Journal of Political Science 31: 13–40. 2002. Structures of mutual obligation. Journal of Social Policy 31: 579–96. 2004. Support with strings: workfare as an ‘impermissible condition’. Journal of Applied Philosophy 21: 297–308. 2005. Responsibilities for children’s well-being. Pp. 60–88 in No Time to Lose: The Wellbeing of Australia’s Children, ed. Margot Prior and Sue Richardson. Melbourne University Press. 2008. Women’s work: its irreplaceability and exploitability. In The Illusion of Consent: Essays after Carole Pateman, ed. Daniel O’Neill, Mary Lyndon Shanley and Iris Marion Young. University Park: Pennsylvania State University Press. Forthcoming. Goodin, Robert E., Bruce Headey, Ruud Muffels and Henk-Jan Dirven. 1999. The Real Worlds of Welfare Capitalism. Cambridge University Press. Goodin, Robert E. and Deborah Mitchell, eds. 2000. The Foundations of the Welfare State. Aldershot: Edward Elgar. Goodin, Robert E., Antti Parpo and Olli Kangas. 2004. The temporal welfare state: the case of Finland. Journal of Social Policy 33: 531–52. Goodin, Robert E., James Mahmud Rice, Michael Bittman and Peter Saunders. 2005. The time-pressure illusion: discretionary time vs free time. Social Indicators Research 73: 43–70. Goodnow, Jacqueline J. and Jennifer M. Bowes. 1994. Men, Women and Household Work. Oxford University Press. Gordon, Linda. 1979. The struggle for reproductive freedom: three stages of feminism. Pp. 107–32 in Capitalist Patriarchy, ed. Zillah R. Eisenstein. New York: Monthly Review Press. 1994. Pitied But Not Entitled: Single Mothers and the History of Welfare, 1890–1935. Cambridge, MA: Harvard University Press. Gornick, Janet C. and Jerry A. Jacobs. 1996. A cross-national analysis of the wages of part-time workers: evidence from the United States, the United Kingdom, Canada and Australia. Work, Employment and Society 10 (1: March): 1–27. Gornick, Janet C. and Marcia K. Meyers. 2003. Families that Work: Policies for Reconciling Parenthood and Employment. New York: Russell Sage Foundation. Gornick, Janet C., Marcia K. Meyers and Katherin E. Ross. 1997. Supporting the employment of mothers: policy variation across fourteen welfare states. Journal of European Social Policy 7: 45–70. Griffin, James. 1986. Well-being. Oxford: Clarendon Press. Gustafsson, Siv. 1994. Childcare and types of welfare states. In Sainsbury 1994, pp. 45–61.

Bibliography

437

Haataja, A. 1998. Tasaetu, tarveharkinta vai ansioperiaate? Sosiaalipolitiikkamallit, mikrosimulaatiot ja tyo¨tto¨mien taloudellinen asema. Sosiaalipolitiikan laitos. Turun yliopiston julkaisuja. Sarja C. Osa 141. Turku. English summary: Means tested, flat rate or earnings related benefits? Social policy models, microsimulation and the economic position of the unemployed. Department of Social Policy. Annales Universitatis Turkuensis, C:141. Turku. Hakim, Catherine. 2000. Work-Lifestyle Choices in the 21st Century. Oxford University Press. 2003a. Competing family models, competing social policies. Family Matters 64: 52–61. 2003b. Public morality versus personal choice: the failure of social attitude surveys. British Journal of Sociology 54: 339–45. Hale, Sir Matthew. 1778. The History of the Pleas of the Crown. London: Sollom Emlyn. Hamermesh, D. S. and J. Lee. 2005. The Economics of Workaholism: We Should Not Have Worked on This Paper. National Bureau of Economic Research Working Paper No. 11566. Available at: www.nber.org/ papers/w11566. 2007. Stressed out on four continents: time crunch or yuppie kvetch? Review of Economics & Statistics 89: 374–83. Hancock, Keith. 1979. The first half-century of Australian wage policy. Journal of Industrial Relations 21: 129–60. Hantrais, Linda. 1993. Women, work and welfare in France. Pp. 116–37 in Women and Social Policies in Europe: Work, Family and the State, ed. Jane Lewis. Aldershot: Elgar. Harvey, Andrew S. 1984. Special time budget measures. In Harvey et al. 1984, pp. 116–35. Harvey, Andrew S., Alexander Szalai, David H. Elliott, Philip J. Stone and Susan N. Clark. 1984. Time Budget Reseach: An ISSC Workbook in Comparative Analysis. Frankfurt: Campus Verlag. Haveman, Robert H. 1993. Who are the nation’s truly poor? Problems and pitfalls in (re)defining and measuring poverty. Brookings Review 11 (1: Winter): 24–7. Haveman, Robert H. and Andrew Bershadker. 1998. Self-reliance as a poverty criterion: trends in earnings-capacity poverty, 1975–1992. American Economic Review (Papers & Proceedings) 88 (2: May): 342–7. 2001. The inability to be self-reliant. Review of Income & Wealth 47: 335–60. Haveman, Robert H. and L. F. Buron. 1993. Escaping poverty through work – the problem of low earnings capacity in the United States, 1973–88. Review of Income and Wealth 38: 1–15.

438

Bibliography

Hawrylyshyn, O. 1977. Towards a definition of non-market activities. Review of Income and Wealth 23: 79–96. Heidenheimer, Arnold J., Hugh Heclo and Carolyn Teich Adams. 1990. Comparative Public Policy: The Politics of Social Choice in America, Europe, and Japan. 3rd edn. New York: St Martin’s Press. Hernes, Helga. 1984. Welfare State and Woman Power. Oslo: Universiteitsforlaget. Hillsman, Sally T. 2002. Best practices along the criminal justice process: criminal fines as an intermediate sanction. Speech to ‘Beyond Prisons Symposium’, Kingston, Ont. Available at: www.csc-ssc/gc/ca/text/ forum/bprisons/speeches/11_e.shtml. Hirsch, Fred. 1976. Social Limits to Growth. Cambridge, MA: Harvard University Press. Hochschild, Arlie Russell. 1989. The Second Shift: Working Parents and the Revolution at Home. New York: Viking. 1997. The Time Bind. New York: Metropolitan Books. Holloway, Sue, Sandra Short and Sarah Tamplin. 2002. Household Satellite Account (Experimental) Methodology. London: Office of National Accounts, National Statistics. Available at: www.statistics.gov.uk/ hhsa/hhsa/resources/fileattachments/hhsa.pdf. Honore´, Tony. 1982. The Quest for Security: Employees, Tenants, Wives. London: Stevens. International Labour Office (ILO). 1935/1996. Convention concerning the reduction of hours of work to forty a week [signed 1935; came into force 1957]. Vol. I, pp. 261–3 in International Labour Conventions and Recommendations. Geneva: ILO. Ironmonger, Duncan. 1996. Counting outputs, capital inputs and caring labour: estimating gross household product. Feminist Economics 2 (3): 37–64. Jacobs, Jerry A. and Kathleen Gerson. 2000. The overworked American debate: new evidence comparing ideal and actual working hours. Pp. 71–95 in Work and Family: Research Informing Policy, ed. Toby Parcel and Daniel B. Cornfield. Thousand Oaks, CA: Sage. 2004. The Time Divide: Work, Family and Gender Inequality. Cambridge, MA: Harvard University Press. Jahoda, Marie, Paul F. Lazarsfeld and Hans Zeisel. 1933/1971. Marienthal: The Sociography of an Unemployed Community, trans. by the authors with John Reginait and Thomas Elsaesser. Chicago: Aldine-Atherton; originally published 1933. Jephcott, P. with N. Seear and J. H. Smith, under the direction of R. Titmuss. 1962. Married Women Working. London: Allen & Unwin.

Bibliography

439

Johansson, Sten. 1973. The Level of Living Survey: a presentation. Acta Sociologica 16: 211–19. Juster, F. Thomas. 1985a. Conceptual and methodological issues involved in the measurement of time use. In Juster and Stafford 1985b, pp. 19–32. 1985b. Preferences for work and leisure. In Juster and Stafford 1985b, pp. 333–51. Juster, F. Thomas, Paul N. Courant and Greg K. Dow. 1985. A conceptual framework for the analysis of time allocation data. In Juster and Stafford 1985b, pp. 113–31. Juster, F. Thomas and Frank P. Stafford, eds. 1985a. Introduction and overview. In Juster and Stafford 1985b, pp. 1–18. 1985b. Time, Goods and Well-Being. Ann Arbor: Survey Research Centre, Institute for Social Research, University of Michigan. 1991. The allocation of time: empirical findings, behavioral models and problems of measurement. Journal of Economic Literature 29: 471–552. Kahneman, Daniel, Ed Diener and Norbert Schwarz, eds. 1999. Well-Being: The Foundations of Hedonic Psychology. New York: Russell Sage Foundation. Kajanoja, Jouko. 1999. Lasten pa¨iva¨hoito investointina [Child Day Care as an Investment]. VATT-tutkimuksia 50. Helsinki: Valtion taloudellinen tutkimuskeskus / Government Institute of Economic Research. Kamerman, Sheila B. 2003. Welfare states, family policies, and early childhood education, care, and family support: options for the Central and Eastern European (CEE) and Commonwealth of Independent States (CIS) countries. Paper presented to the Council of Europe and UNESCO ‘Consultation Meeting on Family Support Policy in Central and Eastern Europe’, September 3–5, 2003, Budapest, Hungary. Available at: www.childpolicyintl.org/intpublications.html. Kamerman, Sheila B. and Alfred J. Kahn, eds. 1978. Family Policy: Government and Families in Fourteen Countries. New York: Columbia University Press. Kamerman, Sheila B., Michelle Neuman, Jane Waldfogel and Jeanne BrooksGunn. 2003. Social Policies, Family Types and Child Outcomes in Selected OECD Countries. OECD Social, Employment and Migration Working Papers, No. 6. Paris: OECD. Available at: www.childpolicy. org/oecd-workingpaper_6_2003.pdf. Kangas, Olli. 1992. The politics of universalism: the case of Finnish sickness insurance. Journal of Social Policy 21: 25–52. Kangas, Olli and Joakim Palme. 1996. The development of occupational pensions in Finland and Sweden: class politics and institutional

440

Bibliography

feedbacks. Pp. 211–40 in Privatization of Social Policy, ed. Michael Shalev. London: Macmillan. Kant, Immanuel. 1785. Groundwork of the Metaphysics of Morals. In Kant, Practical Philosophy, trans. and ed. Mary J. Gregor. Cambridge University Press, 1996. Karasek, R. 1979. Job demands, job latitude and mental strain: implications for job redesign. Administrative Science Quarterly 24: 285–308. Keynes, John Maynard. 1930. Economic possibilities for our grandchildren. Pp. 358–73 in Essays in Persuasion. London: Macmillan, 1931. King, Desmond. 1995. Actively Seeking Work? The Politics of Unemployment and Welfare Policy in the United States and Great Britain. University of Chicago Press. King, Desmond S. and Jeremy Waldron. 1988. Citizenship, social citizenship and the defence of welfare provision. British Journal of Political Science 18: 415–43. Korpi, Walter. 1983. The Democratic Class Struggle. London: Routledge and Kegan Paul. 1989. Power, politics and state autonomy in the growth of social citizenship: social rights during sickness in eighteen OECD countries since 1930. American Sociological Review 54: 309–28. 1990. The development of the Swedish welfare state in a comparative perspective. Stockholm: Swedish Institute. Reprinted in Goodin and Mitchell 2000, vol. II, pp. 314–23. 2000. Faces of inequality: gender, class and patterns of inequalities in different types of welfare states. Social Politics 7: 127–91. Koven, Seth and Sonya Michel. 1990. Womanly duties: maternalist policies and the origins of welfare states in France, Germany, Great Britain and the United States, 1880–1920. American Historical Review 95: 1076–108. Kymlicka, Will. 2001. Contemporary Political Philosophy: An Introduction. 2nd edn. Oxford University Press. Lai, D. W. and J. R. McDonald. 1995. Life satisfaction of Chinese elderly immigrants in Calgary. Canadian Journal on Aging 14: 536–52. Lake, Marilyn. 1992. Mission impossible: how men gave birth to the Australian nation – nationalism, gender and other seminal acts. Gender and History 4 (3: Autumn): 305–22. Land, Hilary. 1994. The demise of the male breadwinner – in practice but not in theory: a challenge for social security systems. Pp. 100–15 in Social Security in Transition, ed. Sally Baldwin and Jane Falkingham. New York: Harvester-Wheatsheaf. Land, Hilary and Hilary Rose. 1985. Compulsory altruism for some or an altruistic society for all. Pp. 74–96 in In Defence of Welfare, ed. Philip Bean et al. London: Tavistock.

Bibliography

441

Landers, Renee M., James B. Rebitzer and Lowell J. Taylor. 1996. Rat race redux: adverse selection in the determination of work hours in law firms. American Economic Review 86: 329–48. Langer, E. J. and J. Rodin. 1976. The effects of choice and enhanced personal responsibility for the aged. Journal of Personality and Social Psychology 34: 191–8. Lasch, Christopher. 1977. Haven in a Heartless World: The Family Besieged. New York: Basic Books. Layard, Richard. 2005. Happiness: Lessons from a New Science. London: Allen Lane. Lebergott, Stanley. 1984. The Americans: An Economic Record. New York: Norton. Lefaucheur, Nadine. 1996. ‘At risk families’ and ‘family risks’: a brief history of social welfare provision for single parent families in France. Vol. I, pp. 419–38 in Comparing Social Welfare Systems in Europe, ed. Bruno Palier. Paris: MIRE. Lenoir, Re´mi. 1991. Family policy in France since 1938. In Ambler 1991, pp. 144–86. Lewis, Jane. 1980. The Politics of Motherhood: Child and Maternal Welfare in England, 1900–1939. London: Croom Helm. 1986. Lone parent families: politics and economics. Journal of Social Policy 18: 595–600. 1990. Women’s two roles: Myrdal, Klein and post-war feminism. Pp. 167–88 in British Feminism in the Twentieth Century, ed. Harold Smith. Cheltenham: Elgar. 1992. Gender and welfare regimes. Journal of European Social Policy 2: 159–73. 1997. Gender and welfare regimes: further thoughts. Social Politics 4: 160–77. 2001. The End of Marriage? Individualism and Intimate Relations. Cheltenham: Elgar. Lewis, Jane and Gertrud A˚stro¨m. 1992. Equality, difference and state welfare: labor markets and family policies in Sweden. Feminist Studies 18: 59–87. Liljestro¨m, Rita. 1978. Sweden. In Kamerman and Kahn 1978, pp. 19–48. Linder, Steffan B. 1970. The Harried Leisure Class. New York: Columbia University Press. Lindgren, Jarl. 1978. Finland. In Kamerman and Kahn 1978, pp. 270–94. Luce, R. Duncan and Howard Raiffa. 1957. Games and Decisions. New York: Wiley. Lundberg, Shelly. 1985. Tied wage-hours offers and the endogeneity of wages. Review of Economics & Statistics 67: 405–10.

442

Bibliography

Lundberg, Shelly and Robert A. Pollak. 1994. Noncooperative bargaining models of marriage. American Economic Review (Papers & Proceedings) 84 (2: May): 132–7. Lynd, Robert S. and Helen Merrell Lynd. 1937. Middletown in Transition. New York: Harcourt, Brace. Machan, Tibor R. 2000. Tax slavery. Available at: www.mises.org/ story/410. Macintyre, Stuart. 1985. Winners & Losers: The Pursuit of Social Justice in Australian History. Sydney: Allen & Unwin. MacKinnon, Catharine A. 1989. Toward a Feminist Theory of the State. Cambridge, MA: Harvard University Press. McRae, Susan. 2003. Constraints and choices in mothers’ employment careers: a consideration of Hakim’s Preference Theory. British Journal of Sociology 54: 317–38. Mahoney, Barry with Judith A. Greene, Julie Eigler and Joan B. Green. 1996. How to Use Structured Fines (Day Fines) as an Intermediate Sanction. US Department of Justice, Bureau of Justice Assistance Monograph NCJ 156242. Washington, DC: Bureau of Justice Assistance. Available at: www.vera.org/publications/publications_5. asp?publication_id=96. Malthus, T. R. 1826. An Essay on the Principles of Population. 6th edn. London: J. Murray. Marmot, Michael G. 2004. The Status Syndrome: How Social Standing Affects our Health and Longevity. New York: Henry Holt. Marmot, Michael G., George Davey Smith, Stephen Stansfeld, Chandra Patel, Fiona North, Jenny Head, Ian White, Eric Brunner and Amanda Feeney. 1991. Health inequalities among British civil servants: the Whitehall II study. The Lancet 337: 1387–93. Marshall, T. H. 1949. Citizenship and social class. Reprinted pp. 70–134 in Marshall, Class, Citizenship and Social Development. University of Chicago Press, 1963. Marx, Karl. 1858/1972. Grundrisse, ed. and trans. David McLellan. New York: Harper & Row, 1972. 1858/1973. Grundrisse, trans. Martin Nicholaus. London: Penguin, 1973. 1867/1906. Capital, ed. Frederick Engels, trans. Samuel Moore and Edward Aveling, rev. Ernest Untermann. New York: Modern Library, 1906. 1875/1978. Critique of the Gotha Program. In The Marx–Engels Reader, ed. Robert C. Tucker. 2nd edn. New York: Norton. Masnick, George S. and Mary Jo Bane. 1980. The Nation’s Families. Boston: Auburn Press.

Bibliography

443

Maume, David J. Jr and Marcia L. Bellas. 2001. The overworked American or the time bind: assessing competing explanations for time spent in paid labor. American Behavioral Scientist 44: 1137–56. Merz, Joachim. 2002. Time and economic well-being – a panel analysis of desired versus actual working hours. Review of Income and Wealth 48: 317–46. Mill, John Stuart. 1869. The Subjection of Women. Pp. 425–58 in John Stuart Mill, Three Essays, ed. Richard Wollheim. Oxford: Clarendon Press, 1975. Miller, Pavla. 2005. Useful and priceless children in contemporary welfare states. Social Politics 12: 3–41. Moene, Karl Ove and Michael Wallerstein. 1995. How social democracy worked: labor-market institutions. Politics & Society 23: 185–212. Morgan, Kimberly J. 2005. The ‘production’ of child care: how labor markets shape social policy and vice versa. Social Politics 12: 243–63. 2006. Working Mothers and the Welfare State: Religion and the Politics of Work-family Policies in Western Europe and the United States. Stanford University Press. Morris, Norval and Michael Tonry. 1990. Between Prison and Probation: Intermediate Punishments in a Rational Sentencing System. Oxford University Press. Moynihan, D. P. 1995. ‘It will shame the Congress’: a statement on the Senate floor on September 16, 1995. New York Review of Books 62 (16): 71–2. Murray, Charles. 1984. Losing Ground: American Social Policy, 1950–80. New York: Basic. Myrdal, Alva. 1944. Nation and Family: The Swedish Experiment in Democratic Family and Population Policy. London: Kegan Paul, Trench, Trubner. Myrdal, Alva and Viola Klein. 1956. Women’s Two Roles. London: Routledge & Kegan Paul. Myrdal, Alva and Gunnar Myrdal. 1934. Kris i befolkningsfra˚gan. Stockholm: Bonniers. Myrdal, Gunnar. 1940. Population: A Problem for Democracy, The Godkin Lectures, 1938. Cambridge, MA: Harvard University Press. Nelson, Barbara J. 1990. The origins of the two-channel welfare state: Workmen’s Compensation and Mothers’ Aid. Pp. 123–51 in Women, the State and Welfare, ed. Linda Gordon. Madison: University of Wisconsin Press. Nichols, Shaun, Stephen Stich, Alan Leslie and David Klein. 1994. Varieties of off-line simulation. Reprinted pp. 39–74 in Theories of Theories of Mind, ed. P. Carruthers and P. Smith. Cambridge University Press, 1996.

444

Bibliography

Niemi, Iris. 1993. Systematic error in behavioural measurement: comparing results from interview and time budget studies. Social Indicators Research 30: 229–44. Niemi, Iiris and Hannu Pa¨a¨kko¨nen. 2001. Time Use Changes in Finland through the 1990s. Culture and the Media 2002:2. Helsinki: Statistics Finland. Nowotny, Helga. 1994. Time: The Modern and Postmodern Experience, trans. Neville Plaice. Oxford: Polity. Nozick, Robert. 1974. Anarchy, State and Utopia. Oxford: Blackwell. Nussbaum, Martha C. 1994. The Therapy of Desire: Theory and Practice in Hellenistic Ethics. Princeton University Press. 2006. Frontiers of Justice: Disability, Nationality, Species Membership. Cambridge, MA: Harvard University Press. O’Connor, Julia S. 1993. Gender, class and citizenship in the comparative analysis of welfare regimes: theoretical and methodological issues. British Journal of Sociology 44: 501–18. O’Connor, Julia S., Ann S. Orloff and Sheila Shaver. 1999. States, Markets, Families: Gender and Social Policy in Australia, Canada, Great Britain and the United States. Cambridge University Press. Offe, Claus and Rolf G. Heinze. 1992. Beyond Employment: Time, Work, and the Informal Economy, trans. Alan Braley. Oxford: Polity. Organisation for Economic Co-operation and Development (OECD). 1993. Education at a Glance. Paris: OECD. 1995. Household Production in OECD Countries: Data Sources and Measurement Methods. Paris: OECD. 1996. Education at a Glance. Paris: OECD. 1997a. Education at a Glance. Paris: OECD. 1997b. Making Work Pay: Taxation, Benefits, Employment and Unemployment. Paris: OECD. National Accounts, 1983–1995, Volume II: Detailed Tables. Paris: OECD. 2001. OECD Employment Outlook, 2001. Paris: OECD. 2002a. OECD Employment Outlook, 2002. Paris: OECD. ‘Statistical annex’ available at: www.oecd.org/pdf/M00032000/M0003207.pdf. 2002b. Society at a Glance, 2002. Paris: OECD. ‘Statistical indicators and related data’. Available at: www.oecd.org/EN/document/0,,\\.ENdocument-notheme-5-no-1-22006-0,00.html. 2003. Education at a Glance. Paris: OECD. 2005a. OECD Employment Outlook, 2005. Paris: OECD. www.oecd.org/ document/1/0,2340,en_2649_201185_34855489_1_1_1_1,00.html. 2005b. Society at a Glance, 2005. Paris: OECD. Available at: www.oecd.org/ document/24/0,2340, en_2825_497118_2671576_1_1_1_1,00.html.

Bibliography

445

2005c. What are Equivalence Scales? Paris: OECD. Available at: www.oecd.org/dataoecd/61/52/35411111.pdf. Orloff, Ann Shola. 1993. Gender and the social rights of citizenship: the comparative analysis of gender relations and welfare states. American Sociological Review 58: 303–28. 1997. Comments on Jane Lewis’s ‘Gender and Welfare Regimes: Further Thoughts’. Social Politics 4: 188–202. 2002a. Explaining US welfare reform: power, gender, race and the US policy legacy. Critical Social Policy 22: 96–118. 2002b. Women’s Employment and Welfare Regimes. Social Policy & Development Paper No. 12. Geneva: United Nations Research Institute for Social Development. Orshansky, Mollie. 1969. How poverty is measured. Monthly Labor Review 92: 37–41. Ostner, Ilona. 1993. Slow motion: women, work and the family in Germany. Pp. 92–115 in Women and Social Policies in Europe: Work, Family and the State, ed. Jane Lewis. Aldershot: Elgar. Ostner, Ilona and Jane Lewis. 1995. Gender and the evolution of European social policies. Pp. 159–93 in European Social Policy, ed. Stephan Liebfried and Paul Pierson. Washington, DC: Brookings Institution. Oswald, Andrew J. 1997. Happiness and economic performance. Economic Journal 107: 1815–31. Padavic, Irene and Barbara Reskin. 1994. Women and Men at Work. London: Pine Forge Press. Palda, Filip. 1997. Fiscal churning and political efficiency. Kyklos 50: 189–206. Palme, Joakim. 1990. Models of old-age pensions. Pp. 104–25 in Needs and Welfare, ed. Alan Ware and Robert E. Goodin. London: Sage. Parker, Stephen. 1990. The Australian Child Support scheme. Family Law 20: 210–14. Pateman, Carole. 1988a. The patriarchal welfare state. Pp. 231–60 in Democracy and the Welfare State, ed. Amy Gutmann. Princeton University Press. 1988b. The Sexual Contract. Cambridge: Polity. Perlmuter, L. C. and R. A. Monty. 1977. The importance of perceived control: fact or fantasy? American Scientist 65: 759–65. Peterson, Christopher. 1999. Personal control and well-being. In Kahneman et al. 1999, pp. 288–301. Pettit, Philip. 2001. A Theory of Freedom. Oxford: Polity. Petty, William. 1899. The Economic Writings of Sir William Petty Together with the Observations upon the Bills of Mortality More Probably by Captain John Graunt, ed. Charles Henry Hull. Cambridge University Press.

446

Bibliography

Pharr, Susan J. and Robert D. Putnam, eds. 2000. Disaffected Democracies: What’s Troubling the Trilateral Countries? Princeton University Press. Phillips, Anne. 1983. Hidden Hands: Women and Economic Policies. London: Pluto Press. Phillips, Roderick. 1988. Putting Asunder: A History of Divorce in Western Society. Cambridge University Press. Pigou, A. C. 1932. The Economics of Welfare. 4th edn. London: Macmillan. Piven, Frances Fox and Richard A. Cloward. 1971. Regulating the Poor. New York: Pantheon. Polanyi, Karl. 1944. The Great Transformation. New York: Rinehart. Pollak, Robert A. and M. L. Wachter. 1975. The relevance of the household production function and its implications for the allocation of time. Journal of Political Economy 83: 255–77. Postone, Moishe. 1978. Necessity, labour and time: a reinterpretation of the Marxian critique of capitalism. Social Research 45: 739–88. Pulkkinen, Lea, Marja Kokkonen and Ari Makiaho. 1998. Positive affectivity, self-mastery, and a sense of failure as predictors of self-assessed health. European Psychologist 3 (2): 133–42. Questlaux, Nicole and Jacques Fournier. 1978. France. In Kamerman and Kahn 1978, pp. 117–82. Radin, Margaret Jane. 1996. Contested Commodities. Cambridge, MA: Harvard University Press. Ransom, Roger L. and Richard Sutch. 1977. One Kind of Freedom: The Economic Consequences of Emancipation. Cambridge University Press. Rawls, John. 1971. A Theory of Justice. Cambridge, MA: Harvard University Press. Raz, Joseph. 1986. The Morality of Freedom. Oxford: Clarendon Press. Reagan, Ronald. n.d. Quotes. Available at: www.quotesforall.com/r/ reaganronald.htm. Rhode, Deborah L. and Martha Minow. 1990. Reforming the questions, questioning the reforms: feminist perspectives on divorce law. Pp. 191–210 in Divorce Reform at the Crossroads, ed. Stephen D. Sugarman and Herma Hill Kay. New Haven, CT: Yale University Press. Rice, James Mahmud, Robert E. Goodin and Antti Parpo. 2006. The temporal welfare state: a crossnational comparison. Journal of Public Policy 26: 195–228. Rimlinger, Gaston V. 1971. Welfare Policy and Industrialization in Europe, America and Russia. New York: Wiley. Ringen, Stein. 1988. Direct and indirect measures of poverty. Journal of Social Policy 17: 351–65. Robinson, John P. 1977. How Americans Use Time. New York: Praeger.

Bibliography

447

1985. The validity and reliability of diaries versus alternative time use measures. In Juster and Stafford 1985b, pp. 33–62. Robinson, John P. and Ann Bostrom. 1994. The overestimated workweek? What time diary measures suggest. Monthly Labor Review 117: 11–23. Robinson, John P. and Philip E. Converse. 1972. Social change reflected in the use of time. Pp. 17–86 in The Human Meaning of Social Change, ed. Angus Campbell and Philip E. Converse. New York: Russell Sage Foundation. Robinson, John P., Philip E. Converse and Alexander Szalai. 1972. Everyday life in twelve countries. In Szalai et al. 1972, pp. 113–44. Robinson, John P. and G. Godbey. 1997. Time for Life: The Surprising Ways Americans Use their Time. University Park: Pennsylvania State University Press. 1999. Time for Life: The Surprising Ways Americans Use their Time. 2nd edn. University Park: Pennsylvania State University Press. Robson, W. A. 1976. Welfare State and Welfare Society. London: Allen & Unwin. Roemer, John E. 1982. Exploitation, alternatives and socialism. Economic Journal 92: 87–107. 1988. Free to Lose. Cambridge, MA: Harvard University Press. Rones, Phillip L., Randy E. Ilg and Jennifer M. Garner. 1997. Trends in hours of work since the mid-1970s. Monthly Labor Review 120 (4: April): 3–14. Available at: www.bls.gov/opub/mlr/1997/04/art1abs.htm. Rosanvallon, Pierre. 2000. The New Social Question: Rethinking the Welfare State, trans. Barbara Harshav. Princeton University Press. Ruddick, Sara. 1980. Maternal thinking. Feminist Studies 6: 342–67. 1989. Maternal Thinking. New York: Ballantine Books. Ruggles, Patricia. 1990. Drawing the Line: Alternative Poverty Measures and their Implications for Public Policy. Washington, DC: Urban Institute. Sainsbury, Diane, ed. 1994. Gendering Welfare States. London: Sage. 1996. Gender, Equality and Welfare States. Cambridge University Press. Saunders, Peter. 1994. Welfare and Inequality: National and International Perspectives on the Australian Welfare State. Cambridge University Press. Saunders, Peter, Inge O’Connor and Timothy Smeeding. 1994. The distribution of welfare: inequality, earnings capacity and household production in comparative perspective. SPRC Discussion Papers No. 51 (Nov. 1994). Kensington, NSW: Social Policy Research Centre, University of New South Wales. Sawhill, Isabel V. 1999. Families at risk. Pp. 97–135 in Setting National Priorities: The 2000 Election and Beyond. Washington, DC: Brookings Institution.

448

Bibliography

Schor, Juliet B. 1992. The Overworked American: The Unexpected Decline of Leisure. New York: Basic Books. 1993. Preface to the paperback edition. Pp. xv–xvi in The Overworked American: The Unexpected Decline of Leisure. New York: Basic Books. 2000. Working hours and time pressure: the controversy about trends in time use. Pp. 73–86 in Working Time: International Trends, Theory, and Policy Perspectives, ed. Lonnie Golden and Deborah M. Figart. London: Routledge. Schwartz, Barry. 1975. Queuing and Waiting: Studies in the Social Organization of Access and Delay. University of Chicago Press. Scitovsky, Tibor. 1976. The Joyless Economy: An Inquiry into Human Satisfaction and Consumer Dissatisfaction. Oxford University Press. Self, Peter. 1975. Econocrats and the Policy Process: The Politics and Philosophy of Cost-Benefit Analysis. London: Macmillan. Sen, Amartya. 1980. Equality of what? Tanner Lectures on Human Values 1: 195–220. 1981. Poverty and Famines. Oxford: Clarendon Press. 1983. Poor, relatively speaking. Oxford Economic Papers 35: 153–69. 1985. Commodities and Capabilities. Amsterdam: North-Holland. 1987. The Standard of Living. Cambridge University Press. 1992. Inequality Reexamined. Cambridge, MA: Harvard University Press. 1996. Social commitment and democracy: the demands of equity and financial conservatism. Pp. 9–38 in Living as Equals, ed. Paul Barker. Oxford University Press. 1999. Development as Freedom. New York: Random House. Shafir, E., Peter A. Diamond and Amos Tversky. 1997. Money illusion. Quarterly Journal of Economics 112: 341–74. Shaver, Sheila. 2000. Inequalities, regimes and typologies. Social Politics 7: 215–19. 2002. Gender, welfare, regimes and agency. Social Politics 9: 203–11. Shubik, Martin. 1971. Games of status. Behavioral Science 16: 117–29. Silvers, Anita, David Wasserman and Mary B. Mahowald. 1998. Disability, Difference and Dilemmas: Perspectives on Justice in Bioethics and Public Policy. Lanham, MD: Rowman & Littlefield. Simon, Herbert A. 1951. A formal theory of the employment relationship. Econometrica 19: 293–305. Sirianni, Carmen. 1991. The self-management of time in post-industrial society. In Working Time in Transition, ed. Karl Hinrichs, William Roche and Carmen Sirianni. Philadelphia: Temple University Press. Skinner, Quentin. 2001. The third concept of liberty. Proceedings of the British Academy 112: 237–68.

Bibliography

449

Skocpol, Theda. 1992. Protecting Soldiers and Mothers: The Political Origins of Social Policy in the United States. Cambridge, MA: Harvard University Press. Smeeding, Timothy M. 1997. Time and public policy: why do we care and what instruments are needed? Presented to conference on ‘Timeuse, Non-market Work and Family Well-being’, Washington, DC, 20–21 Nov. 1997. Available at: www-cpr.maxwell.syr.edu/faculty/ Smeeding/papers/timepolicy.pdf. Smeeding, Timothy M. and Joseph T. Marchand. 2003. Family time and public policy in the United States. Mimeo., Center for Policy Research, University of Syracuse. Smeeding, Timothy M., P. Saunders, J. Coder, S. Jenkins, J. Fritzell, A. J. M. Hagenaars, R. Hauser and M. Wolfson. 1993. Poverty, inequality and family living standards impacts across seven nations: the effect of noncash subsidies for health, education and housing. Review of Income & Wealth 39: 229–54. Solow, Robert M. 2000. Welfare: the cheapest country. New York Review of Books 47 (5: 23 Mar.): 20–3. Sorokin, P. A. and R. K. Merton. 1937. Social time: a methodological and functional analysis. American Journal of Sociology 42: 615–29. Sorokin, Pitrim and Clarence Berger. 1939. Time-budgets of Human Behavior. Cambridge, MA: Harvard University Press. Spengler, Joseph J. 1979. France Faces Depopulation: Postulate Edition, 1936–1976. Durham, NC: Duke University Press. Stafford, Frank P. 1985. Issues in modeling behavior with time use data. In Juster and Stafford 1985b, pp. 383–96. Stark, Cynthia A. 2007. How to include the severely disabled in a contractarian theory of justice. Journal of Political Philosophy 15: 127–45. Statistics Finland. 2001. Tulonjakotilasto 1999. Tulot ja kulutus [Income Distribution Survey 1999: Incomes and Consumption] 2001: 16. Helsinki: Tilastokeskus/Statistics Finland. Stinchcombe, Arthur L. 1995. Sugar Island Slavery in the Age of Enlightenment: The Political Economy of the Caribbean World. Princeton University Press. Stinson, Linda L. 1999. Measuring how people spend their time: a time-use survey design. Monthly Labour Review 122 (8: Aug.), 12–19. Available at: http://stats.bls.gov/opub/mlr/1999/08/art3full.pdf. Stoljar, Samuel. 1989. A history of the French law of divorce. International Journal of Law and the Family 4: 137–59. Strasser, Susan. 1982. Never Done: A History of American Housework. New York: Pantheon.

450

Bibliography

Sullivan, Oriel and Jonathan Gershuny. 2001. Cross-national changes in time use: some sociological (hi)stories re-examined. British Journal of Sociology 52: 331–47. Sweden. 1931. Proposition 1931: 188. 1968. The Status of Women in Sweden. Swedish Government’s Report to the United Nations Economic and Social Council. Stockholm: Swedish Institute. Szalai, Alexander, Philip Converse, Pierre Feldheim, Erwin Scheuch and Philip Stone. 1972. The Use of Time. The Hague: Mouton. Tax Foundation. 2005. America celebrates Tax Freedom Day1. Available at: www.taxfoundation.org/taxfreedomday/. tenBroek, Jacobus and Floyd W. Matson. 1966. The disabled and the law of welfare. California Law Review 54: 809–40. Therborn, Go¨ran. 1995. European Modernity and Beyond. London: Sage. Thompson, E. P. 1967. Time, work-discipline, and industrial capitalism. Past & Present 38: 56–97. Tilton, Timothy A. 1990. The Political Theory of Swedish Social Democracy. Oxford: Clarendon Press. Titmuss, Richard M. 1955. The social division of welfare: some reflections on the search for equity. Reprinted pp. 39–59 in The Philosophy of Welfare, ed. B. Abel-Smith and K. Titmuss. London: Allen & Unwin, 1987. 1958. Essays on ‘the Welfare State’. London: Allen & Unwin. 1974. Social Policy. London: Allen & Unwin. Tobin, James. 1970. On limiting the domain of inequality. Journal of Law and Economics 13: 263–78. Tocqueville. Alexis de. 1835. Address to the Royal Academy of Cherbourg. Pp. 1–27 in Tocqueville and Beaumont on Social Reform, ed. and trans. Seymour Drescher. New York: Harper & Row, 1968. Reprinted as: Memoir on pauperism. The Public Interest 70 (1983): 102–20. Townsend, Peter. 1979. Poverty in the United Kingdom. Harmondsworth: Penguin. Travers, Peter and Sue Richardson. 1993. Living Decently: Material Wellbeing in Australia. Oxford University Press. UK Parliament. 1999. Welfare Reform and Pensions Act of 1999. In Acts of the UK Parliament 1999, c. 30. Available at: www.legislation.hmso. gov.uk/acts/acts1999/19990030.htm. UN Department of Economic & Social Affairs, Statistics Division. 2005. Guide to Producing Statistics on Time Use: Measuring Paid and Unpaid Work. ST/ESA/STAT/SER.F.93. New York: United Nations. Available at: http://unstats.un.org/unsd/publication/SeriesF/SeriesF_93E.pdf.

Bibliography

451

UN Statistics Division. 1993. System of National Accounts. New York: United Nations. Available at: unstats.un.org/unsd/sna1993;toctop.asp. US Bureau of the Census. 1965. The Statistical History of the United States from Colonial Times to the Present. Stamford, CT: Fairfield. US Congress. 1996. Personal Responsibility and Work Opportunity Reconciliation Act of 1996 [Welfare Reform Act of 1996]. PL 104–193 [HR 3734], August 22, 1996, 100 Stat 2105. Available at: http:// frwebgate.access.gpo.gov/cgi-bin/getdoc.cgi?dbname¼104_cong_public_ laws&docid¼f:publ193.104.pdf. US Council of Economic Advisers (US CEA). 1999. Families and the Labor Market, 1969–1999: Analyzing the ‘Time Crunch’. Washington, DC: Government Printing Office. Available at: http://clinton4.nara.gov/ media/pdf/famfinal.pdf. US Social Security Administration (US SSA). 1994. Social Security Programs Throughout the World. Washington, DC: Government Printing Office. 2006. Social Security Programs Throughout the World. Available at: www.ssa.gov/policy/docs/progdesc/ssptw/. Va¨isa¨nen, Paavo. 2002. Estimation procedure of the Finnish Time Use Survey 1999–2000. Paper presented to the IATUR Annual Conference, Lisbon, 15–18 Oct. 2002. Available at: http://pascal.iseg.utl.pt/cisep/IATUR/ Papers/Vaisanen102f.pdf. Van der Doef, Margot and Stan Maes. 1999. The Job Demand-Control (-Support) Model and psychological well-being: a review of 20 years of empirical research. Work and Stress 13 (2): 87–114. van der Veen, Robert and Loek Groot. 2006. Post-productivism and welfare states: a comparative analysis. British Journal of Political Science 36: 593–618. van Kersbergen, Kees. 1995. Social Capitalism: A Study of Christian Democracy and the Welfare State. London: Routledge. Van Parijs, Philippe. 1995. Real Freedom for All. Oxford: Clarendon Press. Van Praag, Bernard M. S. and Ada Ferrer-i-Carbonell. 2005. Happiness Quantified: A Satisfaction Calculus Approach. Oxford University Press. van Wanrooy, Brigid and Shaun Wilson. 2006. Convincing the toilers? Dilemmas of long working hours in Australia. Work, Employment and Society 20: 349–68. Vanek, Joann. 1974. Time spent in housework. Scientific American 231 (5 Nov.): 116–20. Varjonen, Johanna and Eeva Hamunen. 1999. Proposal for a satellite account of household production. Statistics Directorate, Organisation for Economic Co-operation and Development, STD/NA(99)17. Paris: OECD. Available at: www.oecd.org/dataoecd/61/62/2680989.pdf. Veenhoven, Ruud. 1984. Conditions of Happiness. Boston: Kluwer Academic.

452

Bibliography

1999. Quality-of-life in individualistic society. Social Indicators 48: 157–86. Vickery, Claire. 1977. The time-poor: a new look at poverty. Journal of Human Resources 12: 27–48. Walby, Sylvia. 1989. Theorizing patriarchy. Sociology 23: 213–34. 1990. Theorizing Patriarchy. Oxford: Blackwell. 1992. Post-post-modernism: theorizing social complexity. Pp. 31–52 in Destabilizing Theory, ed. Miche`le Barratt and Anne Phillips. Oxford: Polity. Walzer, Michael. 1983. Spheres of Justice. Oxford: Martin Robertson. Waring, Marilyn. 1988. If Women Counted: A New Feminist Economics. New York: Harper & Row. Warr, Peter. 1990. Decision latitude, job demands, and employee well-being. Work and Stress 4: 285–94. Warren, Charles. 1926. The Supreme Court in United States History. Rev. edn. Boston: Little, Brown. Weaver, R. Kent. 2000. Ending Welfare as We Know It. Washington, DC: Brookings Institution. Weiner, Bernard. 1985. An attributional theory of achievement motivation and emotion. Psychological Review 92: 548–73. Weir, Margaret, Ann Shola Orloff and Theda Skocpol, eds. 1988. The Politics of Social Policy in the United States. Princeton University Press. Weitzman, Lenore J. 1985. The Divorce Revolution: The Unexpected Social and Economic Consequences for Women and Children in America. New York: Free Press. Wennemo, Irene. 1994. Sharing the Costs of Children. Dissertation Series 25. Stockholm: Swedish Institute for Social Research. Whiteford, Peter and Jonathan Bradshaw. 1994. Benefits and incentives for lone parents: a comparative analysis. International Social Security Review 47 (3–4): 69–89. Whiteford, Peter, David Stanton and Matthew Gray. 2001. Families and income security: changing patterns of social security and related policy issues. Family Matters 60: 24–35. Whitehouse, Gillian. 2001. Recent trends in pay equity: beyond the aggregate statistics. Journal of Industrial Relations 43: 66–78. Wilensky, Harold L. 1961. The uneven distribution of leisure: the impact of economic growth on ‘free time’. Social Problems 9 (1): 32–56. 2002. Rich Democracies: Political Economy, Public Policy and Performance. Berkeley: University of California Press. Wilensky, Harold and C. N. Lebeaux. 1958. Industrial Society and Social Welfare. New York: Russell Sage. Williams, Joan. 2000. Unbending Gender: Why Family and Work Conflict and What to Do about It. Oxford University Press.

Bibliography

453

Wilson, John, Jane Thomson and Anthony McMahon, eds. 1996. The Australian Welfare State: Key Documents and Themes. Melbourne: Macmillan Education. Wollstein, Jarret B. 2005. We must end tax slavery now. Available at: www.free-market.net/resources/lit/end-tax-slavery.html. Wooden, Mark. 1997. Enterprise bargaining and the gender earnings gap. Australian Bulletin of Labour 23: 214–26. World Bank. 1994. Averting the Old Age Crisis: Policies to Protect the Old and Promote Growth. Oxford University Press. 2005. Human Development Report 2005: International Cooperation at a Crossroads: Aid, Trade and Security in an Unequal World. Oxford University Press for the World Bank. Available at: hdr.undp.org/ reports/global/2005/. World Health Organization (WHO). 2005. The World Health Report 2005: Make Every Mother and Child Count. Geneva: WHO. Available at: www.who.int/whr/2005/en/. Young, Iris Marion. 1995. Mothers, citizenship and independence: a critique of pure family values. Ethics 105: 535–6. Young, Michael and Peter Willmott. 1973. The Symmetrical Family: A Study of Work and Leisure in the London Region. London: Routledge & Kegan Paul. Zeckhauser, Richard J. 1973. Time as the ultimate source of utility. Quarterly Journal of Economics 87: 668–75. 1974. Risk spreading and distribution. Pp. 206–28 in Redistribution through Public Choice, ed. Harold M. Hochman and George E. Peterson. New York: Columbia University Press. Zelizer, Viviana A. 1985. Pricing the Priceless Child: The Changing Social Value of Children. New York: Basic Books. 1994. The Social Meaning of Money. New York: Basic Books.

Index

Bold type indicates a figure or table abortion 161, 255 actual spare time 82, 83, 86–8, 89–90, 91, 95–8 adoption 173 agriculture 70–1 Aid to Dependent Children (US) 172–3 Aid to Families with Dependent Children (US) 144, 172, 175 Allocation Parent Isole´ 174 Analyzing the Time Crunch (US) 77 Anglo-Saxon countries 119–20 anti-discrimination policies 134 anti-tax rhetoric 9 atomistic households 241–2, 242 divorce rule 216–17, 230–1, 235, 237, 301–3 Australia 6, 16 discretionary time 56, 63–4, 66 gender regime 155, 157–8, 169, 175–6, 184, 191 household regime 209, 215, 246, 247, 256 time pressure 73–4, 76 welfare regime 126–7, 129, 136, 140, 143, 144–5 Australian Arbitration Court 166 Australian Sole Parent Pension 144 Australian Time-use Survey 55 Austria 123 autonomy 26, 124 concept of 27–30 see also temporal autonomy baby-boomers 163 ‘basket of goods’ approach 116 Becker, G. 206–7 Bernadotte, Marshal Jean-Baptiste 214 birth rates 76, 160, 163–4

454

Bismarck, Otto von 117, 119 blame 109, 214 blaming victims 101–8 external forces and 102–6 income and 106–8, 107 breadwinner rules 203–8, 294–8 divorce 218–19, 232, 235, 305–11 male-breadwinner 203–5, 294–7 most-efficient breadwinner 205–8, 207, 297–8 ‘paternity penalty’ and 235 Britain 118–19 Butler, E. 15 California 215 caregivers 156, 195 Centre for Population, Poverty and Policy Studies 20 character-testing 157, 172–3 charity 172–3 Charity Organization Society 172 child-care subsidies 53, 128–30, 240–1 gender regimes and 177, 181–2, 183–4, 186, 188–9, 191, 192–3 welfare regimes and 134, 136, 139–41, 143, 149 childless 140, 142–3, 183, 192, 202 dual-earner couples 91, 93–5, 96, 110–11, 143 single-earner couples 91 children allowances 173 benefit 129, 157, 240 ‘child poverty’ 138 costs 46–7, 91 custody of 236–8, 237 as pets 217, 265–6 responsibility for 202–3

Index choice 6, 30, 264–6 conditions of 84–5, 111–12 household 199–200, 202, 239 private 254–6 time-pressure illusion and 99–101, 102–3, 108, 110 Clinton, President Bill 69 Cloward, R. A. 117 Code Napole´on 159 Collier, Mary 72 common patterns 61, 65 Commonwealth Public Service (Australia) 155 comple´ment familial 159 ‘conditionalities’ 117 Conference on the Care of Dependent Children (US) 172 Constitution (Germany) 173 consumption 79 contraception 161 conventional dual-earners see dualearner households ‘corporate associations’ 119 corporatist regimes 6, 25–6, 63–4 gender and 178, 181, 183, 186, 188, 189, 192–5 impact on stay-at-home parents 146–8, 147 underlying values 195 welfare and 121, 123–4, 124, 125–7, 130, 137, 140–1, 150 ‘cost-sharing’ approach 215 Council of Economic Advisers (US) 69, 77 country differences 65–6 gender regime 169–71, 170, 175, 175 welfare regime 125–8, 125 criminal sentencing 9 custodial sentences 12–14 custody, child 214–15, 219, 236–8, 237 day fines 13–14 ‘de-commodification index’ 126 debt peonage 32 definitions discretionary time 34, 52, 293 poverty 116 spare time 36, 52, 132, 274–5 welfare regimes 115–17 ‘demographic panic’ 162–3

455 ‘depopulation’ 158 ‘deserving poor’ 171–3 diary-based exercises 21, 21n ‘diminishing marginal utility’ 59 disabled 3n discretionary time defined 34, 52, 82, 132, 293 distribution of 61–6 importance of 60 operationalized 34–53, 293 validated as measure 54–60 see also post-government discretionary time; pregovernment discretionary time dismissal 155, 161 divorce 201 divorce rates 76 divorce rules 203, 213–23, 235, 255, 256–7, 301–24 atomistic 216–17, 230–1, 235, 237, 301–3 breadwinner 218–19, 232, 235, 305–11 equitable 268–9 financially egalitarian see financially egalitarian divorce gendered see gendered divorce self-reliant 217–18, 230, 238, 303–5 strictly egalitarian see strictly egalitarian divorce dual-earner gender regime 167–9 dual-earner households 202, 209–10, 239, 299–300 atomistic individuals and 241–2, 242 childless 91, 93–5, 96, 110–11, 241–2, 242, 243 discretionary time and 228–9, 232 equal temporal contribution and 247–9, 249 gender equality and 234–5 gender regimes and 165, 185–6, 194 gendered divorce and 243–5, 244 male breadwinners and 245–6, 246 time pressure and 75, 106, 108 with children 143, 146, 236, 242–3, 243 dual-role model 162 e´cole maternelle 129, 188, 193

456 education 128, 129, 139 nursery 161 pre-school 129, 161, 241 egalitarian rules 26, 121, 145, 202, 210–13, 229, 240, 300–1 equal monetary contribution 211–12, 300–1 equal temporal contribution 212–13, 301 see also financially egalitarian divorce; strictly egalitarian divorce employment -oriented pro-natalism 160–3, 163–4 relations 156 status 187 England 32 English Poor Law 117 environment, public 256–7 equal financial contributions 221 equal living standards 215 equal monetary contribution rule 211–12, 212n, 232, 300–1 equal temporal contribution rule 212–13, 221, 229, 232, 235, 301 equality culture of 269 gender 145, 168, 182, 240 social 26 ‘equivalence scale’ 49 ‘equivalent income’ 42 Esping-Andersen, G. 123, 124, 125, 126, 163–4 e´tat providence 119 Europe 118, 120, 123, 158 European Union (EU) 41 EU-15 74 familiaux ministers 158–9 family allowances 129, 157, 240 benefits 173 ideology 159 policies 77 structures 146 traditional values 150, 173 fathers see lone fathers; lone parents; parents

Index female-friendly see social-democratic regimes feminism 153, 154, 156, 214 financially egalitarian divorce 220–1, 232, 238, 245, 252–3, 311–7 gendered divorce and 249–50, 250 strictly egalitarian divorce and 250–1, 251 findings see major findings fines, day 13–14 Finland 6, 23–4 discretionary time 50, 56, 63–4, 66 gender regime 164, 169–71, 175, 181–4, 188, 191, 193–5 household regime and 209, 246, 247, 256 time pressure 73–4, 87, 91 welfare regime 129, 134, 135–7, 140–1, 143, 146 Finnish Time Use Survey 23, 55 ‘fiscal churning’ 120 flexible working hours 267–8 Folbre, N. 217 ‘forced labour’ 14 France 6, 269 discretionary time 63–5, 65–6 gender regime 169–71, 174–5, 178, 181–6, 189, 191–3, 195 household regime 209, 241, 246, 247, 256 maternalism and 158–60 time pressure 73–4, 76, 97 welfare regime 123, 129, 134–5, 143, 147 ‘free time’ see spare time freedom 26, 27–8, 124 French Revolution 214 Friendly Societies 119 gender equality 145, 168, 182, 240 macro-politics of 157 micro-politics of 156 gender gap 157, 169, 178–81, 180, 192 alternative household rules and 233, 234–5, 234 time pressure illusion and 89–90, 89 gender regimes 24–6, 153–76 country classification 169–71, 170 lone mothers 171–6, 175

Index maternalism 157–64 models of 164–9 patriarchy 155–7 temporal perspective 177–91 autonomy and 192–5, 194 mothers, impact on 185–91, 185, 187, 189, 190, 194 parents, impact on 182–4, 182, 184 women, impact on 178–82, 179, 180 traditional categorizations 193–4 gender roles 123 gendered divorce 220, 232, 235, 245, 252, 305 dual earners and 243–4, 244 financially egalitarian divorce and 249–50, 250 German Socio-Economic Panel (GSOEP) 57, 105 Germany 6, 13, 16 discretionary time 63–4, 65–6 gender regime 155, 169–71, 173–4, 179–86, 189, 191–5 household regime 209, 241, 246, 247, 250, 256 time pressure 73–4, 76, 90, 97, 105 welfare regime 1, 119–20, 123, 129, 137, 140–1, 145–50 government tax as time 14–16, 15 working mothers and 100 see also gender regimes; household regimes; welfare regimes ‘Great Transformation’ 118 Hakim, C. 100, 100n, 109, 111, 153–4, 164 hard paternalist line 115–16 ‘harmonization’ of free time 268 Hegel, G. W. F. 26, 124 heroic acts 109–10 historical perspective divorce 214 gender regimes 155–6, 159 lone mothers 171–6 time pressure 70–2 welfare regimes 117–22 Home Care Allowance (Finland) 164

457 home-oriented pro-natalism 158–60, 163–4 see also stay-at-home parents homemakers 203–4, 208, 219 hours worked 77 ‘house-husbands’ 207 household labour division of 203–4, 205 see also unpaid household labour household regimes 199–223, 200, 224–38, 293–324 alternative 201–3, 231 breadwinner rules 203–8, 294–8 divorce rules 213–23, 301–24 dual-earner rule 209–10, 299–300 egalitarian rules 210–13, 300–1 gender regime impact on 185–91, 185 income 59 methodology 224–8 state impact on 239–53 major 252–3 policy instruments and norms 239–41 temporal perspective 228–32, 229, 231, 233 autonomy and 254–7 changing rules 245–51, 246, 248, 249, 250, 251 changing types 241–5, 242, 243, 244 custody and 236–8, 237 gender gap, impact on 234–5, 234 ‘paternity penalty’ and 235, 236 time pressure illusion and 90–3, 92, 93–5, 93, 95–8, 96, 97 types 40, 90 welfare regime impact on 141–3, 142 hunter-and-gatherer societies 70, 204 income 131 ‘effect’ 78–9 ‘equivalent’ 42 household 59 -maintenance programmes 128 net disposable 42 poverty level 44 -sharing approach 215 surveys 19–24, 23, 324–5 time pressure and 77–80 victims and 106–8, 107

458 independence 166 female 160 ‘indoor relief’ 118 industrial relations policy 134 industrial revolution 71, 77 industrialization 118–19, 121 infant mortality 138 insurance 119–20, 127 Jahoda, M. 31n Japan 81 Kant, I. 29 kindergarten 129, 188, 193 Kingdom of Ends (Kant) 29 Korpi, W. 170 labour division of 203–4, 205 ‘forced’ 14 -market policy 134 see also paid labour; unpaid household labour Le Play, Fre´de´ric 119 leave, parental 129–30 legal matters 32–3, 203 legislation Code Napole´on 159 English Poor Law 117 Married Women’s Poverty Act (UK)32 Social Security Act (US) 172 Swedish Social Services Act (Socialtja¨nstlag) 121–2 ‘leisure class’ 79 Leo XIII, Pope 165 ‘lesser-eligibility’ 118 Lewis, J. 159, 171 liberal regimes 6, 13, 24–6, 63–4, 111 gender and 178, 188, 192–4 impact on lone parents 144–5 welfare and 122–4, 124, 125–7, 130, 137, 140, 150 life satisfaction 57–60, 58 lifestyle choices 111–12 living standards, equal 215 lone fathers 94 lone mothers 94–5, 185–6, 195, 225

Index gender-regime impact on 188–91, 189, 190 gender regimes and 171–6, 175 lone parents 64–5, 77, 91, 93–4, 203, 228 liberal regime impact on 144–5, 150 social-democratic regime impact on 145–6 state impact on discretionary time 141–2, 143 ‘luck egalitarians’ 109, 111 Luxembourg Income Study (LIS) 17, 19–23, 22n, 23, 46, 61, 82, 86, 135, 224, 324–5 macro-politics of gender 157 macroeconomic policy 134 major findings 261–3 implications 263–6, 267–70 male-breadwinner gender regime 165–7 male-breadwinner households 162, 171, 193 division of labour 201, 203–5, 240, 294–7 divorce rule 218–19, 232, 235, 253, 305–8 dual earners and 245–6, 246 gender equality and 234 modified 159 most-efficient breadwinner and 246–7, 248 time pressure illusion and 95–8, 97 manufacture 71 market-oriented regimes see liberal regimes marriage forced 255 rates 76 Married Women’s Poverty Act (UK) 32 Marx, Karl 32, 239 maternalism 157–64 employment-oriented pro-natalism 160–3, 163–4 home-oriented pro-natalism 158–60, 163–4 means-testing 127, 129, 130, 157, 174–5 measuring rods 7–19 ‘micro-politics of gender’ 156 middle ages 71

Index Mill, John Stuart 32 ‘minimally necessary’ time 37–8, 38, 48 Ministry of Social Affairs (Sweden) 160–1 models, gender regime 164–9 money 105, 131 ‘illusion’ 99 see also income; time and money mortality, infant 138 most-efficient breadwinner households 205–8, 207, 219, 229, 235, 297–8, 308–11 male bread-winners and 246–7, 248 mothers ‘dual-role’ model 162 gender-regime impact on 185–91, 185, 194 coupled 187–8, 187 unmarried 144, 178 working 110, 190 see also lone mothers Mother’s Aid (US) 157, 172–3 Mother’s Pensions (US) 172 Multinational Comparative TimeBudget Research Project 51, 75–6 Multinational Time Use Study (MTUS) 19–23, 22n, 23, 61, 77, 82, 86, 224 multiple-earner households 44 mutual benefit societies 119 Mutualite´ 119 Myrdal, A. 160–1 Myrdal, G. 160 Napoleon III 119 National Accounts 8, 10–12 National Study of Families and Households 105 ‘near-poor’ 84 necessary time 36–40, 88 ‘minimally’ 37–8, 38 in paid labour 40–8, 278–93 in personal care 50–1, 275 in unpaid household labour 48–50, 275–8 variables 205, 208–10, 212–13, 217–21, 223 necessity realms of 34–5 defined 4–6

459 ‘negative liberty’ 26 ‘new gender contract’ 163 ‘no-fault revolution’ 214–15 ‘non-agreement point’ 216 nursery education 161 ‘obedience’ 159 ‘objective welfare’ 115 obligatory tasks see necessary time ‘opportunity’ 131, 133 ‘cost method’ 11 Organisation for Economic Co-operation and Development (OECD) 8, 10, 16, 125, 157, 170, 173 orphanages 173 orphans 139, 157, 171–3 paid labour 34, 273–74 necessary time in 40–8, 278–93 time demands 73–4 women and 74, 75, 123 parenthood, voluntary 161 parents 111 dual-earner 143, 146, 242–3, 243 gender regime impact on 182–4, 182, 184 impact on discretionary time 143 parental leave 129–30 social-democratic regime impact on 145–6, 150 state impact on 137–41, 139, 141 time-pressure costs 91 time pressure illusion and 89–90, 89 see also lone parents; stay-at-home parents Parkinson’s Law 98 paternalist line, hard/soft 115–16 ‘paternity penalty’ 235, 236, 238 patriarchy 155–7 structures of 155 patterns common 61, 65 family life 76 Paul VI, Pope 165 pensions 128, 240 policies 127 systems 163 personal care 35, 271–2

460 personal care (cont.) necessary time in 50–1, 275 Piven, F. F. 448 planning 134 policies 77, 127, 134 policy instruments 188, 191 of dual-earner model 168–9 of male-breadwinner model 167 norms and 239–41 politics, women in 157 poor, deserving 171–3 poor law 139, 157, 171 ‘poor relief’ 26, 117–18, 122–3, 124 Population Commission (Sweden) 160–1 post-government discretionary time 51–3 gender regimes and 178–81, 179, 182 household regimes and 228 welfare regimes and 133–6, 135, 139, 141–2, 146 potential spare time 82, 89 ‘potential welfare’ 132 poverty child 138 defining 116 line 41–2, 43n, 48, 49, 116 research 84 time 9, 84 time and money 16–19 Powell, Enoch 121 pre-government discretionary time 51–3 gender regimes and 179–81, 179, 182, 186, 192 welfare regimes and 133–6, 135, 139–40, 142–3, 146 pre-school education 129, 161, 241 ‘preference theory’ 100 prison sentences 12–14 pro-natalism 178, 195 countries 169, 171 employment-oriented 160–3, 163–4 home-oriented 158–60, 163–4 programmes, welfare 126–7 income-maintenance 128 service-delivery 129 proportionality 109 quality of life 57–60, 58

Index Raz, J. 29 Reagan, President Ronald 14 ‘replacement cost method’ 11, 11n research poverty 84 welfare state 126 responsibilities, in divorce 218 Revenu Minimum d’Insertion 174 rights of citizenship 121 risk-pooling 119–20 role reversals 207 Roman republic 70 Roosevelt, President Theodore 172 rural life 70–1 safety net 26 Samuelson, P. 102 ‘Satanic mills’ 71 Scandinavia 13, 18, 121, 123, 128, 148, 153, 163, 211 see also Finland; Sweden Schor, J. B. 99, 101, 102–3, 109 self-reliant divorce rule 217–18, 230, 238, 303–5 service-delivery programmes 129 single-earner households 44, 146–8, 147, 185–6, 226 childless 91 see also stay-at-home parents single-member households 216, 226, 228 single parents see lone mothers; lone parents slavery 29–30, 32–3 wage 32 ‘social assistance’ 122–3 ‘social cohesion’ 26, 124 Social Democratic Party (Sweden) 161, 162 social-democratic regimes 6, 25–6, 63 gender and 178, 181–2, 183, 185–6, 187–8, 192–4 impact on parents 145–6 welfare and 122–4, 124, 125–7, 130, 135–7, 140–1, 149–50 ‘social equality’ 26, 124 Social Security Act (US) 172 social status 104 soft paternalist line 115–16 Soldaten des Arbeit 119

Index solidarity 121 spare time 51–2 actual 82, 83, 86–8, 89–90, 91, 95–8 defined 36, 52, 132, 274–5 potential 82, 89, 91 subjective time pressure and 56, 57–8 ‘special needs’ 35 standard of living 115 Statistics Finland 23 Statistics Office (UN) 10 status 127 goods 79–80 social 104 symbols 78 stay-at-home parents 178, 193, 204, 241 corporatist regime impact on 146–8, 147, 150, 164 la me`re au foyer 159 Stoics 30, 32–3 strictly egalitarian divorce 222–3, 232, 238, 252, 317–24 financially egalitarian divorce and 250–1, 251 subjective time pressure 54–6, 56, 69 subsidies 269–70 see also child-care subsidies ‘substitution effect’ 78–9 Superior Council for Natality (France) 159 surveys 19–24, 23 Sweden 6, 13–14, 269 discretionary time 63–5, 66 gender regime 156, 167, 169–71, 174, 181–4, 188–94 household regime 199, 200, 201, 209, 214, 256 maternalism and 160–3 time pressure 73–4, 76, 86–8, 94 welfare regime 123, 129, 135–6, 140–1, 146, 149 Swedish Level of Living Survey 117 Swedish Social Services Act (Socialtja¨nstlag) 121–2 ‘symmetrical household’ 210–11 System of National Accounts 10 tax codes 240 credits 129, 240

461 as government time 14–16, 15 ‘slavery’ 14 tax and transfers 53, 269–70 gender regimes and 177, 181, 183–4, 186, 188–9, 191–3 welfare regimes and 134, 136–7, 139–41, 143, 149 Tax Freedom Day 14–16, 15 temporal autonomy 9, 27–60, 133 discretionary time operationalized 34–53, 293 gender regimes and 192–5, 194 household regimes and 254–7 validation of measure 54–60 value of 27–34 welfare regimes and 149–50 Temporary Assistance to Needy Families (US) 175 Theorizing Patriarchy (Walby) 155 time and money 3–26 life satisfaction and 58 measuring rods 7–19 poverty 16–19, 17 properties, time 3–7 surveys 19–24, 23 welfare and gender regimes 24–6 welfare measures 132–3 see also discretionary time ‘time bind’ 29 ‘time crunch’ 69, 77 time poverty 9, 84 time pressure 69–80 concept of 83–6 effects of 76–7 historical perspective 70–2 income rise and 77–80 sources of 73–7, 86–8 subjective 54–6, 56, 69 see also time-pressure illusion time-pressure illusion 36, 81–98, 99–112 blaming victims 101–8 by gender and parental status 89–90, 89 by household types 90–3, 92, 93–5, 93, 95–8, 96, 97 choice/options 109–11 components of 87–8, 88 concept of 83–6 magnitude of 86–8, 87

462 time-pressure illusion (cont.) state neutrality and 111–12 see also time pressure time-use files 48, 48n studies 77, 83, 132 surveys 19–24, 23, 54–5 Tocqueville, Alexis de 118 traditional family see corporatist regimes training 128 transfers see tax and transfers transportation policies 134 travel time 46–8, 273, 291–2 United Nations Statistics Office 9 United States (US) 6, 14–15, 16 discretionary time 50, 63–5, 66 gender regime 153, 157–8, 169, 172–3, 175, 183, 188, 191–5 household regime 199, 200, 201, 204, 209, 215, 246–7, 250, 256 time pressure 69, 73–4, 75–6, 91, 94, 104–5 welfare regime 129, 137, 140, 143, 144–5 universal-breadwinner see dual-earner gender regime universality 125, 126–7 unmarried mothers 144, 178 unpaid household labour 34–5, 156, 227, 272–3 necessary time in 48–50, 275–8 time demands 74–6 urban areas 71 variation 61–5 by gender 62 by household type 62 victims see blaming victims wage rates 4–5, 59, 212–13, 226 prime-aged worker 107–8, 107 ‘slavery’ 32 working wives 207–8, 207 Walby, S. 155–6 Weber, Max 165 ‘welfare effort’ 125, 125 welfare regimes 6, 24–6, 115–30, 131–48

Index child-care provision 128–30 country classification 125–8, 125 defined 115–17 history of 117–22 temporal perspective 133–7, 135, 137 autonomy and 149–50 household types, impact on 141–3, 142 money and time 132–3 parents, impact on 137–41, 139, 141 subgroups, impact on 143–8, 147 three worlds of 122–4, 124 well-being 57–60 widows 144 and orphans 139, 157, 171–3 withdrawal rules 227–8 see also divorce rules Woman’s Labour, The: An Epistle to Mr Stephen Duck; in Answer to his Late Poem, called The Thresher’s Labour (Collier) 72 women burdens on 77 choices 100, 110 dismissal 155, 161 gender regime impact on 178–82, 179, 180 paid labour and 74, 75, 123 in politics 157 work and 72, 130, 204, 207–8, 207 work-centred 164 see also gender regimes Women’s Clubs 172 work -centred women 164 -life balance 153–4 -life trade-offs 39 mothers and 110, 190 -‘spend squirrel cage’ 79 -time flexibility 267–8 women and 72, 130, 204, 207–8, 207 ‘workfare’ 171 see also paid labour; unpaid household labour Workers’ Encyclical (Leo XIII) 165 Workmen’s Compensation (US) 172